Deep Learning in Dementia
Deep Learning for the Differential Diagnosis of Dementia from Multi-Modal Neuroimaging Data
Deep Neural Nets
DeepMentia: The Project
DeepMentia is an inter-disciplinary research project funded by BMBF for the differential diagnosis of dementia. The differentiation is challenging and requires the integration of multi-modal neuroimaging data such as magnetic resonance imaging (MRI) and positron emission tomography (PET), along with genetic and demographic information. We will build on latest advances in representation learning to design novel neural networks that can extract relevant information in multi-modal data, potentially yielding a leap in performance. We will work on differentiating between patients with Alzheimer’s disease, frontotemporal lobar degeneration, Lewy body disease, and vascular dementia, which are the four most common forms of dementia. To understand the decision-making process of the network, novel methods for network interpretability will be developed.
Magnetic Resonance Imaging
Positron Emission Tomography
Deep Neural Networks
Advance diagnostics of different forms of dementia
Improve our understanding of the decision-making process
Prof. Christian Wachinger
Prof. Wachinger coordinates the project and envisions to transform the healthcare system with intelligent algorithms. A computer scientist by training, he has years of experience in working at the intersection to medicine. Dr. Wachinger is an expert in machine / deep learning, medical image analysis, computer-aided diagnosis, and neuroimaging. He has received his training with leading researchers in the field of automated image analysis at TU Munich, MIT, Harvard, and ENST Paris, and is now leading the lab for Artificial Intelligence in Medical Imaging at LMU Munich.
Dr. med. Igor Yakushev
Dr. Yakushev is a clinician scientist, specialist in Nuclear Medicine, senior physician and Head of Neuroimaging at the Dept. of Nuclear Medicine, TUM. He is a Principle Investigator of the TUM Neuroimaging Center and an associate faculty within the Munich Center for Neurosciences. He leads the research group “Multimodal imaging of normal and pathological cognition” with the focus on molecular imaging of dementing disorders. Dr. Yakushev is well aware of the clinical perspective for differential diagnosis and knows the requirements for integrating automated systems in clinical workflows.
Lab for Artificial Intelligence in Medical Imaging
Department of Child and Adolescent Psychiatry
Ludwig Maximilian University
Department of Nuclear Medicine
Klinikum rechts der Isar
Technical University of Munich