Schiffer team
LARGE-SCALE ARTIFICIAL INTELLIGENCE FOR BRAIN MAPPING
Christian Schiffer - Team leader
UNCOVERING MICROSTRUCTURAL BRAIN ORGANIZATION WITH THE HELP OF AI
Facing the challenges of brain complexity and big data analytics, research at INM-1 addresses human brain organization on its different scales, hand in hand with brain simulation and the investigation of learning in artificial and biological neural networks to advance our understanding of cognition, which opens perspectives to discover novel mechanisms for AI.
The research group is embedded into the Big Data Analytics group headed by Timo Dickscheid, which develops methods for high throughput analysis and 3D reconstruction of microscopic image data, and neuroinformatics solutions for high-resolution brain atlases. In this context, our Helmholtz AI young investigator group focuses on the development of artificial intelligence methods for large-scale microstructural mapping of the human brain. Our work aims at data-driven holistic characterization of human brain architecture at the microstructural scale, based on petabyte-scale microscopic datasets reflecting cyto-, fiber-, and chemoarchitecture as complementary organizational principles. Driven by continuously increasing image resolutions and data volumes, our methods are designed for large-scale distributed operation on High-Performance Computing (HPC) systems.
Research lines
- Medical Imaging with Deep Learning: We will continue our research directions to develop fully-automated, high accurate solutions that save export labor and efforts, and mitigate the challenges in medical imaging, i.e. i) the availability of a few annotated data, ii) low inter-/intra-observers agreement, iii) high-class imbalance, iv) inter-/intra-scanners variability and v) domain shift. Our research portfolio can be categorized into Learn to Recognize, Adapt, Learn, Reason and Explain, incorporate prior knowledge.
- Federated Learning in Healthcare: We will focus our research on developing innovative deep Federated Learning algorithms that can distill and share the knowledge among AI agents in a robust and privacy-preserved fashion. Research topics include, but not limited to, i) handling distributed DL models with data heterogeneity including non i.i.d, and domain shifts, ii) developing explainability and quality control tools for distributed models, and iii) robustness to poisoning models.
- Affordable AI and Healthcare: In addition, we are also interested in developing affordable AI solutions suitable for poor-quality data generated by low infrastructure and point-of-care diagnosis.
