Modeling and Artificial Intelligence in Ophthalmology

Artificial intelligence in practice: measuring its medical accuracy in oculoplastics consultations

Adam J. Neuhouser — 2024-05-10

Purpose: The aim of this study was to investigate the medical accuracy of responses produced by Chat Generative Pretrained Transformer 4 (Chat GPT-4) and DALLE-2 in relation to common questions encountered during oculoplastic consultations.

Methods: The 5 most frequently discussed oculoplastic procedures on social media were selected for evaluation using Chat GPT-4 and DALLE-2. Questions were formulated from common patient concerns and inputted into Chat GPT-4, and responses were assessed on a 3-point scale. For procedure imagery, descriptions were submitted to DALLE-2, and the resulted images were graded for anatomical and surgical accuracy. Grading was completed by 5 oculoplastic surgeons through a 110-question survey.

Results: Overall, 87.3% of Chat GPT-4’s responses achieved a score of 2 or 3 points, denoting a good to high level of accuracy. Across all procedures, questions about pain, bruising, procedure risk, and adverse events garnered high scores. Conversely, responses regarding specific case scenarios, procedure longevity, and procedure
definitions were less accurate. Images produced by DALLE-2-were notably subpar, often failing to accurately depict surgical outcomes and realistic details.

Conclusions: Chat GPT-4 demonstrated a creditable level of accuracy in addressing common oculoplastic procedure concerns. However, its limitations in handling case-based scenarios suggests that it is best suited as a supplementary source of information rather than a primary diagnostic or consultative tool. The current state of medical imagery generated by means of artificial intelligence lacks anatomical accuracy. Significant technological advancements are necessary before such imagery can complement oculoplastic consultations effectively.

Efficient semianalytical investigation of a fractional model describing human cornea shape

Marwan Abukhaled — 2024-05-10

Purpose: This study presents a novel application of the semianalytical residual power series method to investigate a one-dimensional fractional anisotropic curvature equation describing the human cornea, the outermost layer of the eye. The fractional boundary value problem, involving the fractional derivative of curvature, poses challenges that conventional methods struggle to address.

Methods: The analytical results are obtained by utilizing the simple and efficient residual power series method. The proposed method is accessible to researchers in all medical fields and is extendable to various models in disease spread and control.

Results: The derived solution is a crucial outcome of this study. Through the application of the proposed method to the corneal shape model, an explicit formula for the curvature profile is obtained. To validate the solution, direct comparisons are made with numerical solutions for the integer case and other analytical solutions available in the literature for the fractional case.

Conclusion: Our findings highlight the potential of the proposed method to significantly contribute to the diagnosis and treatment of various ophthalmological conditions.

Using mathematics to avoid blindness in diabetics (Part 2): eliminating re-emergent diabetic retinopathy caused by blood thinners

Arieh Helfgott — 2024-02-13

Purpose: To report the clinical experiences of author AH, who calculated that modest stepwise lowering of arterial blood pressure can reverse (i) re-emergent diabetic retinopathy (DR) caused by antiplatelet and anticoagulant agents, even in the presence of continued use of the latter necessary agents, or (ii) DR induced by common or severe hypertension and so, (iii) simultaneously treat both of AH’s vascular and ocular medical conditions.

Methods: In instances of DR and visual impairment with evidence of exudate formation, blood pressure adjustments were applied, based on mathematical models of retinal exudate production developed by one of the authors (AH). Specifically, the model was used to calculate a critical arterial blood pressure below which retinal exudate formation should cease. Antihypertensive agents were then increased gradually until the desired lower target blood pressure was achieved and DR eliminated. Optical coherence tomography (OCT) was used to test for therapeutic effectiveness.

Results: In four different clinical situations, which included blood thinners or hypertension, control of retinal exudate formation and elimination of re-emergent DR was achieved solely by blood pressure lowering and confirmed (with OCT) by return, to normal, of retinal measurements and vision.

Conclusion: While the evidence presented here is derived from clinical examples in one person and not from a statistically justified large study, this approach to the control of retinal exudate formation offers very effective unintrusive management of a common vision-threatening aspect of DR. In particular, this approach avoids laser treatments and the challenging experience of commonly administered intraocular injections. Clinical and mathematical evidence is presented that treatment with abundant vitamin B1 (300 mg) and vitamin D results in partial cure of DR. A cure to DR has not been reported before.

Future perspectives: The reversal of DR and potentially age-related macular degeneration (ARMD), with safe and simple measures, is an incredibly worthy management goal for these two very common conditions. The possibility demands urgent evaluation with what should be zero- or low-risk clinical trials.