Performance of ChatGPT in Ophthalmic Registration and Clinical Diagnosis: Cross-Sectional Study (Preprint)

Abstract

<sec> <title>BACKGROUND</title> Artificial intelligence (AI) chatbots such as ChatGPT are expected to impact vision health care significantly. Their potential to optimize the consultation process and diagnostic capabilities across range of ophthalmic subspecialties have yet to be fully explored. </sec> <sec> <title>OBJECTIVE</title> This study aims to investigate the performance of AI chatbots in recommending ophthalmic outpatient registration and diagnosing eye diseases within clinical case profiles. </sec> <sec> <title>METHODS</title> This cross-sectional study used clinical cases from &lt;i&gt;Chinese Standardized Resident Training–Ophthalmology (2nd Edition)&lt;/i&gt;. For each case, 2 profiles were created: patient with history (Hx) and patient with history and examination (Hx+Ex). These profiles served as independent queries for GPT-3.5 and GPT-4.0 (accessed from March 5 to 18, 2024). Similarly, 3 ophthalmic residents were posed the same profiles in a questionnaire format. The accuracy of recommending ophthalmic subspecialty registration was primarily evaluated using Hx profiles. The accuracy of the top-ranked diagnosis and the accuracy of the diagnosis within the top 3 suggestions (do-not-miss diagnosis) were assessed using Hx+Ex profiles. The gold standard for judgment was the published, official diagnosis. Characteristics of incorrect diagnoses by ChatGPT were also analyzed. </sec> <sec> <title>RESULTS</title> A total of 208 clinical profiles from 12 ophthalmic subspecialties were analyzed (104 Hx and 104 Hx+Ex profiles). For Hx profiles, GPT-3.5, GPT-4.0, and residents showed comparable accuracy in registration suggestions (66/104, 63.5%; 81/104, 77.9%; and 72/104, 69.2%, respectively; &lt;i&gt;P&lt;/i&gt;=.07), with &lt;i&gt;ocular trauma&lt;/i&gt;, &lt;i&gt;retinal diseases&lt;/i&gt;, and &lt;i&gt;strabismus and amblyopia&lt;/i&gt; achieving the top 3 accuracies. For Hx+Ex profiles, both GPT-4.0 and residents demonstrated higher diagnostic accuracy than GPT-3.5 (62/104, 59.6% and 63/104, 60.6% vs 41/104, 39.4%; &lt;i&gt;P&lt;/i&gt;=.003 and &lt;i&gt;P&lt;/i&gt;=.001, respectively). Accuracy for do-not-miss diagnoses also improved (79/104, 76% and 68/104, 65.4% vs 51/104, 49%; &lt;i&gt;P&lt;/i&gt;&amp;lt;.001 and &lt;i&gt;P&lt;/i&gt;=.02, respectively). The highest diagnostic accuracies were observed in &lt;i&gt;glaucoma&lt;/i&gt;; &lt;i&gt;lens diseases&lt;/i&gt;; and &lt;i&gt;eyelid, lacrimal, and orbital diseases&lt;/i&gt;. GPT-4.0 recorded fewer incorrect top-3 diagnoses (25/42, 60% vs 53/63, 84%; &lt;i&gt;P&lt;/i&gt;=.005) and more partially correct diagnoses (21/42, 50% vs 7/63 11%; &lt;i&gt;P&lt;/i&gt;&amp;lt;.001) than GPT-3.5, while GPT-3.5 had more completely incorrect (27/63, 43% vs 7/42, 17%; &lt;i&gt;P&lt;/i&gt;=.005) and less precise diagnoses (22/63, 35% vs 5/42, 12%; &lt;i&gt;P&lt;/i&gt;=.009). </sec> <sec> <title>CONCLUSIONS</title> GPT-3.5 and GPT-4.0 showed intermediate performance in recommending ophthalmic subspecialties for registration. While GPT-3.5 underperformed, GPT-4.0 approached and numerically surpassed residents in differential diagnosis. AI chatbots show promise in facilitating ophthalmic patient registration. However, their integration into diagnostic decision-making requires more validation. </sec>

Ähnliche Arbeiten