Soufiane Ben Haddou


Soufiane Ben Haddou completed his Bachelor’s degree in Artificial Intelligence in 2021 at the University of Amsterdam, followed by a Master’s degree in the same field in 2023 at the same university. During his master’s thesis, he focused on mesh reconstruction of the left ventricle from LGE segmentations using a transformer variational autoencoder. He examined how well the latent representation captured the geometry by measuring meaningfulness and representativeness through predictions on the latents and generating new samples respectively. He was supervised by Dr. Erik Bekkers from AMLab and Dr. Fleur Tjong from AUMC.Now embarking on a PhD, Soufiane is working on the prediction of Dilated Cardiomyopathy (DCM), with the goal of prematurely predicting DCM using multimodal solutions. These solutions incorporate various modalities such as MRI, ECHO, ECG, genetic information, and clinical data. His work is under the supervision of Prof. Ivana Isgum, a Professor in AI and Medical Imaging at the University of Amsterdam, and Prof. Connie Bezzina, a professor specializing in the Genetics of Cardiac Disorders.His involvement is with the QurAI group at the University of Amsterdam and Amsterdam UMC. DCM, the focus of his research, is a heart muscle disorder characterized by the thinning and stretching of the heart muscle in the left ventricle, complicating blood supply to the body. The project, ‘DCM-NEXT’, aims to develop new models through multimodal deep learning and computer vision for earlier diagnosis and better treatment of DCM, utilizing clinical, imaging (CT, MRI, ECHO), ECG, and genetic data. The project’s ambition is to integrate these models to harness the predictive value of each modality.