I. Introduction
Deep Learning is a newly emerged technique resulting from the rapid and upward progress of artificial intelligence.
This is actually a method of machine learning, which consists of: feeding a machine with massive data, so that it can process and model it autonomously.
Ophthalmology, especially retin, is an area where it is serious; AI plays a prejudicial role given the use of a multitude of digital images.
The AI will thus succeed & agrave; detect abnormalities in the eyes and generate algorithms that identify the specifics of each pathological case.
II.Discussion
1.How can AI helps oculists?
Due to the difficulty; the interpretation of scanners by humans and the large number of documents to process on a daily basis, review times are getting longer.
The AI can make the same diagnoses as a doctor in a reduced analysis time. just a few seconds.
The solution recommended by Deep Learning will help the medical body to help the medical body. sort out patients requiring urgent treatment and therefore avoid any risk of the disease developing.
The AI thus offers the possibility; to speed up patient care. [1]
2.What is the use of AI in neuro-ophthalmology??
- Artificial intelligence enables us to analyze and manage data on diseases that affect millions of people, such as diabetes. Diabetic retinosis is difficult to achieve. Detect because in the initial stages affected people do not notice any changes in their vision. Artificial intelligence makes it possible, in this case of silent pathologies, to achieve early detection thanks to the analysis of the symptoms of patients with a history of diabetes. Software is developed in this context ; after taking a photo of the elect with a retinal camera, they allow doctors to check for the presence of any signs that may indicate an early stage of the disease. diabetic retinopathy. [2]
- New AI technologies, namely Deep Learning, allow precise and rapid interpretation of the appearance of the optic nerve, on simple digital images of the fundus of the eye
- These techniques allow the precise identification of papillary edema associated with intracranial hypertension
- The performance of these technologies is comparable & agrave; that of experts in neuro-ophthalmology, which confirms the usefulness; AI and its undeniable role in accelerating the diagnosis and therefore the care of patients in the field of ophthalmology. [3]
- In this section, we will be able to recall the role of AI in predicting the development of AMD: AMD or Age-Related Macular Degeneration is a pathology affecting the eye, and more precisely the retina. It results from the deterioration of the macula, which is the central area of the retina of the eye where visual acuity is greatest. This zone allows the vision of details and colors with precision. Macular degeneration is the consequence of aging leading to a progressive loss of central vision. AI has made it possible for AMD to assess the urgency of referring a patient to a retinal specialist based on the analysis of 3D OCT * images. In fact, several studies have shown that the AI model has eliminated errors made by physicians in estimating the degree of urgency. Clearly, then, AI will allow the interpretation of data that grows in volume with new imaging devices, and faster analysis for more subjects, for the benefit of patients. [4]
3.Current and future capacities
- Detect abnormalities: microaneurysm, hemorrhage, venous dilation, etc.
- Quantify these anomalies.
- Find and classify diseases.
- Guiding therapy.
- Detect repeat offenses.
- Establish prognostic factors
4.Examples of applications that use AI for ophthalmology
The growing use of smartphones and connected tablets among seniors makes it possible to consider personalized assessment of visual acuity (VA) and its monitoring at home. ForeseeHome is an application that can be launched at home and allows a simple test to be performed daily to check for minimal changes in vision. It is based on the principle of early detection of distortions. The monthly reports are sent directly to the doctor who is then alerted.
In the same vein, the Odysight app, by Tilak , was added. launched in 2019 for the follow-up of maculopathies: early detection of the evolution of the disease by generating an alert of drop in AV response to anticipate the response and therefore the thesis response Early detection of disease progression by generating a low AV alert. [5]
III.Conclusion
In the light of all the above, it could be said that AI techniques in the health field especially in the field of ophthalmology have resulted in the facilitation of the task of ophthalmologists and thus shorten the time required. for diagnosis. As a result, Deep Learning in ophthalmology will make it possible to sort out patients requiring rapid treatment or to detect pathologies which put the patient's life in danger.
Finally, remember that despite; help and utility What AI in ophthalmology provides could not replace ophthalmologists who would have to use all these technologies while keeping their surgical strategies, their empathies and especially their visions of the sets of problems of a patient.
So, we have to be very careful. the evolution of AI techniques in ophthalmology and ophthalmology their impacts on quality; of our medical care which should always respect the sequence of steps; the diagnosis and then the prognosis or the suggestion of treatments, these systems will certainly modify our care pathway.
Reference: [1] letter-neurologist / utility-artificial-intelligence-neuro-ophthalmology [2] artificial-intelligence-from-google-ophthalmology-service [3] / artificial-intelligence-futur -de-l-ophtalmologie / [4] J De Fauw JR Ledsam B Romera-Paredes Clinically applicable deep learning for diagnosis and referral in retinal disease. Nat Med 2018 (24) [5] Querques G, Querques L, et al. Preferential hyperacuity perimeter as a functional tool for monitoring exudative age-related macular degeneration in patients treated by intravitreal ranibizumab.
