Can ChatGPT Vision Diagnose Melanoma? An Exploratory Diagnostic Accuracy Study - Supplemental material
Published: 12 January 2024| Version 1 | DOI: 10.17632/j2t89njrft.1
Contributors:
, , , Description
- Mendeley Supplemental Table 1. Raw study data Including ISIC image ID, Histopathologically Confirmed Diagnosis, Top-3 Differential Diagnoses by ChatGPT Plus, and corresponding Top-1 and Top-3 diagnoses evaluations (0=No, 1=Yes). - Mendeley Supplemental Figure 1. Representative examples of images from the dataset. - Mendeley Supplemental Figure 2. Examples of images of falsely diagnosed pigmented lesions by ChatGPT Vision.
Files
Steps to reproduce
1. Download the images from the ISIC archive (see table for ISIC archive reference numbers) 2. Upload the image to ChatGPT Vision 3. Ask the following question to ChatGPT Vision: What is the top 3 differential diagnosis?
Institutions
Hospital Clinic de Barcelona, Leids Universitair Medisch Centrum Geneeskunde, Antoni van Leeuwenhoek Nederlands Kanker Instituut
Categories
Artificial Intelligence, Dermatology, Artificial Intelligence Diagnostics, Melanoma, Nevus, ChatGPT