Evaluating the Diagnostic and Treatment Recommendation Capabilities of GPT-4 Vision in Dermatology
Bibliographic record
Abstract
Abstract Background The integration of artificial intelligence (AI) in dermatology presents a promising frontier for enhancing diagnostic accuracy and treatment planning. However, general purpose AI models require rigorous evaluation before being applied to real-world medical cases. Objective This project specifically evaluates GPT-4V’s performance in accurately diagnosing and generating treatment plans for common dermatological conditions, comparing its assessment of textual versus image data and its performance with multimodal inputs. Beyond the immediate scope, this study contributes to the broader trajectory of integrating AI in healthcare, highlighting the limitations of these technologies, as well as their potential to enhance efficiency, and education within medical training and practice. Methods A dataset of 102 images representing nine common dermatological conditions was compiled from open-access websites. Fifty-four images were ultimately selected by two board-certified dermatologists as being representative and typical of the common conditions. Additionally, nine clinical scenarios corresponding to these conditions were developed. GPT-4V’s diagnostic capabilities were assessed in three setups: Image Prompt (image-based), Scenario Prompt (text-based), and Image and Scenario Prompt (combining both modalities). The model’s performance was evaluated based on diagnostic accuracy, differential diagnosis, and treatment recommendations. Results In the Image Prompt setup, GPT-4V correctly identified the primary diagnosis for 29 of 54 images. The Scenario Prompt setup showed a higher accuracy rate of 89% in identifying the primary diagnosis. The multimodal Image and Scenario Prompt setup also achieved an 89% accuracy rate. However, a notable bias towards textual data over visual data was observed. Treatment recommendations were evaluated by the same two dermatologists, using a modified Entrustment Scale, showing competent but not expert-level performance. Conclusion GPT-4V demonstrates promising capabilities in dermatological diagnosis and treatment recommendations, particularly in text-based scenarios. However, its performance in image-based diagnosis and integration of multimodal data highlights areas for improvement. The study underscores the potential of AI in augmenting dermatological practice, emphasizing the need for further development, and fine-tuning of such models to ensure their efficacy and reliability in clinical settings.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".