Scribd
Upload
14 views

Deep Learning

Uploaded by

sharmaswastik048
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
14 views

Deep Learning

Uploaded by

sharmaswastik048
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 19

Project Progress Presentation

On
Kidney Stone Detection and Classification from
CT Scanned Images Using Deep Learning Techniques

Presented By GUIDED BY
. SARFARAJ ALI (205240)
Mr. P.C YADAV
ANKITA SINGH (2152067 )
POORNIMA KUMARI (2152071)

DEPARTMENT OF ELECTRONICS ENGINEERING


FACULTY OF ENGINEERING AND TECHNOLOGY
(Uma Nath Singh Institute of Engineering and Technology)
VEER BAHADUR SINGH PURVANCHAL UNIVERSITY JAUNPUR (U.P)
CONTENTS

Introduction
Project Concept
Review of Related Work
Project Objectives
Proposed of design Methodology
Project Progress
Application
References and Bibliography
Introduction
 Kidney stone disease is one of the risks for life
throughout the world a n d the majority of people with
the stone formation in the kidney at the initial stage
do not notice it a s disease a n d it d a m a g e s the
organ slowly.
 The most common symptom of kidney stone is a pain
in the upper back. There c a n be blood in the urine
a n d also urinary tract infection.
 The purposed work is used to detect a n d classify the
kidney stone using CT images.
 The Bilinear Interpolation a n d the median filter is used
for image resizing a n d noise removal in pre-processing.
Detection of Kidney stone in the Computed Tomographic
( C T ) images using Level set Segmentation a n d few
more algorithms, implemented using Colab Software
Project Concept
 In this presentation, we explore the opportunities and
potential of deep learning within the realm of CT
scans, with a particular focus on segmentation and
classification.
Review to Related Works

S.NO RESEARCH AUTHORS DATASET METHODOLOGY RESULTS


PAPER
1. Kidney Stone
Detection and
Genemo, M. (2023) metadata, Real-time dataset 3D-CNN Accuracy: 98.2%, Sensitivity:
94.1%, Specificity: 97.4%
classification based on
deep Learning
Approach

2. Deep learning
computer vision
Kristian M. Black, Hei
Law†, Ali Aldoukhi, Jia
Louis C. Herring and Co.,
Orlando, FL
(CNN), ResNet-101 Accuracy: 94%, Sensitivity:
97.83%, Specificity: 97.83%
algorithm for Deng† and Khurshid R.
detecting kidney stone Ghani(CNN), ResNet-
composition 101

3. A deep learning
system for automated
Daniel C. Elton1 Evrim
B. Turkbey1 Perry J.
NNMC-CTC from University of 3D U-Net33 Accuracy: 92.2%, Sensitivity:
95%, Specificity: 88%
kidney stone detection Pickhardt2 Ronald M. Wisconsin Medical Center
and volumetric Summers1
segmentation on
noncontrast CT scans
Project Objectives

Objective(1) Collection of Images


This involves gathering a dataset of images that contain
kidney stones or related features for analysis and training
purposes.
Objective(2) Preprocessing of data images
Before feeding the images into the segmentation algorithms,
preprocessing steps like noise reduction, normalization, resizing,
and contrast enhancement may be necessary to improve the
quality of the data.
Objective(3) Classification
The trained CNN model can then classify the detected
kidney stones based on their characteristics, providing
detailed information on stone type, size, and location. Objective(4) Performance Evaluation
This data can inform personalized treatment strategies Evaluate the performance of the selected image segmentation
method in detecting kidney stones. This involves quantitatively
analyzing the accuracy, sensitivity, specificity, and other relevant
metrics of the segmentation results compared to ground truth or
expert annotations.
Proposed of Design Methodology

Image Feature Feature


Kidney Image Classification
Preprocessing Exraction selection
Convolutional Neural Networks(CNN) for CT
Scan Analysis
Image Preprocessing 11
Deep learning models for kidney stone
detection often begin with image
preprocessing, where the CT scan images Feature Extraction
are standardized, normalized, and to 2 Convolutional neural networks (CNNs)
isolate the relevant anatomical structures.
are particularly well-suited for medical
image analysis, as they can automatically
extract relevant visual features from the
CT scans, such as stone size, shape, and
Classification and 3 density.
Quantification
The trained CNN model can then classify
the detected kidney stones based on their
characteristics, providing detailed
information on stone type, size, and
location. This data can inform
personalized treatment strategies.
Collection of Data Images of CT Scanned
Flowchart
Classification Algorithm and Confusion Matrix
Training and Validaton Accuracy

The result of training model with validation revealed in Figures a and and detailed in Table
2. Figure
B shows loss per epoch (entire image dataset) training.
It can be seen from the graph that when number of epoch increases, loss decreases. This
indicates that with more training, the model will be less susceptible
to losses.
Figure a shows that with increasing number epoch of training, accuracy increases.
Training of model validation Loss
The study considered loss per epoch (entire image dataset) for the training.
This is presented in Figures .
It can be seen from the graph pattern that number of epoch increases while loss decreases.
This indicates that with more training, the model will be less susceptible to losses.
The progress of training with increase in number of epoch is shown in graph.
The research further presents elapased time and increasing batch number.
This revealed the relationship between loss and training accuracy.
Normal Images
Cyst Images
Project Progress
Applications

4.Research and
Education

5.Quality Assurance
in Pathology

6.Clinical Decision
Support etc.
References and Bibliography

1.Genemo, M. (2023). Kidney Stone Detection and Classification Based On


Deep Learning Approach. (DOI: 10.59287/ijanser.2023.7.4.545) Volume 7, pp.
38-42, 4 2023

2. Kristian M. Black, Hei Law Ali Aldoukhi, Jia Deng† and Khurshid R. Ghani.
Deep learning computer vision algorithm for detecting kidney stone
composition. | doi:10.1111/bju.15035

3. Yingpu Cui, Zhaonan Sun, Shuai Ma, Weipeng Liu, Xiangpeng Wang,
Xiaodong Zhang, Xiaoying Wang . Automatic Detection and Scoring of Kidney
Stones on Noncontrast CT Images Using S.T.O.N.E. Nephrolithometry:
Combined Deep Learning and Thresholding Methods DOI: 10.1007/s11307-
020-01554-0

4. Mol Imaging Biol (2021) 23:436Y445 DOI: 10.1007/s11307-020-01554-0


World Molecular Imaging Society, 2 Deep Segmentation Networks for
Segmenting Kidneys.

You might also like