Deep Learning in Medical Image Analysis
Imaging is a cornerstone of medicine. The number and volume of radiology exams is growing rapidly thereby tremendously increasing the workload of radiologists. Deep learning methods can potentially extract more information from images, more reliably, more accurately, and most notably fully automatically.
In the DLMedIA programme novel deep learning technology is developed that enables successful application to medical image analysis, for specific solutions for personalized and precision medicine.
In five interrelated projects, we address unique characteristics of medical image data that currently limit applicability of deep learning: high dimensionality of the data, limited availability of labeled data, and sensitivity to differences in acquisition protocols. We develop approaches to optimally involve human experts in the construction of deep learning systems.
Close collaboration with clinicians ensures access to large datasets needed for successful application of deep architectures, and clinical relevance of the solutions.
Involvement of outstanding companies in medical imaging ensures development of new products for radiology, pathology and ophthalmology