Abstract
The early diagnosis and detection of cataracts based on the image recognition of AI-driven systems are propounded in this chapter. The categorization puts cataracts in the bucket of avoidable blindness, and early detection is the key to prevent it. A cataract is undoubtedly a story for every home, and more information and education on eye health is always a scope for functional society. A minimum of 2.2 billion people suffer from vision impairments, and one billion could have been prevented. The chapter introduces itself to the significance of early detection and the challenges faced during manual diagnosis. The chapter then briefs the methodology, the collection build-up, and the method to pre-process a large dataset of labelled eye images. Image recognition techniques, feature extraction, and deep learning algorithms are availed to train a robust machine learning model. In healthcare technology, diagnosing disease through picture recognition is a significant contribution and is a revolution in medical history.
Top1. Introduction
All over the world, an eye condition namely Cataract has topped the list of prompts for blindness, the incidence of cataract is seen with aging changes, 3.9% in 55-64 years and 92.6% for 80 years or older than that (Lee & Afshari, 2017) (Liu, Wilkins, Kim et al, 2017). The year of 2010 marked the 10.8 million individuals with the blindness due to cataract (Khairallah et al., 2015) and it is also anticipated to alarm to 40 million by 2025 by ages and noteworthy life expectancy (Pascolini & Mariotti, 2012).
Cataracts degrade the vision quality and, along with vision, degrade the quality of life overall, burdening individual lives. The growing age is itself a challenge and is a transitory phase where there are physical, mental, emotional, and psychological changes one is going through. Cataract, being the aging changes, compound likelihood of dementia (Lee et al., 2022) menace of tumble, collision or road traffic accident (Foss, 2022) and distinctly affect the perception of the position of the life and in the long run leads to more fatality rate (Keay et al., 2022).
India accounts for 20.5% and China accounts for 20.9% according to the World Health Organization in the year 2010 visual impairment report. These two countries were the preponderance out of all in terms of blindness rate. India uplifted itself and made several remarkable furtherance. The WHO action plan for universal eye health 2014-2019 intent a curtailment in the prevalence of avoidable visual impairment by 25% till 2019 with the baseline of 2010. The WHO in the year 2010, had put a figure of blindness as 0.68% and visual impairment of 5.30% in India. The contemporary research shows a degradation of blindness by 47.1% which initially was in the range of 0.68% to 0.36% and for visual impairment by 51.9% which was initially in the range of 5.30% to 2.55%. The landmark to achieve the 25% reduction in visual impairment has been triumphed by India (Measures taken to control blindness cases, n.d.).
Various factors are enumerated for the immense weight of cataract blindness. The fact of enhanced life expectancy in the elderly age are associated with cataract nature. As the ICMR- UVR study indicates pervasiveness of ultraviolet radiation. The target to reduce the burden of cataract were affected during the phase of COVID-19 pandemic. There needs to be education, understanding in secluded place as well and workability of treatment available. The restrained opportunity of treatment facilities and huge discrepancy in the surgery of cataract state wise indicates the establishment of treatment for cataract set up (Measures taken to control blindness cases, n.d.).
Advancements in science and medicine have led humanity to a secure increase in life expectancy. As per the World Health Organization, doubling the population over 60 is expected by 2050 and will require radical societal changes. The diagnosis of cataracts needs to be revolutionized considering the cataract burden on public health; the conventional strategy to diagnose the cataract will, therefore, be a real problem in which the clinical examination will remain the key to the operative decision. The understanding of semiology and clinical forms of cataracts, along with the concept of technology and blending the knowledge of technology with the medical diagnosis will be fruitful and definitely will make the diagnosis easy, convenient, accessible, and affordable, and definitely will screen a large group of people in less period.
4.95 million People in the category of blind which is like 0.36% of total population, 35 million visually impaired and 0.24 million blind children are approximated in India (Wadhwani et al., 2020). The key factor to reduce the prevalence of blindness and visual impairment like as in case of cataract contributing mainly to its data are expeditious detection and treatment (Mannava et al., 2022). In India cataract being the constant in addition to refractive error remained the foremost prompt of blindness and visual impairment.
Key Terms in this Chapter
Cataract: Cataract is an aging change and its just like other body organ which show the sign of aging like greying of hair, loosening of skin or even low libido. Basically, the lens of the eye gets stiffened and show the sign of aging which ultimately reduces the vision.
Avoidable Blindness: Avoidable blindness is such kind of blindness which can be avoided from blindness if proper intervention is given to it. Cataract is one such example where person can go blind if intervention is not done, but if surgery is done for the same person and new place is placed then the blindness can be avoided.
Machine Learning: Important to artificial intelligence (AI) is machine learning (ML), which allows systems to learn from data automatically, without human intervention. Machine learning (ML) applies algorithms to data in order to help computers learn to recognize patterns, make decisions, and gradually become more efficient. Supervised, unsupervised, and reinforcement learning are three of the most common ML methods.
Image Recognition: It is used to describe the process of teaching computers to recognize objects in digital photos and movies using AI. It is just like how people see and recognize photos, AI systems can now detect and classify visual data based on objects, patterns, or characteristics.
Healthcare Technology: Technology which is being used in the healthcare. This world is revolving around the technology these days and healthcare is one of the core areas for the wellbeing of the society. The amendment between health system and technology will tackle various obstacles of health system. There are various research going on for this noble approach to ease the life where two different areas are coming together.
CNN: One form of deep learning architecture, Convolutional Neural Networks (CNNs), has significant success in picture identification tasks. Convolutional neural networks (CNNs) are able to automatically learn feature hierarchies, which helps them to identify intricate objects and patterns in images.
Deep Learning: A subfield of machine learning known as “deep learning” makes use of multi-layered neural networks. Deep learning, which draws inspiration from the anatomy of the human brain, has played a pivotal role in accomplishing outstanding achievements in domains including autonomous vehicles, natural language processing, picture and audio recognition, and more ( Khang & Hajimahmud, 2024 ).