https://youvenz.github.io/tags/interpretability/Recent content in Interpretability on Rachid Youven ZeghlacheHugoen-usThu, 01 Jan 2026 00:00:00 +0000https://youvenz.github.io/research/explainable-ai-for-health/Thu, 01 Jan 2026 00:00:00 +0000https://youvenz.github.io/research/explainable-ai-for-health/<h2 id="overview">Overview</h2> <p>Deep learning models used in clinical settings must not only be accurate — they must be interpretable. Clinicians need to understand <em>why</em> a model predicts high-risk progression to trust and act on that prediction. Explainable AI (XAI) provides tools to open the black box and reveal which input features drive each decision.</p> <h2 id="techniques-i-apply">Techniques I apply</h2> <h3 id="saliency-maps">Saliency maps</h3> <p>Gradient-based saliency methods (GradCAM, Integrated Gradients, SHAP for images) highlight which spatial regions of a retinal image or OCT scan most influenced the model&rsquo;s prediction. A heatmap overlaid on a fundus image showing drusen, haemorrhages, or neovascularisation has direct clinical meaning.</p>