An Explainable Hybrid AI System for Early Detection of Chronic Diseases from Multimodal Medical Data

An Explainable Hybrid AI System for Early Detection of Chronic Diseases from Multimodal Medical Data

Chronic diseases such as diabetes, cardiovascular conditions, and chronic kidney disease represent a major burden on global healthcare systems due to their progressive nature and late-stage diagnosis. While AI has shown promise in early disease detection, many deep learning models operate as “black boxes,” lacking transparency and limiting clinical adoption. This doctoral project aims to develop an explainable hybrid AI system that combines the predictive power of deep learning with the interpretability of symbolic approaches (e.g., fuzzy logic, decision trees) to support early diagnosis and clinical decision-making. The system will process multimodal medical data, including medical imaging (e.g., CT scans, X-rays), laboratory test results, and data from wearable health sensors. Deep neural networks (e.g., CNNs, Transformers) will be trained to extract high-level patterns, while symbolic models will be used to generate interpretable diagnostic rules and explanations. Feature attribution methods and explainability modules (such as SHAP or LIME) will be embedded to visualize the contribution of each variable to a given prediction. The platform will also allow physicians to interact with and validate AI decisions through a transparent interface, increasing trust and usability in clinical environments. From a scientific perspective, the project will advance hybrid AI integration, medical decision explainability, and trustworthy AI design. Societally, it will contribute to improved early detection, personalized treatment, and preventive care, especially in resource-constrained settings. The project directly supports SDG 3 (Good Health and Well-Being) by aiming to reduce morbidity and mortality through timely, data-driven, and explainable interventions.

Candidates are expected to carry out their research full-time within the structures of EUROMED University.

The PhD candidate will develop an explainable hybrid AI framework for chronic disease prediction by combining deep learning and symbolic reasoning. They will manage multimodal data processing, train interpretable models using CNNs, Transformers, and tools like SHAP or LIME, validate them on medical datasets, and contribute to publications in medical AI and explainable systems.

The candidate should hold a Master’s in AI, Data Science, or Health Informatics, with experience in deep learning, data fusion, and hardware implementation. Strong Python skills, familiarity with TensorFlow/PyTorch, and interest in medical AI and model interpretability are required, along with analytical rigor, teamwork, and fluency in English (and preferably French).

CV, a cover letter, the PhD project, diplomas, and academic transcripts.

The application file must be sent to the Doctoral Studies Center (CEDoc) of the Euro-Mediterranean University of Fes by email no later than October 24, 2025, to the following contacts:

Administrative Affairs Officer of the CEDoc: Mrs. Boutaina Jai Mansouri : : b.jai-mansouri@emadu.ueuromed.org)

Director of Research and of the CEDoc: Prof. Abdelghafour Marfak : : a.marfak@euromed.org)

• Pr. Loubna OURABAH (l.ourabah@ueuromed.org)