The diagnosis of Alzheimer’s disease (AD) involves the comprehensive assessment of cognitive symptoms, functional abilities, and neurological signs by specialized physicians. On the other hand, the diagnosis of its underlying neurobiology and pathology requires neuroimaging and biomarkers, which may provide evidence on the accumulation of amyloid-beta and tau proteins within the brain. Given the complexity of AD diagnosis, it is easy to imagine and predict that there is a real possibility that newly developed Artificial Intelligence methods, particularly the Chat-Generative Pre-Trained Transformer (ChatGPT), could aid in this complex process. In the ever-evolving landscape of artificial intelligence, ChatGPT stands as a remarkable achievement in the ongoing quest to enhance human-machine communication. It operates on the principles of deep learning, employing an extensive neural network model with the capacity to comprehend and produce text with a nuanced understanding of context, tone, and intent.
ChatGPT is increasingly finding applications in various areas of human activities, medicine included. In fact, there is a growing body of research that emphasizes the utility of ChatGPT in various medical contexts, such as patient education, interaction, and the dissemination of healthcare information. However, with its use in the medical field, some controversy is also arising. For instance, there are reports of inconsistency in the diagnostic response to a disease when the same set of information is used and analyzed by ChatGPT.
At times, these types of results could be ascribed to the different versions of ChatGPT employed in different studies. However, this is not always the case. Nevertheless, this aspect in the diagnostic response to ChatGPT is without a doubt a timely and pressing concern that needs some further consideration. I strongly believe that physicians and other healthcare providers, patients, caregivers, and even healthy adults who are utilizing ChatGPT as a diagnostic tool for AD should be very cautious and not consider ChatGPT as a definitive and fully reliable diagnostic aid. At least for the moment. Here are some examples that could better explain the current state of the art when it comes to ChatGPT and clinical diagnoses.
In one study, ChatGPT, although showing accuracy in identifying the myths associated with AD, was not 100% aligned with the clinician opinion on the cases. Another paper demonstrated that the responses provided by ChatGPT in the context of hepatocellular carcinoma surveillance and diagnosis are inaccurate and unreliable. In summary, if we take into consideration all the available evidence, we should exercise great caution in the utility of ChatGPT as a reliable diagnostic tool in any stage of AD. Importantly, we should not forget that artificial intelligence technologies like ChatGPT are somewhat still at their early stages of development. However, I do believe that future development and advancements in the field and the implementation of more appropriate data training will ultimately improve the capability of ChatGPT to diagnose AD and other health conditions accurately and consistently.
Domenico Praticò, MD, holds the position of the Scott Richards North Star Charitable Foundation Chair for Alzheimer’s Research and serves as a Professor and the Director at the Alzheimer’s Center at Temple, as well as a Professor of Pharmacology at Lewis Katz School of Medicine at Temple University.
For more information on the research conducted by Dr. Domenico Pratico, please visit this link.
Connect with Dr. Domenico Pratico through LinkedIn, Facebook, Twitter, Medium.
Stay updated with the work happening at Dr. Domenico Pratico's lab by visiting the Pratico Lab website.
Comments