Cataract is a common eye ailment and therefore the cataract detection via Machine Learning could aid ophthalmologists to detect this age-related ocular disorder on time.
Pavani Tripathi, Yasmeena Akhter, Mahapara Khurshid, Aditya Lakra, Rohit Keshari, Mayank Vatsa and Richa Singh have discussed the possibility of cataract detection via Machine Learning in their research paper titled “MTCD: Cataract Detection via Near Infrared Eye Images” that forms the basis of the following text.
In the words of the researchers:
Globally, cataract is a common eye disease and one of the leading causes of blindness and vision impairment. The traditional process of detecting cataracts involves eye examination using a slit-lamp microscope or ophthalmoscope by an ophthalmologist, who checks for clouding of the normally clear lens of the eye. The lack of resources and unavailability of a sufficient number of experts pose a burden to the healthcare system throughout the world, and researchers are exploring the use of AI solutions for assisting the experts. Inspired by the progress in iris recognition, in this research, we present a novel algorithm for cataract detection using near-infrared eye images. The NIR cameras, which are popularly used in iris recognition, are of relatively low cost and easy to operate compared to ophthalmoscope setup for data capture. However, such NIR images have not been explored for cataract detection. We present deep learning-based eye segmentation and multitask network classification networks for cataract detection using NIR images as input. The proposed segmentation algorithm efficiently and effectively detects non-ideal eye boundaries and is cost-effective, and the classification network yields very high classification performance on the cataract dataset
Importance of this research: Cataract detection statistics
Cataract is one of the primary causes of blindness worldwide. In India, Cataract is responsible for 66.2% of blindness cases, 80.7% severe visual impairment cases, and 70.2% moderate visual impairment cases in the 50+ age group, according to the National Blindness and Visual Impairment Survey of India, 2015- 19.
The proposed solution, MTCD, helps detect Cataract Detection even in remote areas where professionals and resources might be unavailable. MTCD is presented as a low cost, accessible, and easy-to-use solution for cataract detection.
How MTCD Works
Here’s a simple step-by-step process explaining the functioning of the proposed method:
- Pupils are dilated with the help of eye drops.
- A NIR (Near-infrared) camera takes the eye image.
- Pyramidnet segments iris and pupil patterns from the image of the eye.
- Classification network accomplishes 2 tasks:
- Classifies the image as healthy or unhealthy
- Classifies the image as pre-cataract, post-cataract & others.
The research paper explains in detail the various steps involved in the process.
The researchers’ MTCD approach results were trained & tested with datasets such as IIITD Cataract Surgery Dataset and IIITD Alcohol dataset. The results of the proposed method were promising, according to the authors of the study.
A deep learning pipeline for Cataract Detection is proposed in the research paper. This technique would be specifically valuable in environments where expert availability is a constraint. Moreover, it also removes the subjectivity associated with the Ophthalmologist’s discretion. The researchers have concluded that
- Effective Cataract Detection is possible in the NIR domain.
- Even in challenging scenarios, the proposed segmentation algorithm was effective in detecting Iris & Pupil boundaries.
- The approach aids automated decision support system, and the overall Cataract Detection results obtained via this technique were encouraging.
The researchers have also made their findings and datasets public to spur further research in this area.
Source: Pavani Tripathi, Yasmeena Akhter, Mahapara Khurshid, Aditya Lakra, Rohit Keshari, Mayank Vatsa and Richa Singh’s “MTCD: Cataract Detection via Near Infrared Eye Images”