“Prediction or Proxy?” Reconsidering Deep Learning‐Based Models for Secondary Loss of Response in Crohn's Disease

Key Points

• The research questions the assumptions behind deep learning models in predicting secondary loss of response in Crohn's disease.
• Evaluated existing deep learning models used for prediction in clinical settings.
• Reviewed implications of these models on patient outcomes and treatment decisions.
• Found inconsistencies in the predictions made by deep learning models.
• Indicated a need for better validation and accuracy of these models in clinical practice.