Volume 6, Issue 6, June – 2021 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Analysis of Various Medical Image File Formats

K Asha Deepa Mahadev
Assistant Professor, CSE Department Assistant Professor, CSE Department
BNM Institute of Technology BNM Institute of Technology
Karnataka, India Karnataka, India

Abstract:- A medical image is a portrayal of the inward can be portrayed. What the mathematical worth of the pixel
design or capacity of an anatomic locale. It has a variety communicates relies upon the imaging methodology, the
of picture components called pixels (2 dimensional) or obtaining convention, the remaking, and at last, the
voxels (3 dimensional). The quantity of pixels used to postprocessing. The first category is about the depth of pixel.
portray the field-of-perspective on a specific securing The pixel depth also called as depth of bit. It allows to store
methodology is an outflow of the detail with which the life the file in specific range. It works on the equation 2 raise to
structures or capacity can be portrayed. The the power n where n represents the depth of pixel. For
communication of the pixel data relies upon the imaging example, an raster which is having 8 bit can include 256
methodology, the obtaining convention, the remaking, and values which is considered as unique and its starts from zero
at last, the post-processing. This paper presents a and extends till 255. Bit depth has different type which
demystifying outline of major file formats right now usually depicts the limit of values for storage. The following
utilized in clinical imaging: Neuroimaging Informatics limits are for unsigned data type. If it is a single bit, it starts
Technology Initiative (Nifti) and Digital Imaging and from 0 and end at 1. If it is a two bit, it starts from 0 and end
Communications in Medicine (Dicom). Concepts which is at 4. If it is a four bit, it starts from 0 and end at 1. If it is a
common to all document designs, like pixel profundity, eight bit, it starts from 0 and end at 255. If it is a sixteen bit, it
photograph metric understanding, metadata, and pixel starts from 0 and end at 65535. If it is a thirty-two bit, it starts
information, are first considered. Then, at this point, the from 0 and end at 4294967295.The following limits are for
attributes and qualities of the different formats are signed data type. If it is eight bits, it starts from -128 and end
examined. at 127. If it is sixteen bits, it starts from -32768 and end at
32767. If it is thirty-two bits, it starts from -2147483648 and
Keywords:- Medical Imaging. Image Processing, File end at 2147483647. Real and imaginary part of complex data
Formats, DICOM, Nifti. consider as pairs of real numbers. Therefore, pixel depth of
complex data is double the single real number. Second is
I. INTRODUCTION about the Data Information. Information that explains the
image is called as metadata. For any image, information
A medical image data set comprises ordinarily of at about the image presents at the backside of pixel data. This
least one pictures addressing the projection of an anatomical information presents at the initial part of the file as a header
volume onto a picture plane (projection or planar imaging), a which consists of sizes, resolution, photometric interpretation.
progression of pictures addressing slim cuts through a volume Metadata plays an important role in the medical image file
(tomographic or multi slice two-dimensional imaging), a format. Various modalities produce information about the
bunch of information from a (volume or three-dimensional creation of images. For example, attributes of MRI are related
imaging), or numerous securing of the equivalent to sequence of pulse which is used. For example, e.g.,
tomographic or volume picture over the long haul to create a information about the time, angle how u flipped, acquisitions
unique series of acquisitions (four-dimensional imaging). count etc. positron emission tomography uses a radioactive
Medical images has different file formats which explains the drug and MRI uses magnetic fields which is so strong,
arrangement of the data within the image and also depicts the gradients of the magnetic field, and radio waves to produce
ways of pixels interpretation. All image file formats should be picture of the various organs in the human body. Metadata
analyzed which includes information of pixels, analysis of portion presents in post operation file format explain the
medical data and metadata. Then the various file formats are depth of pixel data. Modality produced images are different
given. At the end, limitations and advantages of the file from the post operational images which comprises different
formats are deliberated, with few contemplations about the contents. Metadata are a powerful tool to comment and make
upcoming instructions in the area of medical image file full use of information which is related for medical and
formats. investigation purposes and to arrange and fetch into archives
images and relevant data. Third is about the Pixel
Fundamental Ideas Information. Pixel Information is the numerical values which
A medical image is a portrayal of the inward design or can be stored in the medical images. It can be either integers
capacity of an anatomic locale. It has a variety of picture or floating point numbers. It is used to represent by the bytes
components called pixels (2 dimensional) or voxels (3 which is the form of numerical values. CT, PET images are
dimensional). The quantity of pixels used to portray the field having 16 bits to represent each pixel information and it is in
of-perspective on a specific securing methodology is an the form of integer. The characteristics of each file formats
outflow of the detail with which the life structures or capacity are different. Each can be represented in the form of either

IJISRT21JUN1188 www.ijisrt.com 1615

Volume 6, Issue 6, June – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
signed or unsigned integer. Integer data type is suitable for provides the data about the properties of image such as image
front end but it is not sufficient for the post-processing. size, rows and column size, attributes and pixel resolution.
Floating point is the best way of representing naturally and The header provides the patient’s medical information and
also it is suitable for post-processing. As floating point also basic information when DICOM images opens in the
consists of real and imaginary parts, by passing is easy even folder. Patient information can be identified easily by
for the complex data type. Two bytes storage is not sufficient verifying the images which is coded. But still the header
when the pixel information is stored. Problem arises in the information be lost sometimes if the file format is converted
LSB and MSB of the pixel information when representing in to any other formats like PNG, GIFF etc. Sometimes the
2 bytes. This problem arises with respect to the processor. header data will be lost usually for the purpose of
Pixel information starts at the standard location and anonymization. Some few software’s can achieve this
sometimes in the variable location which depends upon the anonymization. PACS stands for Picture Archiving and
header length. In either way, the information can be stored in Communications System. It is a database where all the
uncompressed format. Compression can be categorized into medical images can be saved. The server of Picture Archiving
two techniques. First is lossy and another one is lossless. and Communications System should have enough space to
Lossless compression is the greatest advantage with respect to store high quality DICOM files as it tends to consume more
the storage. Benefit of lossy technique is the information loss space. PACS server is owned by each and every hospital.
cost. It's anything but clear under which conditions the Hospital has the setup to store each and every medical image
perusing of the pictures or potentially the quantitative post- automatically in its PACS server. Before storing the images,
handling systems are not affected by information loss. Then file converts to various formats such as Joint Photographic
again, the appropriation of lossy techniques with a low or Experts Group, Portable Graphics Format, Tagged Image File
moderate loss of information instead of lossless ones can be Format, Graphics Interchange Format. Both lossless and lossy
considered. Recovering exact data which is original and compression techniques can be supported by Joint
recovering complete data from the compressed data is a Photographic Experts Group. Compression specification can
tedious task which can be allowed by very few compression be applied and thereby the amount of lost data can be
forms. Referring this technique as “lossless” or “reversible” controlled. Basic Joint Photographic Experts Group supports
compression schemes. Take a consumer application as an lossy compression. Lossless compression can be supported
example. While decompressing text documents, character by next version of Joint Photographic Experts Group. Images
changes is not acceptable irrespective of frequency of can be partitioned as ROI with respect to specific areas.
occurrence. Applications of medical images may expect the Lossless compression takes only that particular area as an
exact restoration of the input and hence lossless compression input whereas the lossy compression takes remaining parts as
is widely used. Lossy or irreversible compression occurs an input. Image sharing between the system is easy in case of
when the output which is compressed are not similar to the JPEG. If the DICOM is in the form of JPEG, it is easy to
original, but still the loss amount and loss type is adequate for upload in the websites and also helps in the PowerPoint
some application. Lossy compression is widely used in pages presentations. On the other hand, publications of paper is not
of web browser and also continuously applied the colorful be supported with this type of images. Both lossless and lossy
images which is extracted from digital cameras. compression techniques can be supported by Tagged
Image File Format. Various journals support Tagged
II. FILE FORMATS Image File Format as it produces the images of higher
quality. But it is not apt for the PowerPoint presentations and
Medical image file formats can be categorized into two also for the websites as the size of the file is very large.
formats. Standardization of the images is in the format of Lossless compression type is used by Portable Graphics
DICOM where the images can be generated by modalities. To Format. Clarity and Brightness which is very important
improve the post- processing, the Nifti can be used. Medical features in an image can be controlled by Portable Graphics
images are always stored in two ways. First way is using meta Format. As the header components of DICOM file, metadata
data and image data which can be stored in one single file. concept used in images in PNG which provides greater
Second way is storing the two in different files. benefits. Various journals support PNG as it produces the
images of higher quality and also it suits for the PowerPoint
A. DICOM presentations and also for the various websites. Mostly this
DICOM stands for Digital Imaging and format is not in use today as it is the oldest imaging formats.
Communications in Medicine which is used to share the It leads to improper compression and also it has very minimal
medical images, which is used to view the medical images, features. This format is mostly used in websites and it also
store the medical images and finally used to retrieve the supports lossless type of compression.
medical images. It is widely accepted format. It follows
certain rules to manage the correct information conforms to B. Nifti
set protocols to maintain accuracy of information transmitted Nifti (Neuroimaging Informatics Technology Initiative)
through medical images. Each and every DICOM medical is a data format for the storage of Functional Magnetic
image has two components. First one is header and the Resonance Imaging (fMRI) and other medical images. The
second is the original image itself. First component has the Nifti format is adapted from Analyze, developed by
data which explains the image contains important information Biomedical Imaging Resource (BIR) at Mayo Clinic. Nifti
about patient. It contains all the patients basic and necessary and Analyze are still compatible – Nifti simply adds
information such as their name, sex, birth date and age. It also additional fields. There are two variants of the Nifti format:

IJISRT21JUN1188 www.ijisrt.com 1616

Volume 6, Issue 6, June – 2021 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
NIfTI-1 and NIfTI-2. NIfTI-2 improves the data types globally especially in the field of image processing and the
supported by NIfTI-1, as well as precision and voxel size; binding of information and results within a medical image
however, the two formats are not binary compatible. related to the clinical context is a big concern.
Currently only NIfTI-1 is supported by the Nifti reader. Nifti
Files can be extended to include additional information, such REFERENCES
as attribution. Nifti formats can be read/write using reader
and writer. Nifti Reader – Reads one or more raster slices [1]. Michele Larobina & Loredana Murino, Medical Image
from one or more volumes in a single Nifti file. Nifti Writer - File Formats, Society for Imaging Informatics in
Writes a series of raster slices into one or more volumes into Medicine 2013 , DOI 10.1007/s10278-013-9657-9 , 13
a single Nifti file. December 2013.
[2]. NIFTI documentation. Http://nifti.nimh.nih.gov/nifti-
C. Other file formats 1/documentation (accessed in June 2013)
Various other file formats with inhibit compression [3]. Tsalafoutas IA, Metallidis SI: A method for calculating
techniques also available which is following particular the dose length product from CT DICOM images. Br J
standard. Various applications and Graphical system are Radiol 84(999):236– 43, 2011.
using the other file formats. Those formats are available in [4]. Channin DS, Mongkolwat P, Kleper V, Rubin DL: The
Group Portable Graphics Format, Tagged Image File Format, Annotation and Image Mark-up project. Radiology
Graphics Interchange Format. Due to insufficient support of 253(3):590–2, 2009.
compression techniques and restrictions in bit depth, it is not [5]. Patel V, Dinov ID, Van Horn JD, Thompson PM, Toga
used much for medical imaging. Text file and data files such AW: LONI MiND: metadata in NIfTI for DWI.
as ZIP can be used in the images with the compression Neuroimage 51(2):665–76, 2010
techniques. Other file formats are not efficient since the [6]. Andrew J. Hanson. Visualizing Quaternions. Morgan
techniques which used not provides the benefits of structure Kaufmann Pub, 2006
of image. [7]. Brett M, Johnsrude IS, Owen AM: The problem of
functional localization in the human brain. Nat Rev
III. ELABORATION Neurosci 3(3):243–9, 2002
[8]. Jenkinson M, Beckmann CF, Behrens TE: Woolrich
CT scanners and MRI scanners which falls under digital MW, and Smith SM. FSL. NeuroImage 62(2):782–790,
imaging modalities stores various medical images. Fetching 2012.
those medical images is a tough task. File need to be [9]. https://docs.safe.com/fme/html/FME_Desktop_Docume
encrypted and decrypted when it is available. Each and every ntation/FME_R eadersWriters/nifti/nifti.htm
file contains information of four types. First is unmodified [10]. Centre for evidence-based purchasing, A beginner's
image data which is also compressed. Second is the basic guide to virtual private networks in a Picture Archiving
personal information about the patient. Third is related to the and Communication System environment: CEP 05094;
examination of patient’s health. Direct method is retrieving NHS PASA March 2006 [online] (2006).
the information of the patient’s medical data. In PACS area, [11]. http://www.cep.dh.gov.uk/cepproducts/Catalogue.aspx
data usage is distributed which deals with elaborative [12]. Clunie, d., dicom, http://www.dclunie.com/
information and also can retrieve the abstract details in order [13]. Dreyer, k.j., mehta, a., thrall, j.h., pacs: A guide to the
to merge the images into three-dimensional data. This process digital revolution. Springer, Place, Published.(2002)
is more tedious, and it needs in-depth understanding of the [14]. Integrating the healthcare enterprise, www.IHE.net
way the files are built. The three different formats which are [15]. Health level 7, www.HL7.org
very popular are fixed, block and tag based. The layout of [16]. Huang, h.k., pacs: Basic Principles and Applications,
each file is same which is referred as fixed format. Header 2nd edn, John Wiley & sons, inc. (2004)
present in the file points to the information is referred as
block format. File item individually contains the own
information of its length. Out of all three, very popular format
is the second one which is block format. The beginning part
of the header has very few pointers for each block having
larger size. Almost all the blocks are in present in the same
location and at fixed length.

IV. CONCLUSION

There is a tremendous growth in the field of image file

formats. DICOM is the recommended standard in the image
encoding which can created by diagnostic modalities. It
provided the greater advantages in the medical field and also
leads to the development of technology in the advancement of
medical imaging. In the future, there can be evolvement of
new file formats which is useful in post-processing. The
importance of image specific information is increasing

IJISRT21JUN1188 www.ijisrt.com 1617