OPTOMETRY AND ARTIFICIAL INTELLIGENCE (AI)

 

BY EJIOFOR ESTHER CHIOMA (O.D) 

Artificial intelligence (AI) is defined as the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, or decision-making. It utilizes software to stimulate cognitive function such as learning and problem solving. The use of AI is being applied in the areas of sports, agriculture, administration, and even human relations. Therefore, healthcare, and as would be discussed specifically in this piece, Optometry, is surely no exception to the gains that could be derived from Artificial Intelligence. The ease, speed and accuracy that could be mined through exploration of solutions in AI are limitless.

AI algorithms have been described for image analysis in retinal diseases including diabetic retinopathy, agerelated macular degeneration, retinopathy of prematurity, retinal vascular occlusions, and retinal detachment; they have also been described for use in glaucoma, keratoconus, cataract, refractive error, intraocular lens power calculations, and when planning strabismus surgeries. At present, AI is being used to improve diagnostic and administrative aspects in Optometry such as: Optical Coherence Tomography (OCT), Visual Field Testing and even the Electronic Health Records (EHR) and I believe more fields in Optometry could be improved with the integration of AI. 

In OCT, qualitative data represented in images is quantified and this, through comparisons with normative parameters of different parts of the eye, has aided in sharp and accurate diagnosis, helping to single out the main troubling condition amongst differential diagnosis. Today, the most powerful image analysis methods are based on artificial intelligence (AI). Pathologies such as choroidal neovascularization, precise quantification of retinal perfusion, and reliable disease diagnosis can now be achieved through improved OCT imaging techniques using AI. 

Visual field testing has not been left out, as Artificial Intelligence has increased its diagnostic abilities, to the extent of more accurate retrogression analysis that helps in tracking the extent of visual field loss due to different conditions, including glaucoma. On the part of patient welfare and administration, Electronic Health Records, as aided by Artificial Intelligence has helped to achieve a very balanced approach to patients’ diagnosis by giving an objective view combined with the subjective view of the optometrist.. It has also improved the guaging of patient satisfaction, tracking, collation and extraction of patients’ data and even the determination of a patient’s diagnosis through analysis of signs, symptoms and history. All these are definitely advantages of AI in Optometry and so, Artificial Intelligence must be embraced in Optometry.

However, the involvement of AI in Optometry must not remain limited to clinic-based tests but rather, should be broadened to integrate artificial intelligence in more aspects of the profession such as development of self-testing applications on mobile devices.. I look forward to the development of phone and computer applications which will direct every day device users to keep phones or computer screens at an approximate distance of 40 centimetres to get an estimate of their visual acuity. This would help them know themselves that their vision is not optimal and would cause them to report to eye clinic. Apart from visual acuity, applications could also be developed for amsler grid testing for visual field estimation and color vision testing. At the end of the tests, it would be clearly stated that the results are nothing but a good estimate and I believe this will direct the patients to the clinic to get a doctor’s diagnosis after a comprehensive testing rather than push him or her to self-medicate. 

One of the goals of incorporating artificial intelligence in healthcare include the hope of achieving a better doctor-patient relationship in handling diagnosis, treatment and management and I believe AI is a means to that end. Patient management and vision therapy could be jointly managed by both patient and doctor with the use of applications that can be used from home on devices, for instance, in cases of binocular vision training. Even frame selection from home, including fit, size and colour, can be aided with the use of Artificial Intelligence. 

Contrary to the fears that AI would replace optometrists in the profession, I believe we can and should use it to our advantage. Courses linking machine learning and natural language processing with Optometry can be incorporated into School Curriculum probably as introductory at Undergraduate level, then more of a specialty at Postgraduate level. I believe strongly that artificial intelligence would not only improve but also, broaden the scope of Optometry and so should be explored.