The document proposes a 2-D compact variational mode decomposition (2-D-C-VMD) based method for automatically classifying glaucoma stages from fundus images. Preprocessed images are decomposed into variational modes using 2-D-C-VMD. Features are extracted from the variational modes and fed into a multiclass least-squares support vector machine classifier to classify images as healthy, early-stage glaucoma, or advanced glaucoma. The proposed method has advantages over existing empirical mode decomposition approaches such as better noise resistance and computational efficiency, allowing for more accurate early detection of glaucoma.
CCS335 _ Neural Networks and Deep Learning Laboratory_Lab Complete Record
DOC-20221128-WA0000..pptx
1. Suku Krishna K V
M.TECH S3 DOI
Guide-Mr. Kiran Babu(Assistant professor)
Reg.no-47321007
2-D Compact Variational Mode Decomposition Based Automatic
Classification of Glaucoma Stages From Fundus Images
2. Index
Abstract
Introduction
Literature review
Existing Method
Drawbacks
Proposed method
Advantages
Applications
Hardware and Software Requirement
Conclusion
References
3. Abstract:
Glaucoma is one of the leading causes of vision loss worldwide.
This problem can be reduced by the early and reliable diagnosis of glaucoma.
Here a method to classify the glaucoma stages (healthy, early-stage, and advanced stage) using a 2-D compact
variational mode decomposition (2-D-C-VMD) algorithm is done.
The preprocessed input images are first decomposed into several variational modes (VMs) employing 2-D-C-
VMD.
Finally, a trained multiclass least-squares-support vector machine (MC-LS-SVM) classifier has been utilized
for classification purpose.
4. Introduction:
Glaucoma is a group of eye diseases, which results in damage to the optic nerve. The main
risk factor is increased intraocular pressure (IOP) in the eye.
The disorders can be divided into two major groups, namely, primary open-angle glaucoma
(POAG) and primary angle closer glaucoma (PACG).
Glaucoma can permanently damage the vision of the affected eye because the effect of
glaucoma is gradually increased, which is difficult to identify earlier until the circumstance
is at a critical stage.
It is necessary to have regular eye examinations so that diagnosis can be made in its early
stages and treated appropriately.
5. Literature Review:
S. No Journal Type with year Authors Title Outcomes
1
IEEE Trans. Med. Imag., vol.
37, no. 11, pp. 2493–2501, Nov.
2018.
H. Fu et al.
Disc-aware ensemble
network for glaucoma
screening from fundus image
Studied about disc-
aware ensemble
network for
glaucoma screening.
2
in Proc. 1st IEEE Int. Conf.
Meas., Instrum., Control
Autom. (ICMICA),
Kurukshetra, India, Jun. 2020,
pp. 1–6.
D. Parashar and D.
Agrawal
Automated classification of
glaucoma using retinal
fundus images
Studied about
classification of
glaucoma
3
IEEE Trans. Inf. Technol.
Biomed., vol. 16, no. 1, pp. 80–
87, Jan. 2012
S. Dua, U. R. Acharya, P.
Chowriappa, and S. V. Sree
Wavelet-based energy
features for glaucomatous
image classification
Studied about
glaucomatous image
classification using
wavelet
6. Literature Review:
S.
No
Journal Type with year Authors Title Outcomes
4
Ophthalmology
Glaucoma, vol. 2, no. 1,
pp. 36–46, Jan. 2019.
A. T. Nguyen, D. S.
Greenfield, A. S.
Bhakta, J. Lee, and
W. J. Feuer
Detecting glaucoma progression using
guided progression analysis with OCT
and visual field assessment in eyes
classified by international classification
of disease severity codes
Studied about
glaucoma stages
progression
5
Med. Eng. Phys., vol. 34,
no. 2, pp. 129–139, Mar.
2012.
T.-C. Lim, S.
Chattopadhyay, and
U. R. Acharya
A survey and comparative study on the
instruments for glaucoma detection
It’s a comparative
study glaucoma
detection
6
IEEE J. Biomed. Health
Informat., vol. 21, no. 3,
pp. 803–813, May 2017.
S. Maheshwari, R. B.
Pachori, and U. R.
Acharya
Automated diagnosis of glaucoma using
empirical wavelet transform and
correntropy features extracted from
fundus images
Studied about
empirical wavelet
transform and other
feature extracting
methods for
diagnosing glaucoma.
7. Existing Method:
Automated diagnosis of glaucoma using digital fundus images based on empirical wavelet transform
(EWT).
The EWT is used to decompose the image, and correntropy features are obtained from decomposed EWT
components.
These extracted features are ranked based on t value feature selection algorithm.
Then, these features are used for the classification of normal and glaucoma images using least-squares
support vector machine (LS-SVM) classifier.
The LS-SVM is employed for classification with radial basis function, Morlet wavelet, and Mexican-hat
wavelet kernels.
8. Disadvantages in Existing Method:
EMD approach having drawbacks such as
• boundary distortion
• noise sensitivity (Sn)
• mode mixing
• the lack of mathematical proof.
10. Proposed Method:
• Support Vector Machine(SVM) is a supervised machine learning algorithm used for both classification
and regression.
• The goal of the SVM algorithm is to create the best line or decision boundary that can segregate n-
dimensional space into classes so that it will easily put the new data point in the correct category in the
future. This best decision boundary is called a hyperplane.
• Least-squares support-vector machines (LS-SVM) are least-squares versions of support-vector
machines (SVM).
• Here one finds the solution by solving a set of linear equations instead of a convex quadratic
programming (QP) problem for classical SVMs.
11. Proposed Method:
• Least-squares SVM classifiers were proposed by Johan Suykens and Joos Vandewalle. LS-SVMs are a
class of kernel-based learning methods.
• Green channel images have been extracted from the RBG image because it contains finer details for
down-streaming analysis.
• Further, Using contrast-limited histogram equalizations (CLAHE) to improve contrast and pixel
intensity. Then, 2-D-C-VMD has been used for ID.
• Then, various features are computed from the first variational mode (VM).
• Then, linear discriminant analysis (LDA) has been applied for the reduction of dimensionality.
• Afore, a trained multiclass least squares-support vector machine (MC-LS-SVM) classifier has been
utilized for the classification task.
12. Advantages of Proposed Method:
SVM Classifier
• SVM works relatively well when there is a clear margin of separation between classes.
• SVM is more effective in high dimensional spaces.
• SVM is effective in cases where the number of dimensions is greater than the number of samples.
• SVM is relatively memory efficient
• A prevalent and effective decomposition method is the variational mode decomposition (VMD).
• Compared to EMD decomposition, VMD has excellent noise resistance, better decomposing performance,
and stability and can be also utilized for feature extraction and fault diagnose
14. Hardware & Software Requirements:
Software: Matlab R2020a or above
Hardware:
Operating Systems:
• Windows 10
• Windows 7 Service Pack 1
• Windows Server 2019
• Windows Server 2016
Processors:
Minimum: Any Intel or AMD x86-64 processor
Recommended: Any Intel or AMD x86-64 processor with
four logical cores and AVX2 instruction set support
Disk:
Minimum: 2.9 GB of HDD space for MATLAB only, 5-8
GB for a typical installation
Recommended: An SSD is recommended A full installation
of all MathWorks products may take up to 29 GB of disk
space
RAM:
Minimum: 4 GB
Recommended: 8 GB
15. A newly introduced 2-D-C-VMD-based algorithm has been used for ID
It has various advantageous properties such as sharp boundaries, fully adaptive,
and non-recursive multiresolution technique.
VMD has been employed to decompose preprocessed fundus images into different
VMs.
It will have less computation complexity with better Ac and speed.
Effective for early and more accurate detection of glaucoma.
Conclusion
16. References:
[1] H. Fu et al., “Disc-aware ensemble network for glaucoma screening from fundus image,” IEEE Trans. Med.
Imag., vol. 37, no. 11, pp. 2493–2501, Nov. 2018.
[2] D. Parashar and D. Agrawal, “Automated classification of glaucoma using retinal fundus images,” in Proc.
1st IEEE Int. Conf. Meas., Instrum., Control Autom. (ICMICA), Kurukshetra, India, Jun. 2020, pp. 1–6.
[3] S. Dua, U. R. Acharya, P. Chowriappa, and S. V. Sree, “Wavelet-based energy features for glaucomatous
image classification,” IEEE Trans. Inf. Technol. Biomed., vol. 16, no. 1, pp. 80–87, Jan. 2012.
17. References:
[4] A. T. Nguyen, D. S. Greenfield, A. S. Bhakta, J. Lee, and W. J. Feuer, “Detecting glaucoma progression
using guided progression analysis with OCT and visual field assessment in eyes classified by international
classification of disease severity codes,” Ophthalmology Glaucoma, vol. 2, no. 1, pp. 36–46, Jan. 2019.
[5] T.-C. Lim, S. Chattopadhyay, and U. R. Acharya, “A survey and comparative study on the instruments for
glaucoma detection,” Med. Eng. Phys., vol. 34, no. 2, pp. 129–139, Mar. 2012.
[6] S. Maheshwari, R. B. Pachori, and U. R. Acharya, “Automated diagnosis of glaucoma using empirical
wavelet transform and correntropy features extracted from fundus images,” IEEE J. Biomed. Health Informat.,
vol. 21, no. 3, pp. 803–813, May 2017.