Our research proposal pursues four specific objectives:
1.) The development of deep learning methods for diagnosis of PSVD based on histological image data, including multicentre blinded validation.
2.) Combination of image-based diagnosis prediction with clinical data to develop a multimodal deep learning system.
3.) Identification of features that are recognized as predictive for PSVD
4.) Finding novel clusters and previously unknown biological subgroups of PSVD based on reverse engineering approaches.