Targeted Therapies and AI for Retinal Conditions - Uses and Limitations

Introduction:

Targeted therapies aim to utilize the body's immune system to identify cancer-causing cells and destroy them. Artificial intelligence has introduced various techniques to diagnose various eye diseases. The AI technology-based diagnosis provides more accurate results than other clinical diagnostic methods.

What Is Targeted Therapy?

It is an emerging therapy that limits systemic toxicity and enhances tumor treatment. The concept of enabling the body's own immune mechanism to detect cancerous cells and destroy them led to the development of targeted cell therapy. The use of target therapies for diseases of the eyes cause immune and inflammatory complications. The neural and vascular networks of the eye make it more susceptible to side effects.

How AI Is Used in Diagnosing Diabetic Retinopathy?

• Diabetic retinopathy is a common complication of diabetes causing damage to blood vessels of the eyes resulting in mild to moderate vision problems. Severe cases of diabetic retinopathy can lead to blindness. Early screening, timely referral, and treatment for diabetic retinopathy is the accepted way to reduce vision problems.

• Tele Retinal Screening - AI Technology used to diagnose the early stage of diabetic retinopathy. AI Technology uses Deep Learning (DL) systems and various algorithms that help in the accurate screening of diabetic retinopathy. This AI technology and algorithms are utilized for fundus screening and photography. Following the AI test results, the patient referral is done for further clinical evaluation by an ophthalmologist.

• Point-of-Care AI-DR Software - The United States food and drug administration (FDA) approved the autonomous diabetic retinopathy and diabetic macular edema software. This software helps in the screening of more than mild retinopathy in adults with diabetes. This screening is done by taking a fundus photographic examination.

• Eye Art - Another AI-based Technology for diabetic retinopathy screening. This test also received FD approval in 2020.

• ‘Mydriatic’ Smartphone-Based AI Technologies - The world's smartphone retinal imaging system that received FDA approval. These types of out-of-clinic device screen retinopathy provide a benefit for those people who do not have access to routine eye examinations.

What Are the Challenges in Ai-Based Diagnosis?

• Difficulty in grading the severity of diabetic retinopathy and early stages of pre-diabetic retinopathy.

• Misdiagnosis when retinopathy is associated with other conditions

How Artificial Intelligence Is Used for the Diagnosis of Age-Related Macular Degeneration?

• Age-related macular degeneration (AMD) is an eye disorder resulting in reduced vision common in older people. It occurs due to the breaking down of the macula, a part of the retina that helps in perceiving a clear image. Degeneration develops in both eyes at the same time. Degeneration increases over time and decreases the central vision, in which reading, driving, and recognizing faces will become difficult. Early detection and treatment can delay vision loss.

• For diagnosis of age-related macular degeneration, AI-based algorithms are used in fundus photography; these photographs are carefully curated with good quality. The rational photographs captured from real-world clinical settings are compromised in quality when compared with those generated with AI Technology. AI Technology is also used to access the quality and field view of the captured retinal photographs. This method has been shown to have excellent results.

• Currently, retinography is considered to be an important diagnostic method for AMD in clinics. AI-based algorithms can be used as a helpful adjunct for diagnosis and screening.

• Recently developed DARC detection of apoptotic retinal cells is a retinal imaging technique when accompanied by an AI-aided algorithm that helps in the detection of stressed and damaged retinal cells.

How Is AI Technology Used in the Diagnosis of Glaucoma?

• Glaucoma is a term used to address disorders associated with optic nerve damage, commonly leading to blindness. In glaucoma, fluid buildup in the front part of the eye causes increased pressure on the eye. This rise in Intraocular pressure causes damage to the optic nerve. Other degenerative changes are the loss of ganglion cells and thinning of the retinal nerve fiber layer. It is a common related disease affecting a large population. Timely diagnosis and proper treatment can slow down the disease progression, and vision is improved.

• AI technology for glaucoma screening should be able to evaluate intraocular pressure, optic disc, gonioscopy, and visual field. The AI technology measures intraocular pressure through a contact lens-based continuous monitoring device. Netra AI is an AI tour that evaluates glaucomatous fundus photographs.

• The major drawback is the difficulty in detecting glaucoma in high myopic eyes. A combination approach in diagnosis of glaucoma helps in a more accurate classification of patients. The next level for AI in glaucoma diagnosis will be based on lifestyle behavior, medical history and other optimal logical parameters.

How Is AI Technology Used in Retinopathy of Prematurity (ROP)?

• For a newborn baby, the normal blood vessels grow in the last weeks before a baby is born. When a baby is born prematurely the blood vessels are not fully developed and they grow into other parts of the eye. These grown blood vessels leave scar tissue formation inside the eye. This car tissue causes retinal detachment and loss of vision. This is called retinopathy of prematurity (ROP), the leading cause of childhood blindness in the world.

• AI tools provide retinal image analysis, advanced imaging and informatics in retinopathy of prematurity (ROP). Fully automated ROP screening helps in the overall classification and the need for treatment is predicted.

• Computer-based image analysis (CBIA) provided a more valid method to diagnose ROP. AI has also helped in the development of the ROP severity score. Helps in proper disease monitoring, risk assessment, treatment failure, and recurrence. When compared with human evaluation, CBIA saves a lot of time and effort with more accurate results. In the future, the introduction of CBIA systems in daily clinical practice helps to impact the results of patients with ROP and other changes of the retina.

What Are the Limitations Associated With AI Treatment?

• Complexity and time effectiveness while practicing AI technologies.

• The possibility of AI misdiagnosis is still in the study.

• Legal liabilities arising from incorrect AI diagnosis.

• Errors in diagnosis when a condition is associated with comorbidity.

• Clinical decisions cannot be finalized only by AI-based diagnosis.

Conclusion

The AI tools help clinician diagnose better and easily, with the most accurate results. It also helps with timely corrective treatment conditions. Also, doctors can attend to more people within a limited time, and these emerging healthcare facilities are reducing the number of visits to the doctor. Various researchers are ongoing to make use of AI technology for screening more diseases of the eye.