Optical coherence tomography based diabetic – ophthalmic disease classification and measurement using bilateral filter and transfer learning approach
Keywords:Optical Coherence Tomography (OCT), AlexNet, SVM, ophthalmic diseases, transfer learning, bilateral filter, Michelson type fiber optic interferometer, tomographic image measurement
Optical Coherence Tomography (OCT) is a smooth application of low coherence interferometer with high air resolution and highly sensitive heterodyne detection technology to tomographic image measurement of living organisms. Currently, clinical applications are becoming more widespread in ophthalmology, cardiovascular system, dermatology, and dentistry. The problem with OCT is that the measurement area is as narrow as a few mm compared to other tomographic image measurement techniques, and it was initially applied to ophthalmology. Since then, various researches and developments have been carried out to expand clinical applications. Michelson type fiber optic interferometer is used for image acquisition. This paper presents a classification of ophthalmic diseases caused by diabetes. Bilateral filter is used for image pre-processing and noise removal. A transfer learning approach is implemented which uses AlexNet and Support vector machine (SVM) to classify the images. The AlexNet model is used to extract the features form the images and these features are classified using SVM model. The novelty of the proposed model lies in the use of image denoising using bilateral filter and then classification of the AlexNet features using SVM thereby achieving better classification accuracy with less training data. This also leads to better resource utilization. The ailments under study are Choroidal Neovascularization (CNV), Diabetic Macular Edema (DME), DRUSEN, and NORMAL. The proposed approach produced a higher classification accuracy of 99 % when compared to other deep learning algorithms like CNN, AlexNet and GoogleNet. The precision, sensitivity and specificity are recorded as 0.98, 0.99, and 0.99 respectively.
Copyright (c) 2023 ACTA IMEKO
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).