International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878, Volume-8 Issue-5, January 2020

Assessment on Liver Disease Classification

using Medical Image Processing
A. Bathsheba Parimala, R. S. Shanmugasundaram

 smoothness of picture contains inside segments, area

Abstract: Chronic hepatic disease (CHD) is progressive fatal understanding gives facts about the item to be sketched [3].
disease which is often asymptomatic. CHD has increased One can describe image segmentation technique as
mortality and morbidity rate even in developed countries also. segregating of virtual image into separate vicinity of alike
Invasive and non-invasive methods are used to classify and stage pixels. The goal of segmentation is to convert the photograph
the CHD. In the research, using Ultra Sonographic images (US),
to a form that has extra useful statistics that can be basically
clinical finding and laboratory findings for the staging of CHD is
done. There are three stages of CHD which are Chronic Hepatitis, analysed; it is used to hit upon the numerous object and
Compensated Cirrhosis and Decompensated cirrhosis. For margin in photographs. Thus segmentation of clinical
invasive method, liver biopsy is done followed by histopathological photograph can be referred a shard hassle because pics
examinations. Results of liver biopsy have some complications. commonly have low comparison, lacking margins and
So, non-invasive procedures are used as a safe alternative for liver one-of-a-kind styles of noise [4].
biopsy.This paper presents current various methods of
segmentation based on medical liver images. And also, this paper
focuses on the work of various segmentation and classification
methods that has been proposed to diagnosis many liver diseases.
Keywords: Chronic Hepatic Disease, Computer Aided
Diagnosis, Image Processing, Medical Image Processing,
Non-invasive method.

I. INTRODUCTION

The liver is taken into consideration is one of the most Fig 1: Layout in CAD
crucial organs in the body. Without a liver, no person can live
on because; many organs inside the frame depend on the On the alternative hand, CAD structures are gears that
characteristic of the liver. Diffused liver disorder and focal enhance the detection or analysis of various diseases [5].
liver ailment are the principle kind of diseases that impacts the Diagnoses with the backing of the computer are strategies that
liver. Diffused liver disease affects the overall surface of the supply much record that help physicians to apprehend the
liver including fatty and cirrhotic liver. If the ailment clinical snap shots. As a end result, the correctness of clinical
influences most effective a small section of the liver, it's miles prognosis could be improved, and the time taken in studying
known as as focal liver disorder. Cyst, Haemangioma and an picture with the aid of traditional systems will be decreased
hepatocellular carcinoma are the example of focal liver [6]. The leading stages for detecting sicknesses the use of
disease. In latest times, because of the progression in the CAD structures are the segmentation of a lesion, extraction of
discipline of synthetic intelligence and facts era, growth functions from a lesion, and describing of sicknesses by
within the evaluation of medical pics has a splendid universal means of a classifier [7].
delivered to the early prognosis of diverse illnesses [1].
Therefore, there ought to be a Computer Aided Diagnosis II. PREPROCESSING TECHNIQUES
(CAD) machine that helps to physicians to apprehend a
medical picture. One of the most critical steps in CAD device
is the segmentation procedure that often refers back to the
definition of rationalization of precise shape. The methods of
segmentation technique, collective statistics know-how and
area knowledge to get the best result [2].
Data information taken into consideration because the
presumption about homogeneity, balance and local

Revised Manuscript Received on January 15, 2020.

* Correspondence Author
A. Bathsheba Parimala*, Research Scholar, Department of Computer Fig 2: Pre-Processing Stages
Science, Vinayaka Mission’s Research Foundation, (Deemed to be
University), Salem, Tamilnadu, India. Email: [email protected]
Dr. R. S. Shanmugasundaram, Professor, Department of Computer
Science, Vinayaka Mission’s Research Foundation, (Deemed to be
University), Salem, Tamilnadu, India.

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4167 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

The pre-processing operations [8] basically facilitates iintensity idifferences ibetween

ia ipatch iand iwindows ishifted
subsequent processing of digitized images which get
iin iseveral idirections. i
contaminated with noises during acquisition process and thus 2 SIFT PDE based SIFT ialgorithm ihas ia ilocal
image data gets corrupted, but with help of pre-processing ifeature idetector iand ilocal
techniques the information of all the distorted pixels can be ihistogram-based idescriptor. iIt
inotices isets iof iinterest ipoints
restored. The pre-processing pipeline was used which
iin ian iimage iand ifor ieach
included: 1. Non-linear filtering for noise reduction, 2. Inter – ipoint iit icalculates ia
Slice intensity variation correction, 3. Intra volume intensity ihistogram-based idescriptor
bias field correction, 4. Alignment of different modalities, 5. iwith i128 ivalues. i[15]

Linear alignment of modalities with template, 6. Nonlinear 3 SURF Based Feature ienhancement
under contour ialgorithm ithat icontains ithe
warping of modalities with template. 7. Re-sampling of
isubtraction iof ione iblurred
voxels to template coordination system (B – Spline), 8. itype iof ian ioriginal iimage
Weighted regression for inter volume intensity stan- ifrom iadded, iless iblurred iform

dardization. iof ithe ioriginal i[16]. i

4 Sobel, Differentiat The iSobel ioperator
A. Levels in Image Pre-processing Canny ion based iperforms ia i2-D ispatial
igradient imeasurement ion
The iraw iimage iis icollected ifrom ithe iscan icentre iis inot iimages. iIt iuses ia ipair iof
isuitable ifor idirect iprocessing idue ito ivarious inoise ipresent iin ihorizontal iand ivertical
ithe iimage. iSo iit iis inecessary ito ipre-process iit ibefore igradient imatrices iwhose

iexamination. iThe ilevels iare ilisted ibelow [9] idimensions iare i3×3 ifor iedge
idetection ioperations i[17].
iBlock iLevel iCanny iEdge
Table 1: Levels in Image Pre-processing iDetection iAlgorithm iprovides
No Level Process isimple iedge idetection
. ioperation. iIt ireduces ithe itime
1 Grey scale Almost iall iImage iprocessing itechniques iis iand imemory iconsumption.
conversion ilargely ifocused ion igray ilevel itransformation iThe iBlock iLevel iCanny iedge
itechniques ias igray ilevel itransformation ioperate idetection ialgorithm iis ithe
idirectly ion ipixels. iA igray ilevel iimage ispecial ialgorithm ito icarry iout
igenerally iinvolves i256 ilevels iof igray, iso ifrom ithe iedge idetection iof ian
ithe ihistogram, ithe ihorizontal iaxis ispans ifrom i0 iimage i[18].
ito i255[10] 5 SURF PDE based SURF ialgorithm iis iused
2 Image To ienlarge ia ideduced ithe igiven iimage isize idue ito iits idominant
Resize iin ipixel iformat. iImage iinterpolation ican ibe iattributes, iwith iscale
idivided iinto itwo idifferent iways, i1) iImage iinvariance, itranslation
idown isampling iand i2) iImage iup isampling. iIt iinvariance, ilighting ifixtures
iis inecessary iwhen iresizing ithe idata ifor iinvariance, ievaluation
imatching ieither ithe ispecific icommunication iinvariance, iand irotation
ichannel ior ithe ioutput idisplay i[11]. iinvariance i& iit icould icome
3 Noise Image inoise iis igenerally iregarded ias ian iacross iobjects iin iimages
Removal iundesirable iimage iduring iimage iacquisition. itaken iunder iexceptional
iNoise iis ioften idefined ias ithe iuncertainty iin ia iextrinsic iand iintrinsic
isignal idue ito irandom ifluctuation iin ithat isignal. isettings i[19].
iFor ian iexample, ithe isalt iand ipepper inoise ican
ibe ieliminated iby iusing idark iframe isubtraction
iand iby iinterpolating iaround idark ibright
Extracted ifeatures iare ianalysed ithrough ithe iutility iof
ipixels[12]. wavelet ipacket iconvert ifor iB-mode iultrasound iliver ipics. iFor
4 Filtering To ienhance ian iimage, iin ifilters iare imainly iyears, ithe iscientific icommunal ihas ibeen ilooking ito
iused ito isuppress ieither ithe ihigh ifrequencies iin idetermine iout ian iactive itechnique iof ifeature iextraction ifor
ithe iimage ii.e. ismoothing ithe iimage ior ithe ilow
iilluminating itexture iinfo iand ithe isucceeding iclassifications.
ifrequencies ii.e. ienhancing ior idetecting iedges
iMethods iof ifeature iextraction ican ibe imore ior iless
iin ithe iimage. iSome iof ithe ifilters iare iWiener,
iMedian, iGaussian ifilters i[13]. icharacterised iinto i3 icategories: ispatial-domain ibased,

5 Contrast It iis ithe idifference ibetween imaximum iand ifrequency iarea iprimarily ibased iand iversion ibased itotally
Enhancemen iminimum ipixel iintensity. iThe imajor iobjective
i[20]. iThe ifeature iextraction imakes iuse iof ithe iwavelet ipacket
t iof ithe icontrast ienhancement isystem iis itwofold:
itransformation. iThe iexperiment ihas iproved ithat ifunctions
iLocally iadaptive ihistogram iequalization iand
ireduction iof idesired iobject isuch ias inoise iand iderived ifrom iwavelet ipacket iremodel ican ihigher

iblocking iobjects i[9]. idiscriminate inormal ielegance ifrom icirrhosis iwhen ibeing

icompared iwith isystems iobtained ifrom iwavelet irework.

III. FEATURE EXTRACTION TECHNIQUES Examining iliver icross-section iimages iis ia imanual iand
ilaborious itask, islowing idown icritical iresearch itoward
Table 2: Algorithms in Feature Detector
S.No Algorithm Classification Process ifinding idifferent icures ifor ipatients iwith iend-stage iliver

. idisease. iIn ithe iresearch ipaper, iapproaches ito ispontaneously

1 Harris Gradient The iHarris icorner idetector icount ihepatocytes icells, icount inuclei iand iclassify iliver
based iis ia istandard itechnique ifor
ivessel itypes, iusing iinput iimages iof icell iboundary iand icell
ilocating iinterest ipoints ion ian
iimage i[14].The imain iidea iis inuclei ibased ion ia idataset iof i21 iimages ipresented. I

ibased ion iMoravec’s idetector

ithat irelies ion ithe
iautocorrelation ifunction iof
ithe iimage ifor imeasuring ithe

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4168 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

Compared ito ia itrained iresearcher, ithe iapproaches iare iable idescribed iare ibased ion ithe ifirst iorder istatistics,
icount icells, isegment ioverlapping inuclei, iand iclassify ivessel ico-occurrence imatrix, iwavelet itransform iattenuation iand
itypes, icontaining iportal ivein, icentral ivein, iand ibile iduct iwith iback iscattering iparameters iand ico-efficient ipre-processing

irealistic iprecision iand iaccuracy. iFuture iwork iincludes ialgorithm, ifeature iExtraction iand iSelection, iClassification

iimproving iclassification iaccuracy iand idetecting iother icell i[25]. iThis idescribes ithe imulti ifeature iand imulti iclassifier

itypes i[21]. iMany ichallenges iwere iresolved iin ithe iproject. isystem ibased ion ipre-processing iultra isound iimage

iAmbiguous icell iboundary iis ithe ibiggest ifactor ithat idecomposition.

icontributes ito iover icounting i1-nuclei icells. Bayesian iclassifier iis ian ieffective imethod iused ito iimprove
An iautomatic icategorized iprocess ito istage iCLD iusing ithe imedical iimage isegmentation. iIn iorder ito iexclude ithe

iultrasound iimages, ilaboratory itests, iand iclinical irecords iare ibackground inoises iof iimages, ineed ipre-processing iof

idescribed. iThe ifirst istage iis icalled iClinical iBased iClassifier iimages. iWhen ithe ipre-processing imethod, iBayesian

i(CBC), idistinguishes ihealthy ifrom ipathologic iconditions. iclassifier iis iused ifor iclassifying iof iparticles iin ithe iimage.

iWhen inon ihealthy iconditions iare inoticed, ithe imethod irefines iBayesian iclassifier iis ia ipowerful iprobabilistic igraphical

ithe iresults iin ithree iexclusive ipathologies iin ia ihierarchical imodel ithat ihas ibeen irealistic iin icomputer ivision. iSome iof

ibasis. iThe istructures iused ias iwell ias ithe iclassifiers i(Bayes, ithe ipresent isegmentation ialgorithms ithe iuse iof iBayesian

iParzen, isupport ivector imachine, iand ik-nearest ineighbour) iclassifier iis ifollowed iand icentered ithe ieffect iof iBayesian

iare ioptimally iselected ifor ieach istage. i[22] iin ithis, ithe iclinical iclassifier iin isegmentation ialgorithms i[26]. iIn imedical iimage

ibased iclassifier iis iused ito imaintenance ithe inon-invasive isegmentation, iBayesian iapproach iis iconsidered ithe ibetter

imethod. iapplication.

In icooperation iTransient iElastography i(TE) iand ileft ilobe The ifeature isets ifor irobust ivisual iobject irecognition iare
iliver isurface i(LLS) iultrasound imay inon-invasively ilocate istudied iby iimplementing ilinear iSVM ibased ihuman idetection

icirrhosis. iThe idiagnostic ivalue iof ithese imethods iis iexamined ias ia itest icase. iWhen ilook iover icontinuing iedge iand igradient

iin ipatients iwith ia isuspicion ibut inot ia idefinite idiagnosis iof ibased idescriptors, iit idemos iexperimentally ithat igrids iof
icirrhosis. iThe iperformance iof iLLS iin ithe idiagnosis iof iHOG idescriptors iparticularly ioutperform iprevailing ifeature
icirrhosis iis iinvestigated iin ipatients iwith iclinical isuspicion iof
isets ifor ihuman idetection. iThe istatement iof iall istage iof ithe
icirrhosis ibut ialso ias ia idifferential idiagnosis ifor ithe ipatients
idesign iis imeasured ion iperformance, ifinishing ithat ine-scale
ipresenting iabnormalities, iwho iwere ichosen iover ithose iwith
igradients, ine iorientation ibinning, icomparatively icoarse
ian iestablished idiagnosis iof icirrhosis iin iorder ito imirror ithe
ispatial ibinning, iand isuperior ilocal icontrast inormalization iin
iclinical isituation iin iwhich ifurther iinvasive iinvestigation
icasing idescriptor iblocks iare iall iimportant ifor iworthy
iwould ibe ipursued. iAlso iit iassess iwhether ia idigital ianalysis
ioutcomes. iIt ioffers iclose ito iperfect iseparation ion ithe ioriginal
iof iimages iof iLLS ipermits ia ivalid, iobjective iand iquantitative
iMIT i(Massachusetts iInstitute iof iTechnology) ipedestrian
iLLS imeasurement. iThe iaccuracy iof iTE iis itested ifor ithe
idatabase; iso ithat ia imore ichallenging idataset iis iintroduced
idiagnosis iof icirrhosis iin ipatients iwith iclinical isuspicion iof
icovering iover i1800 inoted ihuman iimages iwith ia ilarge
ithe icondition. iThe ihypothesis ithat ithe imixture iof iTE iand
icollection iof ipose ivariations iand ibackgrounds [27].
iLLS iis itested iin iorder ithat iit iis iextra icorrect ithan ione

iunmarried iapproach ifor ithe inon-invasive idiagnosis iof

icirrhosis i[23]. iAlthough iit iis ian ieasy iand icost ieffective iway IV. CLASSIFICATION
iof idiagnosis, iit iis inot isuitable ifor iobese ipatient iand iascites
The idefinitive istep iin itumor idetection iis iclassification iof
ipatient.
itumor ias imalignant ior ibenign itumor. iClassification iin
The iadvent iof inear iinfrared iimagery iand iit’s iapplications igeneral iinvolves itwo isteps iof itraining iand itesting.
iin iface irecognition ihas iinstigated iresearch iin icross ispectral
Table 3: Classification types [28]
i(visible ito inear iinfrared) imatching. iRemaining iresearch ihas

ifocused ion iextracting itextural ifeatures iincluding ivariants iof S.No. Classificatio Algorithm used
ihistogram iof ioriented igradients. iThe ieffectiveness iof ithese n Based on
ifeatures iis istudied ifor icross ispectral iface irecognition. iOn 1 Training Parallel ipipelined ialgorithm, imaximum
sample based ilikelihood ialgorithm, iminimum idistance ito
iNear iInfrared i– iVisible iLight i(NIR-VIS) icross ispectral iface
imean ialgorithm i[29]
idatabase, ithree iHistograms iof iOriented iGradient i(HOG)
K imeans iclustering ialgorithm
ivariants iare ianalyzed ialong iwith idimensionality ireduction 2 Parameter Maximum ilikelihood ialgorithm, ilinear
iapproaches iand ilinear idiscriminant ianalysis. iThe iresults based idiscriminant ianalysis

idemonstrate ithat iDense iScale iInvariant iFeature iTransform Artificial iNeural iNetwork i(ANN),
iSupport ivector imachine i(SVM), iDecision
i(DSIFT) iwith isubspace iLinear iDiscriminant iAnalysis i(LDA)
itree iclassifier i[30].
ioutperforms ia icommercial imatcher iand iother iHOG ivariants
3 Pixel Maximum ilikelihood ialgorithm,
iby iat ileast i15%. iIt iis iobserve ithat ihistogram iof ioriented based iArtificial iNeural iNetwork(ANN), iSupport

igradient ifeatures iare iable ito iencode isimilar ifacial ifeatures ivector imachine i(SVM)

iacross ispectrums i[24].

Spectral imixture ianalysis, isub ipixel
iclassifier, ifuzzy iset iclassifier
The iresearch ipaper idiscuss iabout ithe iclassification iof GIS based classification approaches
iChronic iLiver idisease, isemi-automatic iprocedure, ito istage eCognition
ithat idisease ibased ion iultrasound iliver iimages, iclinical iand

iLaboratory idata, iIn ithe icore iof ithe ialgorithm itwo iclassifiers

iare iused i-Support iVector iMachine i(SVM) iK-Nearest

iNeighbour i(KNN).The iclassifiers iare itrained iwith

imultimodal ifeature itest. iThe iresults ishows iimprove ithe

iclassification iof iaccuracy. iThe imost icommon ifeatures

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4169 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

4 Spatial maximum ilikelihood ialgorithm,

element iminimum idistance, iArtificial iNeural
iNetwork i(ANN), iSupport ivector imachine
i(SVM), iDecision itree ialgorithm

Fuzzy algorithm
5 Spatial maximum likelihood algorithm,
information minimum distance, Artificial Neural
Network (ANN)

Frequency-based contextual classifier Liver idisease iis iliberal, iasymptotic iand ihypothetically
Combination of parametric or ifatal iDiseases. iAn iautomatic igraded iprocess ican ibe iused ito
non-parametric and contextual algorithm iclassify iand istage iliver idisease iby iultrasound iimages. iThe
[31] idatabase iused iis ithe iultra-sono-graphic iimages iof iliver

idisease ibeside iwith ithe ihealthy iconditions. iInitially ithe

Table 4: Classification Techniques
icontrast ienhancement iis irealistic ito ithe iinput iimage ithat
Classificat Description
ion method ihelps ito iidentify ithe iobject, iafter ithat idiscrete iwavelet

itransform iis iapplied iwhich isupports ito ieliminate ithe ispeckle

Artificial ANN iis ia isort iof isynthetic iintelligence ithat
inoise, ithen ithe irough icomponent iis iexposed ito iK-mean
Neural iimitates ia ifew ifunctions iof ithe iman ior iwoman
network imind. iANN ihas ia iordinary itendency ifor istoring iclustering iwhich isegments ithe iimage iwith irespect ito ithe

iexperiential iexpertise. iAn iANN icontains iof ia iminimum iEuclidian idistance. iThe iclassification iplan iis
iorder iof ilayers, ievery ilayer iconsists iof ia iset iof
iperformed iusing ithe iclassifier isuch ias iNeural iNetwork. iIt iis
ineurones. iAll ineurones iof ievery ilayer iare irelated
iwith ithe iaid iof iweighted iconnections ito iall iused ito iexamine ithe iLiver idisease iwhich iwill ibe iworthwhile

ineurones iat ithe ipreceding iand isubsequent ito idoctors ifor ithe isecond iopinion. iThe iinput iimage iendures
ilayers[32].
ipre-processing ithat icontains iContrast iEnhancement, iDiscrete

iwavelet itransform iand iSegmentation iusing iKNN iclassifier

Decision It icalculates iclass imembership iby iconstantly i[35]. iThis imethod imay ireduce ibut inot ireplace ithe iliver
tree ipartitioning ia idataset iinto ieven isubsets ibiopsy. iIt ihelps ifor isecond iopinion.
iHierarchical iclassifier icertificates ithe iacceptations
iand irejection iof iclass ilabels iat ieach iin-between
istage. i V. TEMPLATE MATCHING
Support A iassist ivector imachine ibuilds ia ihyper iplane ior Template imatching iis ia itechnique iused ito iverdict iareas iin
Vector iset iof ihyper iplanes iin ia ihigh- ior ilimitless
iphoto ithat iare isimilar ito itemplate. iIn itumor idetection isystem,
Machine idimensional ispace, iused ifor iclass. iGood
iseparation iis iperformed iby imeans iof ithe ihyper ia itemplate iis iused ias ian iaid ifor icoarsely ifinding itumor iby

iaircraft ithat ihas ithe ilargest idistance ito ithe iclosest iway iof iruling iout ithe iedges. iA itemplate iis icreated iand iis
ieducation istatistics ifactor iof iany iclass i(useful
imoved iover ithe ireceived iphotograph isequentially iand ithe
imargin), igenerally ilarger ithe imargin ilower ithe
iarea iin iwhich ithe itemplates imatch ithe iimage iis imarked. iThe
igeneralization imistakes iof ithe iclassifier.
itemplate iis ialways iof ismaller ilength ithan iimage. iThe imethod

iis iof ibrilliant iimportance iand iis iuseful iin isuccessfully

idetecting itumor. iA iCT iliver iimage idiagnostic imachine ito

iunearth ithe iliver iboundary iand itumors iin iliver. iTo iextract
Bayesian iclassifier iis ian ieffective imethod iused ito iimprove
ithe iliver iboundary isegmentation iapproach i“Detect iBefore
ithe imedical iimage isegmentation. iIn iorder ito iexclude ithe
iExtract” iwas iused. iThe idesired iliver iboundary iwas iachieved
ibackground inoises iof iimages, ineed ipre-processing iof
iby iusing iDeformable iActive iModel i(contour imodification
iimages. iWhen ithe ipre-processing imethod, iBayesian
imodel). iClassification isystem ideveloped ion ibasis iof iMPNN
iclassifier iis iused ifor iclassifying iof idebris iin ithe iimage.
i(Modified iProbabilistic iNeural iNetwork) iwas iapplied ifor
iBayesian iclassifier iis ia ipowerful iprobabilistic igraphical
idiscrimination ibetween iheptoma iand ihemageoma iliver
iversion ithat ihas ibeen ipractical iin ilaptop iimaginative iand

iprescient. iSome iof ithe iexisting isegmentation ialgorithms ithe itumor. iThe iclassification isystem iused iin ithis ipaper iis

iusage iof iBayesian iclassifier iis iadopted iand icentered ithe iTexture i–Feature ibased iwhich iused ifeatures.

ieffect iof iBayesian iclassifier iin isegmentation ialgorithms i[33].

iIn imedical iphotograph isegmentation, iBayesian iapproach iis VI. IMAGE SEGMENTATION METHODS
itaken iinto iconsideration ithe ihigher isoftware.
In ithis isection, inumerous isegmentation istrategies iin ithe
The iclassification ialgorithms ilike iNaive iBayes iclassifier, iscientific ipix iprocessing iare iin icomparison. iEGBIS
iC4.5, iSupport iVector iMachines, iand iBack ipropagation
itechnique ihas iefficient igraph ibased itotally iimage
iNeural iNetwork ialgorithm iare icomputed ibased ion ifour
isegmentation. iEDISON itechnique iis ibased ion iimply ishift
icriteria inamely iAccuracy, iPrecision, iSpecificity iand
iclustering i[36]. iImage isegmentation iapproaches iare iin
iSensitivity. iSupport iVector iMachine i(SVM) isplits ithe idata
igeneral ibased itotally ion ione iof ithe i iessential ihomes iof
iinto itwo itypes iof iworking iout iclassification iand imaking ian
iintensity ivalues iwhich ican ibe idiscontinuity iand iresemblance
iN-dimensional ihyper iplane. iThese ifashions iare ifurther
i[37].
irelated ito ineural inetworks. iThe iselected idataset, iKNN, iBack

ipropagation iand iSVM iare igiving ibetter iresults iwith iall iof ithe

ifeature iset icombinations [34].

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4170 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

iSegmentation istrategies ibased ion iincoherence iimage

iattempting ito ifind ian iclean iexchange iinside ithe ilevel iof
idepth. iIf ione ilocated, idivide ithe iphoto iat ithis ialteration.

iHowever, isegmentation itechniques ibased itotally ion

iresemblance iscan ithe iphoto iand ipartition ithe ipicture iinto

iseveral iregions, iwhich ican ibe isimilar iin iline iwith ia ifixed iof

ipredefined isituations. iImage isegmentation istrategies iare

iusually idesigned iand icarried iout ito isome iorgans iof ithe

iframe. iThese istrategies imay ibe iused ifor iseveral iphoto

imodalities ialong iwith iCT, iMRI, iand iUS.

In ivicinity igrowing iapproach itechnique, ithe imethod istarts

ioffevolved ithrough ichoosing ia ipixel ias ia iseed ilocation. iThe
ivicinity igrows iby imeans iof iadding ithe iadjoining ipixels

iwhich iare isimilar icriteria iwith ithe iseed ithe iusage iof

ipredefined istandards iwhich iinclude idepth. iThe iprocess iof

iregion igrowing istops iwhen ithere iis ino ipixels ito ibe iadded. i

In iregion isplitting iand imerging imethod, ithe iprocess istarts

iusing ithe ientire iimage ias ia iseed. iThen iit isplits iinto itypically

ifour isub iregion. iThe iprocess iis irepeated iusing ithe ispited isub

iregion ias inew iseed. iThe iprocess iends iwhen ithere iare inow

iregions ito ibe ispitted. iAfter ithat, ithe ineighbouring iregions

iwith ithe isimilar iproperties ilike iintensity ior igray iscale iare

Fig 3: Segmentation Methods imerged itogether.

A. Thresholding C. Edge-based Segmentation

One iof ithe iessential istrategies iused ifor imedical iimage Edge ibased isegmentation iis ithe imost icommon imethods ifor
isegmentation iis ithresholding i[38]. iIt idivides ithe iphoto iinto i iimage isegmentation. iIt iis iused ifor isensing ithe iwithdraws iin

iareas. iOne iis ireferred ito ias ithe iitem ior iforeground iand ithe ivalue iof iintensity iin iedge ibased imethod i[39, i41]. iIt iworks ion

ilast ipart iof ithe iphoto iis ireferred ito ias ithe iheritage. iTo ithe ibase iof iidentifying iregion iborder iof iimage iquick ichange

itransform ithe iimage iinto ibinary iimage, ithe ivalue iof ithe iedge iin ithe iintensity ilevel i[42]. iA iborder ibetween itwo iregions ithat

i(T) ineeds ito ibe idecided ion ias iit ishould ibe. iThe iobtained irange iin ithe istage iof idepth iis isaid ias iedges iin ian ipicture.

ibinary iphoto icarries ithe iwanted iinformation iapproximately iThese iedges iare iused ito idefine ithe isize iof ithe iitem iand ito

ithe iplace iand ithe iform iof ithe iitem iwhich ican ibe iworried. idistinct ifrom ithe ihistory. iThe iedge idetectors ishould ibe iused

The imotive iof ichanging igrey iphoto ito ibinary ipicture iis ito ito ilocate ipoints iof iimage iwhere iintensity ivariations

iget ieasier iinformation iand ithis iends iin ihandy iprocedure iof inaturally. iSearch ibased imethod iand izero icrossing ibased

iclass ilevel. iWhen ithe iintensity iof ithe ipixel iis imore ithan ithe imethod iare itwo ivital imethods iof iedge idetection.

ibrink iprice, iit'll ibe ilabeled ias iwhite ipixels iand iits icost iset ias

i(1) iand iit's imiles istated ibecause ithe ipart iof ithe iobject. iIf ithe
D. Clustering Techniques
ipixel iprice iis iunderneath ithreshold iprice, ithen iit iwill ilikely A icluster iis ia icollection iof isimilar ipixels iwhich imight ibe
ibe iindicated ias iblack ipixels iand iset ias i(0) iand iit's ifar imultiple ito ithe ipixels iinside ithe igreater iclusters i[43].

imeasured ias ipart iof ibackground. iClustering imethods imake iclustering iboth ithrough idividing

In icommon, ithere iare itwo ithresholding iused. iSpecifically ior iby imeans iof igrouping ipixels. iIn ipartitioning ikind,

iglobal ithresholding iand inearby ithresholding. iThe ithreshold iclustering ibegins iwith ithe ioverall ipicture iand isegment iinto

ifee iT iis ikept ipersistent iin iinternational ithresholding. iA ismaller icluster. iIn igrouping itype ithe iclustering istarts iwith

idrawback iin iworldwide ithresholding iis ithe isegmentation ieach ielement ias ia iunique icluster iand icombines ithese iclusters

imethod imight ialso ifail iif ithe ilighting iof ithe iphotograph iis ito iget ilarger iclusters. iThe itwo ikinds iof iclustering iare

inot iuniform. iThe ithreshold icost iT iisn't ipersistent iin ilocal isupervised iclustering iand iunsupervised iclustering.

ithresholding, iso imore ithan ione ithreshold ivalues ican ibe iused iSupervised iclustering imethod iprerequisites ithe iinteraction iof

ito iequalize ithe iillumination idifficulty. ihumans ito iregulate ithe iclustering ivalues, iIn iunsupervised

iclustering imethod, ithe iclustering icriteria iis iresolute iby iitself.

B. Region-based Segmentation iClustering iconsists iof iFuzzy iC-means ialgorithm i(FCM) iand

Region iprimarily ibased isegmentation iworks ion ithe ibase iof iFuzzy iC-means ialgorithm i(FCM).

isimilarity ia ifew iof ithe ipixels iin ia iarea. iThe itechniques iare

iused ito iselect ithe iarea iimmediately i[39, i40]. iThese

E. FCM Algorithm
itechniques imixed ithe ipixels irelated itraits i(which iinclude Step 1: It regulates the centres of the clusters and fills the
iintensity) iinto iareas. iThere iare itwo imethodologies iinside ithe association matrix with random values among zero and 1.
iarea ibased istrategies. iOne iis iarea ideveloping itechnique iand Step 2: It defines the centres of the fuzzy clusters
ithe iopposite ione iis iarea isplitting i/ imerging iapproach. Step 3: It calculates the fee
fraction

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4171 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

Step 4: It training session a new attachment matrix and repeat VIII. RELATED WORK
the stairs until an top-quality end result is acquired
Lesion isite iselection iby imorphological iimprovement iand
F. PSO Algorithm icontextual isegmentation i[49] istates ithat istatistical imodel

ireinforced iapproach ifor ienhanced isegmentation iand

Step 1: It set a populace of debris with velocities and random
iextraction iof idoubtful imass iarea iin imammographic iimages.
role within the seek area
iWith ia iappropriate istatistical ireport iof imany idiscriminate
Step 2: The health requirements for each particle is
ifeatures iof itrue iand ifalse isamples ifrom ilocalized iareas, ian
calculated.
ibetter imass idetection imay ibe iattained iin icomputer iassisted
Step 3: It creates the vital updates for the greatest role for
idiagnosis i(CAD). iThe imethod iof iautomatic ibrain itumor
every particle and the worldwide first-rate function for the
idetection ioriginates ithe imethodology ithat isegments ithe ibrain
swam.
itumours iautomatically. iThis imethodology iis ibased ion
Table 5: Segmentation Techniques.
S.No Technique Process icovariance iand igeodesic idistance i[50].

. Gunasundari iet ial i[1] iproposed ia isystem ifor iautomatically

1 Thresholding The ihistogram ibased istrategies iis iRely iat ithe idetecting iof iHepato iCellular iCarcinoma i(HCC) iand
ifulfillment iof ithe iapproximating ithe
ihemangioma iliver itumours ifrom iCT iimages. iThe ischeme
ithreshold iprice ithat isplits ithe i2 ihomogenous
isegmented ithe iliver iusing ihistogram ianalyser iand
iarea iof ithe iitem iand ihistorical ipast iof ian
iphotograph. iHistogram ibased ithresholding iis imorphological ioperations. iTextural ifeatures ifor iimage

ifunctional ito iget iall iviable isame iareas iwithin iclassification i[51] ipresented ione iof ithe iimportant iimage
ithe ipicture i[44].
ianalysis iapplications iof iGray iLevel iCo-occurrence iMatrix
2 Edge Based It iis ia ipopular iimage isegmentation
i(GLCM) iand itexture ifeature. iThis imethod itracks itwo isteps
ialgorithm ifor igray iscale iimages. iTo iindicate
ia ilabelling iof ithe iimage isuch ithat iall ipoints ifor ifeature iextraction ifrom ithe imedical iimages. iIn ithe ifirst

iof ia igiven icatchment ibasin ihave ithe isame istep, ithe iGLCM iis iconsidered, iand iin ithe iother istep, ithe
iunique ilabel iand ia ispecial ilabel idistinct ifrom
itexture ifeatures ibased ion iGLCM iare iconsidered. iDue ito ithe
iall ithe ilabels iof ithe icatchment ibasin iis
icomplex istructure iof ispread itissues isuch ias iWM, iGM, iand
iassigned ito iall ipoint iof ithe iwater ished i[45]. i
i iCSF iin ithe iMR iimage iof ibrain, iextraction iof isimilar ifeatures
3 It iis iestablished ion ifeatures iand
iis ia ineeded itask. iTextural ifindings iand ianalysis imay iimprove
Region based ihypotheses imodel iis isuggested. iThe istudy iof
ithe idiagnosis, idifferent istages iof ithe itumor i(tumor istaging),
ithe ifeature iprovides ithe iessential iparameters
ithat iserve ias iinduce iin ithe ihypothesis imodel iand itherapy iresponse iassessment. iChaddad ihas isuggested ia

i[46]. itechnique iof iautomatic ifeature iextraction ifor ibrain itumour

4 Each ipixel ionly igoes ito ia ispecific icluster. idetection ibased ion ithe iGaussian imixture imodel i(GMM)
Cluster based iThe iclustering iis idone icreated ion ieither ipixel
iusing iMR iimages iof ibrain i[52]. iIn ithis, iusing ithe iprincipal
icolour iintensity iof ithe iimage ior itexture,
ilocation, ior ia igrouping iof ithese ifactors. iThe icomponent ianalysis i(PCA) iand ithe iwavelet ibased ifeatures;

iK ivalue ican ibe idesignated imanually, ithe iperformance iof ithe iGMM ifeature iextraction iis iimproved.
irandomly, ior iby ia iheuristic iapproach. iFuzzy iAn iaccuracy iof i97.05% ifor ithe iT1-weighted iand
iC-Means imethod, iIt iis iused ito idetermine
iT2-weighted iand i94.11% ifor iFLAIR iweighted iMR iimages iis
ievery icluster ilocation; iusing ithis itechnique,
igained. iSachdeva iet ial. i ihave ipublished ia imulticlass ibrain
imaximum imembership idefuzzification iand
ineighbourhood ismoothing iis iattained i[47]. itumor iclassification, isegmentation, iand ifeature iextraction

5 In ithis, isimultaneous ihuman idetection iand iimplemented iusing ia idataset iof i428MR iimages[53]. i
Matching isegmentation imerging ilocal ipart ibased iand
Segmentation ifunction iextraction iand imulticlass ibrain
based iglobal ishaped itemplate ibased isystems. iThe
itumor iclassification i[54] iproposed ia isemi-computerized
imethod irelies ion ithe ikey iidea iof imatching ia
ipart i– itemplate itree ito iimages ihierarchically iclassification itechnique ito itake ia ilook iat isteatosis iliver

ito idetect ihumans iand iestimate itheir ipossess itissues icooperate iof iB-test iUS ipix. iSeveral ifunctions ihad
i[48].
ibeen iextracted. iIt icategorized ithe idiseases ihold i3

iexceptional iclassifiers, iANN, iK-Nearest iNeighbours

VII. TRANSFORMATION TECHNIQUES i(KNN), iand iSVM. iThe idevice imay ibe iutilized ito idistinguish

ibetween isteatosis iliver iand ia iwholesome iliver. iThe ithree

Hough itransform iis imainly iused ifor iedge idetection. iEdge
iclassifiers ihave ibeen ieducated ithe iusage iof ithe iten
detection iis ithe ivital imethod ifor iidentifying imeaningful
ipass-validation itechnique. iThe ireceived iresults iuncovered
idiscontinuities iin iintensity ivalues. iClarification iof iimage
ithat ithe ioverall iperformance iof ithe iSVM itype iis imore
icontents iis ione iof ithe ideterminations iin icomputer ivision
isuitable ithan ithe iANN iand ithe iKNN. iThe ioverall
iaccurately iin iimage iprocessing. iHough itransform ibased iline
iperformance iof ithe icautioned imethod ihas ishown ithe iwhole
idetection iusing idissimilar iedge idetection itechniques
iexactness iof i79.77%, i76.92 iand i74.05, ione iat ia itime. iMala
ilike-Robert, iSobel, iPrewitt, iKirch, iLaplacian iof iGaussian,
iet ial. istretched itheir iwork iand iplanned ia inew ioutline. iIt
iLaplacian iand iCanny iare iused. iEven ithough iso imany iedge
icategorized ibenign iand imalignant itumour iusing iComputer
idetection itechniques iare iobtainable iin ithe iliterature, ithe
iTomography idata iby iamalgamating ibiorthogonal iwavelet
ichallenging itask iis ito ithe iresearch igroups ito inotice ithe iexact
itransform iof iLinear iVector iQuantization i(LVQ) inetwork
iimage iwithout inoise ifrom ithe ioriginal iimage. iSeveral iimage
i[55]. iLater ipreliminary ipre-processing, iimage iwas ifinished
ican ihave inoise iin ithem ilike-Heat iwaves, iDust, iSmoke,
iand imultiplied iwith ioriginal iimage ito isegment iliver. iThen
iAcquisition inoise ietc. icanny iedge idetector istretches ibetter
iFCM iclustering iwas iefficient ito isegment ithe iimage iinto
iresult iin inoisy iimage ithan iother ioperator i[20]. iCanny iedge
iliver, ibackground iand
idetector iproposes ithe ibetter iresult iin inoisy iimage ithan iother
itumour. i
ioperator.

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4172 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

Liver ifibrosis iidentification itechnique ifor iultrasound iliver ithresholding, iotsu ithresholding iand iregion ibased
iimages iclassification. iFor ifeature iextraction, iused ithe iFDTA isegmentation. iJianhua iLiu iet ial. iAn ieffective isegmentation
iand ithe iSGLDM ion i64x64 ipixels isub iimage i[56]. iThe ijoint imethod iis iused ifor iedge ideduction. iWater ished iAlgorithm

ifeature ivector, ithus igotten, iwas iused ito idifferentiate i273 i[69]; iRegion iMerging ietc. iare imainly iused. iAnd ithis imethod

ihealthy iand i99 ifibrosis iliver iimages. iTwo itype iof ican iimprove iresolve ithe isegmentation iof ithe ialgorithm.

iclassification imethods inamely iFisher ilinear iclassifier iand iA.S.Thenmozhi iet ial. iIn ithis ipaper, ifuzzy iconnected iand

iSupport iVector iMachine i(leave-one-out ialgorithm) iwere iwatershed itechnique iis iused, ithis itechnique iis iused ifor ibetter

iused. i Nanzhou iPiao iet ial. isuggest ia imethod ifor ivolume iconsistency iof isegmentation iquality. iBy ithis ipaper ithe

icalculation iin i3D ispace iusing ilayers iof i2D iCT iimages. iThe iconcept iof ifuzzy iassociated iis iwithin ithe icontext iof

isystem ican iprovide ipractical isupport ifor imedical isupports isemi-automatic ipicture isegmentation i[70]. i iNuseiba. iM iet ial.

iand isurgery ioptions i[57]. iGrzegorz iChlebus iet ial. ipropose ian iAt ithis ipaper iauthor iproposes ithe iorgan isegmentation iusing

iautomatic imethod ifor iliver itumor isegmentation iwith i2D iNovel iBalloon iForce. iThis iForce iis iused ito ithe isegmentation

iimage ifrom iCT. iThis imethod iattains isegmentation iquality iin iof ithe iLiver. iDLRS imodel iis iused ito iprevent ithe ichallenges

idetection iof ilesions iwhen iassociated ito ihuman iexpert iis i77% iof ithe isegmentation, iimprove ithe ieffectiveness idealing iwith

i[58]. i iover isegmentation iproblem i[71].

Deep ilearning iderived ifrom iautomatic iliver itumor In iAutomatic isegmentation imethods iin imedical iimages
1
isegmentation iin iCT iwith ishape ibased iupon ithe ipost ipaper, ithe iauthor ifocuses ion ithe i segmentation imethod iin

iprocessing isubmit ithe icapable iimage iprocessing iin iliver iMedical iImaging iand iautomatic iImage iSegmentation.

icancer idetection ifocused ithe ineed ifor iinvasive imethod. iThis iComparative ianalysis iimplemented iout iand ialso irecognized

ican ibe irecognized iby ipre-processing, ienhancing, iimage ithe iadvantages iand idisadvantages iof ievery imethod iin

isegmentation iand ivolume icalculation i[59]. iEva iVan iMedical iImaging i[72]. iIt iidentifies ithe iwrong idiagnosis.

iRikxoort iet ial. iproposed, iessential iof ithe imethod iconsist iof iHussein iAlahmeretal. iThis iPaper isuggests ithe iClassification

ivoxel ilabelling iprocedure, iliver iis iassessed iusing istatistical iof iLiver iLesions iusing iCAD iclassification ibased ion imultiple

iclassifier i(KNN). iThis imethod ievaluated ion i10 itest iscans iROI. iAutomated iapproach ifor ia iliver ilesion iclassification

iand iachieved irobustly ias ithe ivolumetric ioverlap ierror iis isystem iin ito ibenign iand imalignant i[73]. iEnhanced ithe

i12.5% ion iaverage iand i15.3% ifor ithe iworst icase. iIn ithis iclassification iaccuracy ito iover i98%. iPatrick iFerdinand

ipaper iused ithe ipre-processing, iresampling, irange iChrist iet.al. iIn ithis ipaper, iproposes iLiver iand iLesion

idetermination, ipost iprocessing, ivoxel iclassification ietc i[60]. isegmentation iby ifully iConvolutional iNeural iNetworks, i3D

iOussemaZayanel iet ial. imade ia istudy ion ithe iproblems iof iconditional iRandom ifields iand ivolumetric isegmentation

iautomatic isegmentation iof ithe iliver iand iproposed ia imethod i[74]. iExamine ithe ihigh iaccuracy iand ispeed ifor iusage. iSajith

iusing iGCC ialgorithm i[61]. iB iBalakumar iet ial. isuggested ia iA. iG., iHariharan i.S, isuggests ithe iLiver iTumor iSegmentation

ifully iautomatic isegmentation imethod ifor iliver iCT iscan. iIn iusing iSpatial ifuzzy iC-Means iClustering. iSegmentation ion

ithe ifirst, idifference iof iunique iimage ienhanced ifor iclear ithe ibasis iautomatic iand isemi-automatic. iIncrease ithe

iboundaries i[62]. iSecond, iSpatial ifuzzy ic imeans iclustering irobustness iof ian iimage iwith ifewer inumbers iof irepetitions

iutilized iextract iliver iarea. iThird, idistance iregularized idegree iand ialso iexpand ithe isegmentation iof ithe iliver iimage ithereby

iset iis iused. iAt iremaining imorphological ioperations iare iused iincreasing idetection iof itumour ieffectively i[75]. iRecognize

ifor iput iup iprocessing. iLeast idistance icounting ion iliver ithe iimproper isegmentation.

isegmentation imethod ifor iCT itest ipicture iliver isegmentation A ivoyage ion imedical isegmentation iprocess istates ithe
i[63]. iAbhijeet iet ial.CAD iSystem iis iused ifor ianalysis iof ithe iimage isegmentation, iusing isegmentation ialgorithms. iIn ithe

iLiver. iHaarWavelet iTransformed iand iClustering itechnique igenesis ithreshold, iseed ipoint iand iedge itracing imethods. iIn

iis iused ito idecrease ithe itime icomplex, icomputational ithe iII igenesis iused ioptimization imodel. iIn iIII igenesis igive
icomplexity ifor ithe iendurance ifor idetection i[64]. iAhmed iM. ithe iprior iinformation iin isegmentation iprocess i[76].
iAnteret ial. iIn ithis ipaper iHybrid isegmentation imethod iused,
iBelgherbi iET ial.proposes iautomated imethod ifor iLiver
i3 isteps iare iused ifor imanner ithe isegmentation i[65]. iIn ithe
ilesions. iWatershed itransform, iFiltering, icontrast
iprimary isection, iby iway iof iadaptive ithreshold ithat
ienhancement. iThe iresult iof isegmentation iwas i92% iof
iinvestigate idepth ivalues iof ithe ineighbourhood iof ievery
idetection irate iand i99% iof iparticularity i[77]. iHuiling iGuo iet
ipixel, iIn ithe isecond isection, imultiscale imorphological
ial. iThis ipaper iprojected ia ikind iof isegmentation imethod ifor
ioperators ito iclear iout itissue. iIn ithe i0.33 isection, icast ioff ithe
iCBCT iimage ibased ion iBayesian inetwork i[78]. iIt iensures
ismall iitems iand ifake itremendous iareas. iThis iapproach
ithat ithe itarget ipoints iof ithe iLiver iorgan iCBCT iimage iare
iupturns iboosting iand igrowing ienhancement iof iweek
iclearly ivisible; iBayesian inetwork imethod ihas igood iability ito
isegmentation. iAntonidoss iet ial. iIn ithis ipaper isuggested ia
idistinguish ipixels iwith irelated itissue iextraction ifrom iliver
itechnique iof iautomated iliver isegmentation ithe iusage iof
iorgan. iHariharan. iSit irecommends ithe isegmentation iusing
iAdaptive ithreshold imethod i[66]. iThis imethod iextracted iliver
itexture ibased iimage ianalysis. iGradient ioperator ilike iSobel,
iarea iappropriately iand ibetter ithe iliver ivicinity. iIna iSingh iet

ial.The icreator iapplied iACO iK-Means iclustering iapproach iis iCany, iPrewitt, iCompas, iRoberts, iGaussian i(etc) iused ifor

iused ifor isegmentation iof iliver iin iaddition ito iphase iof ithe iimage isegmentation. iUsed iRegion igrowing, iSplitting, iand

iLiver ilocation. iMATLAB iis ia itool ioffer iinteractive iMerging ithresholding i(etc) ifor iClassical iImage isegmentation

ienvironment ifor ilayout iand idifficult isolving i[67]. i[79].

This ipaper iperformance ievaluation iof iliver iUS iimages

iusing iimage iprocessing itechnique imention ienhancement

itechnique iis iused ilike iShock iand ispatial ifilter. iThe iprocessed

iimage iis iretrieved ion ithe ibasis iof iPeak isignal ito iNoise iRatio

i[68]. iIt iis iused iapplied isegmentation imethod iare inormal

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4173 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

I This istudy, the iautomatic isegmentation iof iLiver iusing iabove irepetitions iof isegmentation iprocess, iand ifor ieach
iCAD, iRegion igrowing ipromoted ipre iand ipost iprocessing iindividual iobject i[92]. iIna iSingh ideclares iSegmentation iof
ifunction ifor iautomatic isegmentation iof iLiver iand iAlternative iliver ifor itransplantation iplanning iand idetection iof idisease

iFuzzy iC-Means ialgorithm ifor itumor idetection imethod iwas iremains ithe imost ichallenging itask iin imedical iimage

iused i[80]. iAmitha iRaj iet ial. idetect ithe iLiver itumor idetection iprocessing iowing ito ihigh irate iof iinter-patient ivariability iin

iusing iMarkov irandom iFields. iFeature iextraction icomplete ito iliver ishape, isize, ivolume iand ipathologies ior idifferent

icategorize ithem iusing iSVM iclassifier[81]. iLiver iTumor idiseases i[93].

idetection iand isegmentation ibrings ia irobust itool ifor iliver Segmentation iof iLiver iCancer iUsing iSVM iTechniques iby
itumor ianalysis. iQiuyang iet ial. iIn ithis ireport, ibased iupon i ithe iM. iRajeshwaran iet ial. iindicates ithat inaturally ihave imore

imorphology idevelopment iand icorrosion itechnique ibased ion ispackle inoise iin iUltrasound iliver itumor iimage.

iwindow itechnology iand ide-noising iprocessing, ifollowing ito iConsequently iautomatic iidentification iof itumor iin

ithe inecessity iof iliver iimage iorder isegmentation[82]. iLauren iultrasound iliver iimage iis ia istimulating itask i[94]. iPoonam

iK. iHaaitsmabased on isegmentation iof ithe iLiver iusing iROI iDevi iet ial. icomplete istudy ifor iassociating ithe iperformance iof

iselection, iGradient iVector iFlow(GVF),K-Means iclustering, iBack-Propagation iNeural iNetwork iand iSupport iVector

iFuzzy iC-Means iClustering, iiterative iWatershed, iHMM iMachine i(SVM) ifor iclassification iin iliver icancer. iThe

iMeasurements itechnique[83]. iExpand ithe iperformance, iand ipresentation iof ithe itwo imodels iare ias icompared iwith iand

ithe ialgorithms iwere icompared, ifound ito ibe ithe ibest isuited iauthenticated iin iterms iof icorrectness iwithin ithe ipositive irate

ifor ismall itumors. iVishwajit iB.Mohite iet ial. iplanned ia i[95]. iUsman iAkram iet ial. iproposed ia imethod ifor iliver

ischeme ito iidentify ithe iLiver itumor iSegmentation. iIn ithis isegmentation ifrom iCT iimages. iIt iis iinvariant iin iterms iof

ipaper ithe isophisticated ihybrid isystem iare iproposed, iwhich iis ishape, isize iand iintensity ivalues. iThe isystem iinvolves iof

icapable ito isegment iliver itumor ifor ia iliver i[84]. iMaria ithree istages. iIn ithe ifirst istage, ithe iCT iliver iimage iis iacquired

iTsiplakiodu iet ial. ialso iconsidered iautomated idetection iof iand ithe ipre-processing iis idone ito iremove iunwanted inoise

iLiver, iIdentify ithe ifat iregions iusing isteatosis iquantification. iand ito ienhance ithe iimage iquality. iIn ithe isecond istage, iliver

iIn ithe isecond istage iearly icandidate iimage iregions iare iregion iis isegmented ifrom ithe iCT iimage. iIn ithe ithird istage,

iidentified. iGives ithe ihigh iresolution itime i[85]. iMredhulaL iet ithe ipost iprocessing ienhancement iis idone ion ithe isegmented

ial. iused ifuzzy iC iMeans, isubtractive iMethod, iMorphological iliver iregion ito iimprove ithe icontrast iof iliver iregion i[96]. i

iMethod, iclustering itechnique ito idetect iand iclassification iof Detection iand isegmentation iof iabnormal isegment iexisting
itumor. iIn ithis iwork idevelopment iin ithe iexactness iof iresults iin ithe iliver iis ione iof ithe itesting iand iimportant isteps iin

iare iobtained iby inoise iremoval isteps iusing ifuzzy iC-Means itreatment iplanning iwhich ihelps iin iincreasing ithe iexistence iof

iclustering i[86]. ia ipatient. iLiver icancer iincreases ithe ideath irate, isince ithe iside

Analysis iof iLiver itumor ifrom iCT iphotographs ithe iuse iof ieffects icannot ibe ieminent ieven ithe icancer iis iin iits iadvanced

ifirst iorder istatistical ifunctions i[87] ia iassociation iModified istage. iThe iinitial idiagnosis iand ireliable iobserving iis ithe

iK-Mean iIrregularity iand iFirst iorder istatically ifunctions ito ipreeminent iway ito icontrol ithe iimprovement iof ithe

ievaluation iLiver itumor ifrom iCT iImages. iIn ithis ipaintings, imalignancy iand ito irescue ithe ilives. iUltrasound iimaging iis ia

ithe iCT iphotos iare iengaged ithe isegmentation iprocedure iare idifferentiate iamong ithe imost inormally iutilized idiagnosis

iapplied ito ithe iliver iphoto iand ifind ithe iextract ithe iCT iLiver itools ito idifferentiate iand icategorize iindiscretions iof ithe iliver

iboundary iand igreater icategorize iliver iailment. iMohammed iwhich iis ialso ia inon-invasive, isafe itechnique ifor ipatient

iGoryawala iet.Al. iUsed iModified iK-Means iand iParallel iexamination, ibeing ieasy ito iapply, iinexpensive iwhen

icomputing ito idiscover ian icorrect i3-D iliver isegmentation. icompared ito ithe iCT, iMRI, iPET ibased iliver itumor idetection.

iStatistical ianalysis iis iused ito iget ithe icorrectness iof iLiver icancer idetection iplans ihave ilong icalculation itime iand

iconsequences. iIn ithe icurrent istudy iANN ibeen isurveyed iin icomposite iin inature. I

icategorizing ithe iliver itumor i[88]. iA. iMohamed iRiyasey ial. Therefore ito ireduce ithe icomplexity iin ithe icomputerized
icomparative istudy iof iCT iliver iimage isegmentation iusing iprocedure iand ito iimprove ithe idiagnostic iaccurateness. iIn ithe

iglobal ithresh ihold iiterative, iOtsu’s iMethod iand iregion ipaper ithe iauthor ipropose ia inew ioptimal iclassified ifeature

igrowing imethods iare iused. iThe ioutcome ican ihelp ithe ibetter ifusion ibased ion iPenguin iSearch iOptimization iAlgorithm

isegmentation imethod iselection ifor ithe icomputerized iCT i(PeSOA) i[97]. iShikha iMandhan iet ial. iImage isegmentation iis

iliver idiagnosis i[89]. iEvgin iet ial. iprovides ian ibrief iaccount iof ithe iprogression iof isubdividing ithe iimage ito iinto iits iparts ithat

iliver isegmentation imethods iin imagnetic iresonance iimages iare iessential iand iis imeasured ione iof ithe imost iproblematic

iand iillustrate ithe icomparative iresults iof iseven idifferent itasks iin iimage iprocessing. iIt ishows ia itask ithat iis ia inecessity

ipseudo-3D iliver isegmentation iapproaches iselected ifrom iof iany iapplication iand iits ispecific isuccess iis ibased ion ithe

ideterministic i(K-means-based), iprobabilistic i(Gaussian iactual iimplementation iof ithe isegmentation imethod. iFor

imodel-based), isupervised ineural inetwork, iand ideformable inumerous iapplications, isegmentation idecreases ito itracing ian

imodel-based isegmentation itechniques i[90]. iobject iin ian iimage. iIt iinvolves ipartitioning ithe iimage iinto

A iFully iConvolutional iNetwork i(FCN) ifor ithe iprocedure itwo iclasses, ibackground ior iobject. iThe iquantity iof iimage

iof iliver isegmentation i[91] iand iliver imetastases idetection iin iinformation iwhich ican ibe icaptured iby itechnology, ias iit iis inot

icomputerized itomography i(CT) iexaminations. iFCN ihas ifeasible ito imanually iprocess iall isuch iimages i[98].

iestablished ito ibe ia iactual idominant itool ifor iexpressive Segmentation iof iLiver iCancer iUsing iSvmTechniqes i[99]
isegmentation. iTo iexamine ithe iFCN iworking ion ia idefines ithe ivarious iimage iprocessing imethods ifor iautomatic

imoderately ismall idataset iand iparallel iit ito ipatch ibased iCNN idetection iof itumor iin ihuman iliver.

iand isparsity ibased iclassification ischemes. iAssaf iHoogi iet ial.

isuggested ia iunique imethod, ithe iadaptive ilocal iwindow, ifor

iimproving ithe ilevel iset iin isegmentation itechnique. iThe

iwindow iis iassessed iindependently ifor ievery ioutline ipoint,

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4174 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

iThe ichances iof ipersistence ihaving iliver itumor idepends ion iliver ifloor i(obstacles) iand ilesion ilocalizations iis irequired ifor
iearly idetection iof itumor iand ithen iclassification ias imalignant isegmentation iof iliver itumour. iThis ipaper iuses ithe iadaptive
i(cancerous) iand iBenign i(non-cancerous) itumors. iWei iYang ithresholding itechnique ifor ithe isegmentation iof iliver itumour.

iet ial. iEven ithough iinteractive iimage isegmentation ihas ibeen iPriyanka iKumar iet ial. iThere's ia iupward ithrust iin ithe iloss iof

ibroadly iexploited, icurrent iapproaches ipresent iunsatisfactory ilife icharge iof iliver imost icancers isufferers. iThe iprice iof iliver

iresults iin imedical iimage iprocessing. iThis ipaper isuggests ia icancer iis iincreasing iyr iwith ithe iaid iof iyr. iIn inormally, ithe

ifast imethod ifor iinteractive iCT iimage isegmentation iin iwhich iliver imost icancers’s idying iprice iis ivery iexcessive ibecause iof

ithe itumor iregions ishould ibe isegregated ias iforeground ilow ior ino isymptoms, ias iit iis iasymptomatic; iit iis inot icaught

iagainst ithe ihealthy itissues i[100]. i itill ithe iin ifinal istages. i[108]. iLaurent iMassoptier iet ial. iThe

Yuri iY. iBoykov iet ial. iIt istates ithat icertain ipixels iare iinformation iof ithe iliver istructure iwith iblood ivessels

imarked ias i“object” ior i“background” ito iprovide ihard itopography, iliver ifloor iand ilesion ilocalizations iis iin igeneral

iconstraints ifor ithe isegmentation. iAdditional isoft iconstraints iobligatory iin iremedies isimilar iliver iablations iand

icombine iboth iboundary iand iregion iinformation. iGraph icuts iradiotherapy i[109]. iIn ithis ipaper, ipropose ia itechnique ifor

iwere iused ito ifind ithe iglobally iideal isegmentation iof ithe icomputerized isegmentation iof iliver icomplicated igeometries.

iN-dimensional iimage i[101]. iIndrajeet iKumar iet ial. iThe iside iIt iincludes iof imaking iuse iof ia igraph-cut iapproach iinitialized

idetection iis ia itechnique iof idetecting isharp idiscontinuities iin ithrough ian iadaptive ithreshold. iA iparametric iassessment iwith

ian iimage. iThe idiscontinuities iare iunexpected iadjustments iin ithe ioutcomes iexecuted iby iusing iearlier ialgorithms ibased

ipixel iintensity istage i[102]. iThe iold iapproach iof iside itotally ion iactive icontour iis ilikewise ifinished iand ideliberated.

idetection iconsists iof iconvolving ithe iphotograph iwith ian iMain idemerits iof ilively icontour itechniques iend iresult ito ibe

ioperator i(2- iD ifilter) ithat's iconstructed ito ibe itouchy ito inoise. iconquer iand isegmentation iis ibetter-high-quality. iFeasibility

iEdge idetector iis ia iset iof iimportant ilocal iimage iprocessing ito ifrequently iuse igraph-cut itechnique ifor iautomatic iliver

imethod ito ilocate isharp ichanges iwithin ithe idepth icost. i isegmentation iis ialso iworked iout. i

A ifully iautomatic iand irobust ialgorithm ifor irapid Segmentation iof iCT iliver iimages i[110] isupport ito ianalyse
isegmentation iof iliver itissue iand itumors ifrom iCT iscans i[103] ithe ipresence iof ihepatic itumor iand icategorise ithe itumor ifrom

iincluding iliver isurface iand ilesion ilocalization iis inormally iimages iof idiseased ipopulations. i iUsing iregion igrowing

irequired iin itreatments ifor ioccurrence iliver itumor iablations itechnique ito iexamine ithe ineighbouring ipixels iof iinitial iseed

ior iradiotherapy. iA inew imethod iand iequivalent ialgorithm ifor ipoints iand idecide iwhether ior inot ithe ipixel ineighbours ishould

ifast isegmentation iof ithe iliver iand iits iinternal ilesions ifrom ibe iadded ito ithe iregion. iThe iprocess iis iiterative iand iseed

iCT iscans iis iconferred iin ithis ipaper. iMallikarjun iKesaratti iet ipoint iis iselected iinteractively iin ithe isuspected iregion. iThe

ial. iLiver iCancer iis ione iin ievery iof ithe iswiftest igrowing icontour imade iby ithe iregion igrowing ihas ibeen isegmented

icancer iin ithe iworld. iThe iearly idetection iand iidentifying iof iusing iwatershed imethod. iThe isegmented iregion iof itexture

iliver itumor igrowth iis ivital ifor ithe idifficulty iof iliver itumor ifeatures iare iextracted ithrough iGray iLevel iCo-occurrence

igrowth i[104]. iSegmentation iof iliver ifrom imedical iimages iMatrix i(GLCM). iYong iYang, iproposed ito iengulf ithe inoise

ifrom ithe iabdominal ispace iis iimportant ifor idiagnosing iof isensitivity iof iconventional ifuzzy ic-means i(FCM) iclustering

itumor iand ifor imedical iprocesses. iYamini iUpadhyay iet ial. ialgorithm, ia inovel iextended iFCM ialgorithm ifor iimage

iImage isegmentation imentions ia iprocedure iby iwhich ia iraw isegmentation ibe istated iin ithis isurvey. iThis iset iof irules iis

iimage iis idivided iinto inon-overlaying iregions isuch ithat ieach iestablished iby iconverting ithe ireason iof ithe iusual iFCM iset iof

iregion iis ihomogenous iand iassociated. iIn ithis, idirectly irules iwith ia iresult itime iperiod ithat itakes iinto iclarification ithe

icontracts iwith ilocating iand ithus icalculating ithe isize iof iimpact iof ithe ifollowing ipixels ion ithe icentre ipixels i[111]. i

icancer iaffected iarea[105]. iIt ialso isuggests ito idecrease ithe Piotr iPorwik iet ial. iOffer, iImage iprocessing iand ianalysis
itime iand iphysical iefforts iincluded iin ilearning ithe iMR ibased itotally iat ithe inon-stop ior idiscrete ipicture itransforms

iimages iof ithe ipatient, ithus ivalid ithe iprecious itime iof ithe iare iclassic imethods. iThe iphoto ichanges iare iwidely iused iin

idoctors. iThe igoal iof ithis ipaper iis ito isimplify ithe ioffensive iimage ifiltering, istatistics idescription, iand iso ion. iPresently

istudy iproblems iassociated ito ithe istudy iof iMR iimages. i ithe iwavelet itheorems imake iup ivery ipopular imethods iof

The istudy iof iMR iimages irelated ito icancer idetection iin ithe iimage iprocessing, ide-noising iand icompression. iConsidering

iliver ior iabdominal iarea ihas ibeen istimulating. iThe ireason iis ithat ithe iHaar icapabilities iare ithe ibest iwavelets, ithese iforms

icountered ias ithe ishape icomplexity iand iover ilapping iof iliver iare iused iin imany itechniques iof idiscrete iphotograph

iwith iother iorgans. iV.Ulagamuthalvi iet ial. iUltrasound iliver itransforms iand iprocessing i[112]. iAlison iNoble iet ial.

itumor iimage iare inaturally ihave imore ispackle inoise. iEvaluates iultrasound isegmentation imethods, iin ia ibroad

iAutomatic idetection iof itumor iimage iin iultrasound iis ia isense, iconcentrating ion itechniques iestablished ifor imedical

ichallenging itask. iIn ithis iproposed isystem, iTo iapproach ifully iB-mode iultrasound iimages. iFirst, icurrent ic ievaluation iof

iautomatic imachine ilearning isystem ifor iidentifying ithe iliver iarticles iby iclinical iapplication ito ibest ipart iof ithe imethods

icancer itumor ifrom iultrasound iimages i[106]. iFirst iprocess iis ithat ihave i[113].In ithis istudy, iArtificial iNeural iNetwork

isegmenting ithe iliver iimage iby icomputing ithe itextural i(ANN) iand iSupport iVector iMachine i(SVM) ifor iliver icancer

ifeatures ifrom ico-occurrence imatrix iand irun ilength imethod. iclassification. i

iThis iis ione iof ithe ibest imethods ifor isegmentation iof The ipresentation iof iboth imodels iis iassociated iand
iultrasound iliver itumor iimages ias iit iis inot iaffected iby ispeckle iauthenticated ion iBUPA iLiver iDisorder iDataset iin iterms iof

inoise iand ialso ipreserves ithe ispatial iinformation. iFor iaccuracy, isensitivity, ispecificity iand iArea iunder iCurve

iclassification iof iSupport iVector imachine iare igeneral i(AUC) i[114].

ialgorithm ibased ion ithe irisk ibounds iof istatistical ilearning

itheory.

Automatic iLiver iTumour iDetection iin iAbdominal iCT

iImages i[107] igives ithe iinformation iof ithe iliver istructure,

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4175 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

iIn i“Texture ianalysis ifor iliver isegmentation iand iNeural iNetwork i(MPNN) iin iaggregate iwith ifeature
iclassification”, iseveral itextural ianalysis itechniques iproposed idescriptors ithat iare igenerated iby iusing ifractal ifunction
ito iextract iuseful ifeatures ifor iliver itissue iclassification i[115, istatistics iand ithe igrey ilevel ico-prevalence imatrix. iIn ithe

i32], igray ilevel idifferent istatistics i(GLDS), igray ilevel irun ipaper i“A ihybrid iapproach ifor iliver isegmentation”, iapproach

ilength istatistics i(RUNL), igray ilevel ihistogram, ifirst iorder iconsist iof ifour istages. iIntensity ibased ipartition, iregion ibased

iparameters i(FOP), iFourier ipower ispectrum i(FPS) iare ithe itexture iclassification, iInitial iregion iof iliver iand iintensity iof

itechniques iused ito ianalysis ithe iliver isegmentation iand ithe iinitial iregion i[126]. iSome iof ithe itechniques iused ifor

iclassification. iEffective iliver iimage isegmentation ifrom isegmentation iare ithresholding, iregion igrowing, iclustering,

iabdomen icomputer itomography iusing iK-mean itechnique iMarco iRandom iField, iartificial ineural inetworks, iclassifier,

idescribes ian iautomatic iliver iimage isegmentation iusing ithe ideformable imodels iand iatlas iguided iapproach.

iabdomen iCT iimage iusing iK-mean iclustering ifollowed iby ithe Medical iDiagnosis iFor iLiver iCancer iUsing iClassification
imorphological ioperation. iIt ishows ia ibetter iperformance iand iTechniques i[127], icategorised ithe iLiver iCancer iDataset

iminimum iorientation ieffect ion isegmented iimage i[116, i117]. iusing ithe iDecision iTree iInduction i(J48 iDecision iTree), ithat

iDiagnosis iand ianalysis iof iautomated iliver iand itumor ithe ioverall iperformance iof ithe iJ48 idecision itree ihas ibetter

isegmentation ion iCT iuse ithe itechniques ilike iRegion iof ithan iother iclass ialgorithms iand iit itakes ifew itimes ito ibuild

iInterest i(ROI), iGray iLevel iCo-occurrence iMatrix i(GLCM), ithe ichoice itree ithan iothers ialgorithms. iFactors ithat ihave ian

iSupport iVector iMachine i(SVM) iand iFuzzy iC-mean ieffect ion ithe iclassifier’s ioverall iperformance iare: i1. iData iset

iClustering i[118]. iIt iassistances ithe iphysician iand iradiologist i2. iNumber iof ituples iand iattributes i3. iType iof iattributes i4.

ifor iliver itumor idetection iand idiagnosis ifor itumor isurgery. iIn iSystem iconfiguration. iVarious iData iclassification

ithe ipaper iDetection iand iClassification iof iLiver iCancer iusing itechniques ior ialgorithms iare iused ito iresolve ithis iproblem. I

iCT iImages, iit iis ipossible ito idetect ithe inormal iliver iand irest Some icategory itechniques ior ialgorithms iare iDecision itree,
ias iabnormal iliver i[119]. iCLAHE iworks imainly ifor iC4.5, iAssociation irule, iBayesian inetworks, iSupport ivector

iextracting ithe iuseful iresult iwith ivisualization iand iin ibox iplot iMachine, iK-NN, iNeural inetworks iand imany iothers iused ion

iit iis iclearly idifferentiating ithe inormal iliver iregion iwith ithis ipaper. iClassification itechniques iused ithe ischooling iand

iabnormal itumor. iEasily icomparable iwith ithe ibox iplots iusing itake ia ilook iat istatistics iset ito icategorize ithe ifacts iand ito

iboth ithe iCLAHE iand iCVHE ialgorithm. iconstruct ia imodel ior ito ifind iout ithe ihidden iknowledge ifrom

Liver itumor isegmentation iin inoisy iCT iimages iusing ithe idatasets iFeature iextraction ifor ihuman idetection ithe

iDistance iRegularized iLevel iSet iEvolution i(DRLSE) ibased iusage iof iHOG iand iCS-LPB istrategies[128],

ion iFuzzy iC-means i(FCM) iclustering iworks ion ithe ibase iof icentre-symmetric inearby ibinary istyles i(CS-LBP) iand

iFCM ialgorithm, iDRLSE ialgorithm iand imedian ifiltering iHistogram iof ioriented igradients i(HOG) ifunction iextraction

ialgorithm i[120]. iIn ithis ipaper, ithe iliver itumor iidentified iby istrategies iare iexisting. I

ithree istages. iThey iare ipre-processing istage, iprocessing istage The idrawback iof iHOG icharacteristic iextraction, iit
iand idetection istage. iIn ithe ipaper i“Fully iautomatic iliver iproduces ifurthermore iseveral ifunction istyles, iintricate ito

isegmentation icombining imulti-dimensional igraph icut iwith ilook iat iand iis itime iconsuming. Face iRecognition iMethod

ishape iinformation iin i3D iCT iimages”, ian iautomatic iapproach iwith iTwo-Dimensional iHMM i[129], i2D-HMM iand iuses iit

iintegrating imulti-dimensional ifeatures iinto igraph icut ito idiscover ifaces. IThe iapproach ilets iin ifor iquicker iface

irefinement iis ideveloped iand ivalidated i[121]. iMulti iatlas iprocessing iand ipopularity idue ito ithe ifact ithere imay ibe ino

isegmentation, iautomatic igraph iconstruction iand ineed ito itrade ithe i2-D iinput irecords iinside ithe iphotograph

imulti-dimensional igraph icut itechniques iwere iused. iform ito ia ione-dimensional ifacts, iand ifor ithis ireason ino ilack

iRamamoorthy iSuganya isays iclassification iand iretrieval iof iof istatistics icontained iin ithe iphotograph. i[130], iIt iattention

ifocal iand idiffused iliver ifrom iUltra isound iimages iusing ion imodelling ithe icirrhosis iand ihepatocellular icarcinoma,

imachine ilearning itechniques i[122]. iSpeckle ireduction iby ihowever ithe ieveryday ikingdom iand icontinual iviral ihepatitis

iModified iLaplacian iPyramid iNonlinear iDiffusion i(CVH) iare ialso itaken iinto iconsideration. IThe ireason iis ithat

i(MLPND), iMutual iInformation i(MI), iand iMI ibased iimage iof isupplying ia idependable iapproach ifor inon-invasive

iregistration, iimage itexture ianalysis iby iHarlick’s ifeatures, icharacterization iof ithe ievolution iin ithe idirection iof iHCC,

iimage iclassification iand iretrieval iby imachine ilearning ithat iallows iyou ito iprevent ithis imalignant itumor. iImage

ialgorithm. iIgnisha iRajathi istates ithat ia ihybrid icollaborative isegmentation ialgorithms iprimarily ibased ion ifuzzy iclustering

iclassifier iwith iSupport iVector iMachine i(SVM), iK-Nearest i[131].

iNeighbour i(KNN) iand iRandom iForest i(RF) iclassifier iused The iprocess iof idetachment ia ivirtual iphoto iinto idistinct
ito icategorize iliver idiseases i[123]. iClassification iusing iareas iwhich ihave iassociated imatters iincluding igrey idegree,

iHybrid iWhale iOptimization iwith ireplicated iAnnealing iand icoloration, itexture, ibrightness iand iso iforth. iSubsequently

iCollaborative iclassifier. ithat ithe iimage ican ibe imore icomprehensible iand ibeneficial ito

Automatic itype iof iliver idisorder ifrom iextremely isound iinspecting. iOn ithe ibasis iof ipixel iintensity iyou ipossibly ican

iimages ithe iuse iof iGLRLM itexture icapabilities i[124], ithis idifferentiate ithe ibounds iof idifferent igadgets. iIt iidentifies

ipaper ievaluation ithe iimpact iof idiverse ilinear, inon-linear iand iseparate iitem iwithin ian iphoto iand iadditionally idiscover

idiffusion ifilters iin ienhancing ithe isatisfactory iof ithe iliver iboundary ibetween ispecial iareas. iIt imay ibe ilabeled iinto itwo

iultra isound iphotographs ibefore iproceeding ito ithe isubsequent isorts: ineighborhood isegmentation iand iglobal isegmentation.

i ilevels iof ifunction iextraction iand icategory iusing iGray iLevel iLocal isegmentation iis ismall iwindows ion ian ientire

iRun iLength iMatrix i(GLRLM) ifunctions iand iSupport iVector iphotograph iand ideal iwith isegmenting isub iphoto.

iMachines i(SVM). iE-Liang iChen iet ial isays ian iautomatic

idiagnostic imachine ifor iCT iliver iimage icategory. iThe ineural

inetwork iis iencompassed ito icategorize ithe iliver itumors

i[125]. iIt iis ifound iout iby iway iof ia iModified iProbabilistic

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4176 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

Table 6: A Brief thresholding features

Authors Year Technique Features Limitations Evgin, 2015 K-means – Less Higher
Used Mehmet based, Gaussian computation computation
Eva Van 2007 Voxel Incorrect Long [58] model, al time al cost
Rikxoort classification, border can be computation Deformable
[28] Atlas matching found, time model
Localization Abhay 2015 Constrained Normal data The complete
of the border Krishnan Variable is detected classification
M.C. 2009 Functional Best Can [86] Histogram from the is not
Jobin Magnetic correction further Equalization abnormal possible due
Christ Resonance for each improved in (CVHE), data, to some not
[100] Imaging defect, accuracy Contrast Preservation defined
(FMRI), maximized Limited of the parameter
(SPECT), by active Adaptive globalization values
Statistical contour Histogram of an image
Parametric optimization, Equalization
Maps (CLAHE)
(SPM),etc. Rakesh 2017 K-means Ability to Performance
Jiuqing 2010 Wavelet packet Better More Prbhu clustering, extract the can further
Wan transforms, B discriminate effective [84] Thresholding liver region be improved
[89] mode ultra normal class features can techniques, from the
sound liver from be wavelet filter abdomen CT
images, SVM, cirrhosis, investigated image,
Spatial domain More reliable
based, CAD. consistent. IX. COMPARATIVE ANALYSIS
Annalisa 2010 Transient High Cannot be
Berzigott Elastography diagnostic applied to In ithis isection, ito ishow ithe iperformance ianalysis iresults iof
i (TE), Left Live accuracy, some ithe ivarious iclassification ischemes. iThe iclassification
[91] lobe Surface patients like
iprocess iinvolves itwo iparts: ione iis itraining ipart iand ianother
(LLS) Hepatic obese and
Venous liver stiffness iis itesting ipart. iIn iorder ito itrain ithe iMSVM iand iANN ithe

Pressure person ifeatures iwhich iare iextracted ifrom ithe isegmented iregions ithat
Gradient iare iapplied ito ithe iclassifier iin iorder ito itest ithe iproposed
(HVPG),
ischemes. i iTable i1 idemonstrates ithe icomparison iof ivarious
Transjugular
Liver Biopsy iparameter ivalues ibased ion ifeature ivalues. iANN iprovides
(TJLB) ibetter iperformance ithan iother imethods. iTable i2
Ricardo 2011 SVM, Multi Improve the Must be
idemonstrates ithe icomparison iof iparameters ibased ion
Ribeiro model feature classification expanded to
[99] set, KNN, accuracy, incorporate iclassifier iwise. iAmong ithat iANN iwill iprovide ibetter

Radial basis Multi feature more textural iperformance.

kernel classified and multi and
classifier morphologic
system al features Table 7: MEAN, STANDARD DEVIATION (SE), AND
Oussema 2011 Deformable Solve the Less MUTUAL CORRELATION (MC) OF THE SELECTED
Zayane models, difficulties in accuracy FEATURES
[29] Max-flow/ automatic Featur Normal Chronic Hepatic
Min-cut, segmentation es Mean SD MC Mean SD MC
Histogram Glycem 119.34 47.54 0.075 120.67 59.5 0.164
Threshold Tail ia 6
Saima 2011 Gray level Better More texture LDH 418.33 204.5565 0.11 283.34 266. 0.35
Rathore difference discriminatin measure 3 45
[83] statistics, Gray g power than needed for Disease 0.05 0.21 0.05 2.63 1.85 0.155
level run length others, better Cause
statistics, summarize comparative
Fourier power texture study
spectrum, analysis Table 8 Parameter Values
Wavelet feature method Class Best Performance
A. 2013 Various type of Potentially Result can be DR (%) OA (%) J Classifier
Srinivas snake powerful improved Normal 98.12 98.1 0.976 MSVM
an and S. techniques, tools for the when Normal 98.6 98.5 0.984 ANN
Sundara Active Contour segmentation combined Normal 97.34 96.67 0.967 SVM
m of images with Chronic 87.98 89.7 0.88 MSVM
[132] appropriate Hepatic
filter
Chronic 89.8 88.78 0.876 ANN
B. 2014 Fuzzy C-mean Boundaries Ambiguities Hepatic
Balakum clustering, Clear, in
Chronic 85.98 84.7 0.823 SVM
ar [30] Level set Extraction. identification
Hepatic
M. 2014 Support Vector Image can be Image
Rajeshw Machine, Fuzzy analysed and quality has to
aran C-means, classified. be improved.
[62] Pre-processing
Ramshee 2014 ROI selection, Improve Unable to
ja fuzzy C-mean, better describe size
[85] SVM, GLCM, classification , shape and
Guassian filters, , Significant area
Adaptive reduction of

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4177 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

X. RESULTS Computed Tomography", International Journal of Computer

Applications, Vol. 74 – No. 11, 2013.
The imulti-plane iintegrate itechnology ihelps icommunity iget 8. Luoweinbin, “Efficient removal of impulse noise from digital images”,
igreater ihigher iaccuracy, ithat iis iespecially ibenefited ifrom Consumer Electronics, IEEE transaction, pg 523-527, 2006.
iusing ithe i3-D icontextual iinformation iin ithree iorthogonal
9. S. Perumal and T. Velmurugan, “Pre-processing by contrast
enhancement techniques for medical images”, International Journal of
iplanes. iAnd ithe idense iconnection iblended iwith iadverse
Pure and applied mathematics, Vol 118, no. 18, 2018.
ischooling iapproach iin iaddition iimproves ithe imodel ioverall 10. Evans Baidoo, Alex Kwesi Konton, “Implementation of gray level
iperformance idue ito ithe ifact ithe ideep ireuse iof ispatial ifacts image transformation technique”, I.J Modern education and computer
ifrom itumours isequence iand iits iglobal ilook.
science, 2018.
11. Schowengerdt, Robert, “ A remote sensing: Models and methods image
processing”, Academic press, 2006.
12. Manpree Kaur, Sunny Behal, “Study of image denosing and its
techniques”, IJRCSSE, Vol 3, Issue 1, January 2013.
13. Konya, Iuliu Vasile, “Adaptive methods for robust document image
understanding. Diss.”, University of Bonn, 2012.
14. Javier Sanchez, Nelson Monzon, Augustin Salgado, “An analysis and
implementation of the Harris corner detector”, IPOL, 2108.
15. Ehab Salahat and Murad Qasaimeh, “Recent advances in feature
extraction and description algorithms: A comprehensive survey”, March
2017.
https://micro.magnet.fsu.edu/primer/java/digitalimaging/processing/dif
fgaussians/index.html.
16. Samta Gupta, Susmita Ghosh Mazumdar, “Sobal Edge detection
algorithm”, International Journal of Computer Science and
Management result, Vol 2, Issue 2, 2013.
17. Amruta Kabade, Dr. V.G Sangam, “Canny edge detection algorithm”,
Fig. 5 Segmented Image IJARECE, Vol5, issue 5, May 2016.
18. Bhosale Swapuali B, Kayasatha Vijay, Haipale Varsha, “Feature
XI. CONCLUSION AND FUTURE WORK extraction using SURF algorithm for object recognition”, IJTRA,
Vol12, Issue 4, 2014.
This ipaper istudies ithe ione-of-a-kind istrategies iworried 19. Ricardo T. Ribeiro, Rui Tato Marinho, and J. Miguel Sanches,
iwithin ithe ivarious idegrees iin iphoto iprocessing. iEach imethod Classification and Staging of Chronic Liver Disease From Multimodal
imakes iuse iof ione ior igreater iset iof irules ifor iits iprocessing.
Data. IEEE Transactions on Biomedical Engineering. May 2013.
20. Jiuqing Wan and Sirui Zhou, “Features Extraction Based on Wavelet
iThough ievery iset iof irules iexcels iin ia isingle iway, iit ihas isome
Packet Transform for B-mode Ultrasound Liver Images”. International
idifficulty iinside ithe iother. iFor isome iexamples, iinside ithe Congress on Image and Signal Processing 2003.
isegmentation iprocedure ithe iarea ideveloping ialgorithm ihas 21. Fei Yang and AyooluwakunmiJeje, “Automatic Liver Tissue Section
ithe iadvantage iof idecrease itime icomplexity ihowever iit Image Characterization”, Conference proceedings, Stanford University.
2016.
ipossesses ia idrawback ias iover isegmentation. iK iway
22. Annalisa Berzigotti and Juan G. Abraldes,“Ultrasonographic evaluation
iclustering ihas ia ibonus iof igreen isegmentation ibut iit ican
of liver surface and TE in clinically doubtfulcirrhosis”, European
ihandiest ibe iused ifor ilow iassessment iimage. iIn iin iaddition Association for the study of Liver. Journal of Hepatology. 2010.
iexamine, iit iis ifound ithat icombining imore ithan ione ialgorithm 23. TejasIndulalDhamecha, Praneet Sharma, Richa Singh, and Mayank
Vatsa, “On Effectiveness of Histogram of Oriented Gradient Features
igives ia ibetter iend iresult ithan ithe iusage iof iit iman ior iwoman.
for Visible to Near Infrared Face Matching”. IEEE Xplore. 2014.
iOn ithe ibase iof ithis imethod, iif iwe iuse ihigher irunning icombos
24. Ricardo Riberio, RUI Tato Marinho, Jose velosa,FernandoRamalho, J.
iof ialgorithms iin ieach istage iof iphotograph iprocessing, iwe Miguel Sanches, “The Usefulness of Ultrasound in the classification of
ican iimprove ithe idetection iand iclassification iof iliver iillnesses Chronic Liver Disease”. IEEE Xplore. 2011.
iin iin iadvance iranges iand iwith ibetter iaccuracy.
25. M.C.Jobin Christ, K.Sasikumar, and R.M.S.Parwathy, “Application of
Bayesian Method in Medical Image Segmentation”. TECHNIA
International Journal of Computing Science and Communication
REFERENCES Technologies, July 2009.
26. NavneetDalal, “Histograms of Oriented Gradients for Human
1. Gunasundari, S. 2and Ananthi.,“Comparison and evaluation of methods Detection”. International Conference on Computer Vision.2010.
for liver tumor classification from CT datasets”, International Journal of 27. [28] Nupur Thakur, Deepa Maheshwari, “A review of image
Computer Applications, Vol. 39, No. 18, pp. 46-51, 2012. classification techniques”, IRJET, Vol4, Issue11, Nov 2017.
2. Tarek M. Hassan et al, “Medical image segmentation for liver disease – 28. [29] Harsha Pakhale, Deepak Kumar Xaxa, “A survey on diagnosis of
A survey”, International Journal of computer application, Vol 118, No. liver disease classification”, International journal of engineering and
19, May 2015. techniques, Vol 2, Issue 3, 2016.
3. Klaus D. Toennies – “Guide to medical image analysis methods and 29. Dr. S. Vijaya Rani, Mr. S. Dhayanand, “Liver disease prediction using
algorithms”, Springer 2017 SVM and Navie Bayes algorithm”, International journal of science and
4. B. Dong, A. Chien, and Z. SHEN, "Frame Based Segmentation for engineering technology, Vol 4, Issue 4, April 2015.
Medical Images," Communications inMathematical Sciences, Vol. 32, 30. R.V. Patil, S.S. Sannakki, V.S. Rajpurohit, “A Survey on classification
No.4, pp.1724-1739, 2010. of liver diseases using image processing and data mining techniques”,
5. Jinshan Tang, SosAgaian, and Ian Thompson, " Guest Editorial: International journal of computer sciences and engineering, Vol 5, Issue
Computer-Aided Detection or Diagnosis (CAD) Systems", IEEE 3, 2017.
System Journal, Vol. 8, No. 3, September 2014. 31. S. N. Deepa and B. Aruna Devi, “A survey on artificial intelligence
6. Sheenum Marwaha, Himanshu Monga,andShelza, " Automatic approaches for medical image classification”, Indian journal of science
Diagnosis Systems Using Image Processing- A Systematic Study", and technology, Vol 4, No. 11, 2011.
IRACST - International Journal of Computer Science and Information 32. Bendi Venkata Ramana, Prof. M.Surendra Prasad Babu,, Prof. N. B.
Technology & Security (IJCSITS), ISSN: 2249-9555 Vol. 2, No.2, April Venkateswarlu, “A Critical Study of Selected Classification Algorithms
2012. for Liver Disease Diagnosis”. International Journal of Database
7. Gunasundari S and Janakiraman S, "A Study of Textural Analysis Management Systems (IJDMS).2011
Methods for the Diagnosis of Liver Diseases from Abdominal

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4178 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

33. Vishakha V. Hambire1, Dr. S. R. Ganorkar, “Classification of Liver 57. Grzegorz Chlebus, Andrea Schenk, Jan Hendrik Moltz, Bram van
Disease Based on US Images”, International Research Journal of Ginneken, Hans Meine, “Deep learning based automatic liver tumor
Engineering and Technology (IRJET) 2015 segmentation in CT with shape-based post-processing”, Fraunhofer
34. Saurabh Singh, Dr. Ashutosh Datar, “EDGE Detection Techniques MEVIS, Bremen, Germany.
Using Hough Transform”. International Journal of Emerging 58. Yogita Ashok Deore, Prof. Namrata D.Ghuse, “Efficient Image
Technology and Advanced Engineering.2013. Processing Based Liver Cancer Detection Method”, International
35. S. Sundaram and T.M Sridhar, “Cancer identification using Journal on Recent and Innovation Trends in Computing and
probabilistic boundary value edge mapping”, AENSI, January 2014. Communication. Volume: 4 Issue: 12.
36. M.S. Abirami and Dr. T. Sheela, "Analysis of Image Segmentation 59. Eva van Rikxoort, YuliaArzhaeva, and Bram van Ginneken,
Techniques for Medical Images", Proceedings of International “Automatic segmentation of the liver in computed tomography scans
Conference on Emerging Research in Computing, Information, with voxel classification and atlas matching”.
Communication and Applications, 2014. 60. Oussema Zayane1, Besma Jouini1, Mohamed Ali Mahjoub2,
37. Salem Saleh Al-amri, N.V. Kalyankar, and Khamitkar S.D, " Image “Automatic liver segmentation method in CT images”,Canadian Journal
Segmentation by Using Threshold Techniques", Journal of computing, on Image Processing & Computer Vision Vol. 2, No. 8, December 2011.
Vol. 2, Issue 5, ISSN 2151-9617, May 2010. 61. B.BalaKumar,A.Mohamed Syed Ali , “Integrating spatial fuzzy
38. Er. Pratibha Thakur and Er. Nishi Madaan, "A survey of image clustering with level set methods for liver segmentation from computed
segmentation techniques", International journal of research in computer tomography scans”, Computer Applications: An International Journal
applications and robotics, Vol.2 Issue.4, pp. 158-165, April 2014. (CAIJ), Vol.1, No.1, August 2014.
39. P.Elayaraja and M.Suganthi, "Survey on Medical Image Segmentation 62. Dr. K.A.M. SajadHyder, “Least Distance Based Liver Segmentation
Algorithms", International Journal of Advanced Research in Computer Technique (LDBLST) For Liver Segmentation from CT Scan Images”.
and Communication Engineering Vol. 3, Issue 11, November 2014. International Journal of Pure and Applied Mathematics, 2018.
40. K. S. Selvanayaki and R.M.Somasundaram , "A Survey on Image 63. Abhijeet Patil, Prof. Mrs. V. S. Kulkarni, “A Review: Liver Cancer
Segmentation Techniques for Edge Detection", International Detection Algorithm”, International Journal of Advanced Research in
Conference on Innovation in Communication, Information and Computer and Communication Engineering Vol. 5, Issue 6, June 2016.
Computing (ICICIC), 2013. 64. Ahmed M. Anter, Ahmad Taher Azar, Aboul Ella Hassanien , Nashwa
41. Rajiv Kumar , Arthanariee A. M., "A Comparative Study of Image El-Bendary, Mohamed “Abu ElSoud Automatic computer aided
Segmentation Using Edge-Based Approach", International Journal of segmentation for liver and hepatic lesions using hybrid segmentations
Mathematical, Computational, Natural and Physical Engineering Vol. 7 techniques”, Federated Conference on Computer Science and
No. 3, 2013. Information Systems 2013.
42. Santanu Bhowmik, Viki Datta, "A Survey on Clustering Based Image 65. Antonidoss and K.P. Kaliyamurthie,“Segmentation from Images Using
Segmentation", International Journal of Advanced Research in Adaptive Threshold”, Middle-East Journal of Scientific Research, 2014.
Computer Engineering & Technology Vol. 1, Issue 5, ISSN: 2278 – 66. Ina Singh, ,NeelakshiGupta, “Liver image segmentation using ant
1323, July 2012. colony based k means clustering and level set”, International Journal of
43. Rahul Basak, Surya Chakraborty, “Image sementation techniques: A Advances in Electronics and Computer Science, ISSN: 2393-2835
survey”, IRJET, Vol5, Issue4, 2018. Volume-2, Issue-9, Sept.-2015
44. Namrata Puri and Sumit Kaushik, “Analysis of the variants of watershed 67. Smriti Sahu, Maheedhar Dubey and Mohammad Imroze Khan,
algorithm as a segmentation technique in image processing”, ICACCT, “Performance Evaluation of Liver Ultrasound Image using Image
Vol3 Isuue3 2011. Processing Techniques”.
45. Shiuh-Yung Chen, Wei-Chung Lin, Chin-Tuchen, “Split and merge 68. Jianhua Liu, Zhongyi Wang, Rui Zhang,“Liver Cancer CT Image
image segmentation based on localized feature analysis and statistical Segmentation Methods based on Watershed Algorithm”,North China
tests”, CVGIP, Vol53, Issue5, 1991. University of Water Conservancy and Electric Power
46. Rahul Basak, Surya Chakraborty, “Image Segmentation Techniques: A 69. A.S. Thenmozhi , N. Radhakrishnan, “A Study on MRI Liver Image
Survey”, International Research Journal of Engineering and Segmentation using Fuzzy Connected and Watershed Techniques”,
Technology, Vol5, Issue4, 2018. International Journal of Data Mining Techniques and Applications.
47. Zhe Lin, Larry S. Davis, “Shape based human detection and December 2016.
segmentation via hierarchical part – template matching”, IEEE, Vol32, 70. Nuseiba M. Altarawneh,SuhuaiLuo, Brian Regan, Changming Sun,
Issue 4, 2010. “LIVER Segmentation From Ct Images Using A Modified Distance
48. Li. H. Wang Y, Lin KJ, Lo Se, “Lesion site selection by morphological Regularized Level Set Model Based On A Novel Balloon Force”.
enhancement and contextual segmentation” IEEE transaction in 71. Jithy P K, Philomina Simon, “Review on Automatic Segmentation
medical imaging, pg 289-301, 2001. Techniques in Medical Images”, International Journal of Scientific
49. Mohamed Gouskir, et al, “Automatic brain tumor detection and Research in Computer Science, Engineering and Information
segmentation for MRI using covarients of geodesic distance”, Technology , 2017 IJSRCSEIT, Volume 2, Issue 4, ISSN : 2456-3307.
International Conference on multimedia computing and system. 72. Hussein Alahmera,, Amr Ahmeda, “Computer-Aided Classification of
50. R. M. Haralick, K. Shanmugam, and I. Dinstein, “Textural features for Liver Lesions from CT Images Based on Multiple ROI”, International
image classification”, IEEE Transactions on Systems, Man and Conference On Medical Imaging Understanding and Analysis 2016,
Cybernetics, vol. 3, no. 6, pp. 610–621, 1973. MIUA 2016, 6- 8 July 2016, Loughborough, UK.
51. A. Chaddad, “Automated feature extraction in brain tumor by magnetic 73. Patrick Ferdinand Christ, Mohamed Ezzeldin A. Elshaer, Florian
resonance imaging using gaussian mixture models”, International Ettlinger,SunilTatavarty, Marc Bickel, “Automatic Liver and Lesion
Journal of Biomedical Imaging, vol. 2015, Article ID 868031, 11 pages, Segmentation in CT Using Cascaded Fully Convolutional Neural
2015. Networks and 3D Conditional Random Fields”.
52. J. Sachdeva, V. Kumar, I. Gupta, N. Khandelwal, and C. K. Ahuja, 74. Sajith A.G, Hariharan.S., “Spatial fuzzy C-means Clustering based
“Segmentation, feature extraction, and multiclass brain tumor Liver And Liver Tumor Segmentation on Contrast Enhanced CT
classification”, Journal of Digital Imaging, vol. 26, no. 6, pp. Images”. International Journal of Engineering and Advanced
1141–1150, 2013. Technology (IJEAT) ISSN: 2249 – 8958, Volume-4 Issue-3, February
53. Andreia, Andrade, Jose Silvestre Silva, Jaime Santos, “Classifier 2015.
approaches for liver steatosis using Ultra Sound images”, Elsevier, 75. Kumar SN, Lenin Fred, Muthu Kumar, Ajay Kumar, Sebastian
Procedia Technology 5, 2012 (763 – 770). International conference on Varghese, “A Voyage on Medical Segmentation Algorithm”, ISSN
health and social care information system and technologies 0970-938X, Biomedical research, 2018.
54. K.Mala and V.Sathasivam, “Automatic sementation and classification 76. Belgherbi, A., Hadjidj, I., Bessaid, A. “A Semi-automated Method for
of diffused Liver Diseases using wavelet based texture analysis and the Liver Lesion Extraction From a CT Images Based on Mathematical
Neural Network”, IEEE indicon conference 216-219(2005) Morphology”, Journal Of Biomedical Sciences -2013.
55. GUITAO,CAO et al, “Liver Fibrosis identification based on US 77. Huiling Guo, Haizhi Hu, Guobeng Fan, “A novel segmentation method
images”,Annual International conference of the IEEE Engineering in of medical CBCT inage in liver organ based on Bayesian network”,
Medicine and Biology facility(Feb-2005) Biomedical Research (0970 –
56. Nanzhou Piao, Jong-Gun Kim, and Rae-Hong,“Park segmentation of 938X), 2018 Special issue.
cysts in kidney and 3D volume calculation from CT images”,
International Journal of Computer Graphics & Animation (IJCGA)
Vol.5, No.1, January 2015.

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4179 & Sciences Publication
 Assessment on Liver Disease Classification using Medical Image Processing

78. Hariharan S, “Developments in Segmentation of Abdominal Liver CT 104. Yamini Upadhyay, Vikas Wasson, “Analysis of Liver MR Images for
Images for Cancer and Other Disease Lesion Detection”. Annals of Cancer Detection using Genetic Algorithm”, International Journal of
Advanced Biomedical Sciences. February 22, 2019. Engineering Research and General Science Volume 2, Issue 4,
79. S.S. Kumar, R.S. Moni, J. Rajeesh, “Automatic Segmentation of Liver June-July, 2014.
and Tumor for CAD of Liver”, Journal of Advances in Information 105. V.Ulagamuthalvi, D.Sridharan, “Automatic Identification of
Technology, vol. 2, no. 1, February 2011. Ultrasound Liver Cancer Tumor Using Support Vector Machine”,
80. Amitha Raj ,Jayasree M, “Automated Liver Tumor Detection Using International Conference on Emerging Trends in Computer and
Markov Random Field Segmentation”. International Conference on Electronics Engineering (ICETCEE'2012) March 24-25, 2012 Dubai.
Emerging Trends in Engineering, Science and Technology.(ICETEST - 106. Devendra Joshi, Narendra D Londhe, “Automatic Liver Tumour
2015). Detection in Abdominal CT Images”, International Journal of Computer
81. Qiuyang Huang, Yongxiong Sun. B, Liping Huang. C and Pei Zhang, Technology and Electronics Engineering Volume 3, Issue 1, February
“The liver CT image sequence segmentation based on region growing”. 2013.
5th International Conference on Advanced Engineering Materials and 107. Priyanka Kumar, Shailesh Bhalerao, “Detection of Tumor in Liver
Technology (AEMT 2015). Using Image Segmentation and Registration Technique”, IOSR Journal
82. Lauren K. Haaitsma, “Liver tumor segmentation in CT Images”, 2014. of Electronics and Communication Engineering, Volume 9, Issue 2,
83. Vishwajit B. Mohite, P. P. Belagali, “A study of different methods for April 2014.
liver tumor segmentation”, IRJET, Vol 3, Issue 9, 2016. 108. Laurent Massoptier, Sergio Casciaro, “Fully Automatic Liver
84. Maria Tsiplakidou, Markos G. Tsipouras, Nikolaos Giannakeas, Segmentation through Graph-Cut Technique”, Proceedings of the 29th
“Automated detection of liver histopathological findings based on Annual International Conference of the IEEE EMBS Cité Internationale,
biopsy image processing”, 2017. Lyon, France August 23-26, 2007.
85. Mredhula L, M.A. Dorairangaswamy, “Detection and classification of 109. M V Sudhamani, G T Raju, “Segmentation and Classification of
tumors in CT images”, IJCSE, Vol 6, Issue 2, 2015. Tumour in Computed Tomography Liver Images for Detection,
86. Dr. Alyaa H. Ali, Entethar M. Hadi, “Diagnosis of Liver tumor from CT Analysis and Preoperative Planning”, International Journal of Advanced
images using first order statistical features”, IJETT, vol 20 , No. 3, 2015. Computer Research, Volume-4 Number-1 Issue-14 March-2014.
87. Mohammad Goryawala, Magno R. Guillen, SeazGulec, “An accurate 110. Yong Yang, “Image Segmentation By Fuzzy C-Means Clustering
3D liver segmentation method for selective internal radiation therapy Algorithm With A Novel Penalty Term”. Jiangxi University of Finance
using a modified K means algorithm and parallel computing”, IJICIC, and Economics Nanchang 330013, P.R. China,2006.
Vol 8 No. 10(a), 2012. 111. Piotr Porwik, Agnieszka Lisowska, “The Haar–Wavelet Transform in
88. A. Mohamed Riyas, Dr. Mohamed Sathik, “Comparative study of Digital Image Processing: Its Status and Achievements”, Machine
segmentation techniques on CT liver images”, IJTRD, Vol 4(2), 2017. Graphics Vision vol .13, no. 1/2, 2004, pp .79-98.
89. Evgin, Mehmet, “A comparative performance evaluation of various 112. Alison Noble, DjamalBoukerroui, “Ultrasound image segmentation:
approaches for liver segmentation from SPIR images”, Turkish Journal ASurvey”, IEEE Transactions On Medical Imaging, Vol. 25, No. 8,
of Electrical Engineering & Computer Sciences, 2013. August 2006.
90. Avi Ben-Cohen, Idit Diamant, EyalKlang, Michal Amitai, Hayit 113. Sharifah Hafizah Sy Ahmad Ubaidillah, RoselinaSallehuddin,
Greenspan, “Fully Convolutional Network for Liver Segmentation and NoorfaHaszlinnaMustaffa, “Classification of Liver Cancer using
Lesions Detection”, Artificial Neural Network and Support Vector Machine. Proc. of Int.
91. Assaf Hoogi, Christopher F. Beaulieu, Guilherme M. Cunha, “Adaptive Conf. on Advances in Communication, Network, and Computing, CNC.
local window for level set segmentation of CT and MRI liver lesions”, 114. Saima Rathore, Muhammad Aksam Iftikhar, Mutawarra Hussain,
Elsevier 2017. Abdul Jalil, Texture analysis for liver segmentation and classification: a
92. Ina Singh, “A Study of Effective Segmentation Techniques for Liver survey, 2011 Frontiers of Information Technology.
Segmentation, International Journal of Advanced Research in Computer 115. Rakesh Prabhu, A. M. Khan, Effective Liver Image Segmentation from
Engineering & Technology”, Vol4 Issue 4, April 2015. Abdomen Computed Tomography Images Using K-Mean Technique.
93. M. Rajeshwaran, A. Ahila, “Segmentation Of Liver Cancer Using SVM International Journal of Electronics, Electrical and Computational
Techniques”, International Journal of Computing Communication and System September 2017.
Information System, Vol 6, No. 2, April 2014. 116. Pritesh Vora and Bhavesh Oza, “A survey on K-mean clustering and
94. Poonam Devi, Mrs. Poonam Dabas, “Liver Tumor Detection using particle swarm optimization”, International journal of science and
Artificial Neural Networks for Medical Images”, International Journal modern engineering (IJISME), ISSN:2319 – 6386, Vol 1, Isue 3, Feb
for Innovative Research in Science & Technology, Volume 2, Issue 3, 2013.
2015. 117. Ramsheeja. R R, Prof. Sreeraj. R, Diagnosis and Analysis of Automated
95. M. Usman Akram, Aasia Khanum, Khalid Iqbal, “An automated System Liver and Tumor,, November-2014.
for Liver CT Enhancement and Segmentation”, ICGST-GVIP Journal, 118. Abhay Krishan, Deepti Mittal, Detection and Classification of Liver
Volume 10, Issue IV, October 2010. Cancer using CT Images, International Journal on Recent Technologies
96. Rajesh G, SelwinMichPriyadharson A, “Liver cancer detection and in Mechanical and Electrical Engineering (IJRMEE), Volume: 2 Issue:
classification based on optimum hierarchical feature fusion with PeSOA 5.
and PNN classifier”, Biomedical Research 2018. 119. P. Yugander ; G. Raghotham Reddy, “Liver tumor segmentation in
97. Shikha Mandhan, Kiran Jain, “Tumor Segmentation and Automated noisy CT images using distance regularized level set evolution based on
Training for Liver Cancer Isolation”, International Journal of Advance fuzzy C-means clustering”, IEEE explore, January 2018.
Research, Ideas and Innovations in Technology, Volume2, Issue4, 2016. 120. Xuesong Lu, QinlanXie, Yunfeizha, Defeng Wang, “Fully automatic
98. Vinita Dixit, JyotikaPruthi, “Segmentation Of Liver Cancer Using SVM liver segmentation combining multi-dimensional graph cut with shape
Techniqes”, International Journal of Computer Science and Mobile information in 3D CT images” Scientific Reports 8, 2018.
Computing, Vol. 3, Issue. 3, March 2014. 121. RamamoorthySuganya, R. Kirubaharan and S. Rajaram, “Advances in
99. Wei Yang, Xiaolong Wang, Liang Lin, Chengying Gao, “Interactive CT signal processing and intelligent recognition systems”, Springer
Image Segmentation With Online Discriminative Learning”, 18th IEEE International Publishing, Switzerland, 2014.
International Conference on Image Processing, 2011. 122. G. IgnishaRajathi, G. Wiselin Jiji, “Chronic liver disease classification
100. Yuri Y. Boykov Marie-Pierre Jolly, “Interactive Graph Cuts for Optimal using hybrid whale optimization with simulated Annealing and
Boundary & Region Segmentation of Objects in N-D Images ensemble classifier”, Symmetry MDPI – 2019.
International Conference on Computer Vision”, Vancouver, Canada, 123. RagheshKirishnan. K, R. Sudhakar, “ Soft computing application”,
July 2001. AISC 195 PP (611 – 624) Springer Verlag Berlin Heidel berg 2013.
101. Indrajeet Kumar, Jyoti Rawat, Dr. H.S. Bhadauria, “A Conventional 124. E-Liang Chen, Pau-Choo Chung, “An Automatic Diagnostic System for
Study Of Edge Detection Technique In Digital Image Processing”, CT Liver Image Classification”, IEEE Transactions On Biomedical
International Journal of Computer Science and Mobile Computing, Vol. Engineering, VOL. 45, NO. 6, June 1998.
3, Issue. 4, April 2014. 125. RuchaneewanSusomboon, Daniela Stan Raicu, and Jacob Furst, “A
102. Laurent Massoptier, Sergio Casciaro, “A new fully automatic and robust Hybrid Approach for Liver Segmentation”, 2007.
algorithm for fast segmentation of liver tissue and tumors from CT 126. Reetu, Narender Kumar, “Medical diagnosis of liver cancer using
scans”, Article in European Radiology September 2008. classification techniques”,
103. Mallikarjun Kesaratti, Dr. Sunanda Dixit, Prof. B.I. Khodanpur, “A International Journal of Science
Survey on Different Methods for Liver Segmentation”, International and research, Vol 4, Issue 5, May
Journal of Advanced Networking & Applications. 2015.

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4180 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-5, January 2020

127. S.K. Uma, Surjana B.J., “Feature extraction for human detection using
HOG and CS-LPB methods”, International journal for computer
application.
128. JanuszBobulski, “Face recognition method with two dimensional
HMM”, Institute of compute and information science. 9 th International
Confrence on computer recognition systems CORES,2015,pa 317-325.
129. D. Mitrea, S. Nedvschi, M. Lupsor, R. Badea, “Exploring the textural
parameters obtained from ultra sound images for modelling the liver
pathological stages in the evolution towards hepatocellular carcinoma”.
130. R. Ravindraiah and K. Tejaswini, “A survey of image segmentation
algorithms based fuzzy clustering”, International Journal of Computer
Science and Mobile Computing, Volume 2, Issue 7, 200-206, July 2017.
131. A. Srinivasn and S. Sundaram, “Application of Deformable models for
In-Dopth analysis and feature extraction from medical images – A
review”, ISSN 1054 – 6618 Pattern Recognition Image Analysis, 2013
Vol 23, No 2 (296 – 318)

AUTHORS PROFILE

A. Bathsheba Parimala, has received the M.Sc degree

from Manonmaniam Sundarnar University, TamilNadu,
India in 2007 and M.Phil degree from Vinayaka
Missions University,Salem, India in 2009. She is
currently doing Ph.D degree in the Department of
Computer Science at Vinayaka Missions Kirupananda
Variyar Arts and Science College, Salem.Her Research interests Medical
Image Processing and Machine Learning.

Dr. R. S. Shanmugasundaram, has received the BE

degree in Electronics and Communication Engineering
from the University of Madras of India in 1996, ME
degree from the Bharathidasan University of India in
2001 and Ph.D degree from Anna University of India in
2015. His research interest are in medical image
processing, deformable models and segmentation. He is a life member of
ACS and ISTE. Currently he is working as Deputy Director (Academics) of
Vinayaka Mission’s Research Foundation (Deemed to be University),
Salem, Tamil Nadu, India.

Published By:
Retrieval Number: E6217018520/2020©BEIESP Blue Eyes Intelligence Engineering
DOI:10.35940/ijrte.E6217.018520 4181 & Sciences Publication