Agile Bio-inspired Architectures (ABA) is a research project that focuses on medical image recognition of colorectal cancer using the Deep Learning process.
The project aims to improve the recognition of the structure of colorectal cancer samples so that better stratification of patients can be made for more appropriate treatment decisions. Combining 3D medical imaging with an AI application leading to rapid and robust cancer classification, individualized therapy recommendations will be generated for the treating physicians.
In this project, the GWDG is involved in providing the data storage and the medical image recognition. In addition, the GWDG is responsible for the development of new models. Deep Learning will be used to process this data. One challenge is that only very few or not enough data sets are available. That is why new algorithms are being worked on at the same time that are specifically suited for small data sets.