MEDIA - Medical Image Analysis

About

Medical Image Analysis group was founded in 2015 with main focus on fundamentals of Machine Learning followed by Deep learning for medical imaging modalities such as histology, thoracic X-rays, X-ray with fractures, CTs, brain imaging. The most impactful works being V-Net CITE, Agg-Net CITE, Squeeze and Excite CITE.  Currently, the Medical Image Analysis (MedIA) group focuses on top-level research on advanced machine learning and deep learning algorithms aiming to attend medical-image-related problems. Research done by MedIA includes (but is not limited to) interpretability, geometric deep learning, multi-modal data analysis, semi, weakly, and self-supervised learning, meta-leaning, and federated learning. The group also organizes the Deep Learning for Medical Applications (DLMA) and the Graph Deep Learning for Medical Applications (GDLMA) seminars, together with the Machine Learning in Medical Imaging (MLMI) practical course.

Contact Person / Group Coordination

Azade Farshad - azade.farshad@tum.de

Research Partner

TU 
Klinikum Recht Der Isar
Imperial College London 
Nvidia healthcare 
Ulm University
Oxford (VGG group)
Helmholtz AI 
Johns Hopkins 
Harvard
Sharif 
Kyung Hee University Seong Tae Kim
Google

Research Grants and Awards

BigPicture
MCML: Meta-learning & Interpretability
Episodic Semantic Scene Analysis: with CV team
DIVA
COVID
DAAD
DAAD: Doctoral Programmes in Germany 
TUM-ICL JAD 2020

Project Members

Awards

  • Two MICCAI travel awards (Tariq, Mahsa) 2021
  • Best Reviewer award (Ashkan) 2021

Teaching

  • MLMI
  • DLMA
  • GDLMA

Extra- curricular:

  • GCN learning group
  • Scene graph group

Location

Campus Garching

Chair for Computer-Aided Medical Procedures and Augmented Reality,
Boltzmannstr. 3, 85748 Garching b. Munchen