Natural Sciences
Life Sciences
Scientific Computing
Scientific Computing

Sandy Engelhardt, Institute for Medical Informatics, Hochschule Mannheim

Marsilius Arkaden, Turm Süd, K13, Im Neuenheimer Feld 130.2

Heidelberger Kolloquium Medizinische Biometrie, Informatik und Epidemiologie

Cardiac imaging improves on diagnosis of cardiovascular diseases by providing images at high spatiotemporal resolution. Manual evaluation of these time-series, however, is expensive and prone to biased and non-reproducible outcomes. Convolutional neural networks (CNN) have recently shown outstanding performance in medical image processing. We present a method that addresses named limitations by integrating segmentation and disease classification into a fully automatic processing pipeline. We will reflect on our winning contribution for MICCAI's 2017 Automated Cardiac Diagnosis Challenge (STACOM). The fully automatic processing pipeline constitutes an attractive software for clinical decision support in adult and pediatric cardiology due to the visualization of segmentation maps, the comprehensive quantification of cardiologic assessment and the rapid processing speed. In the second half of the talk, we will focus on computer-assisted cardiac surgery applications. We will present novel visualizations of the mitral valve to improve planning of complex mitral valve repair surgeries. Beyond visualization, we present enhanced tools for surgical training and an unique training simulator concept based on 'Hyperrealism'.

Event data:
Import event data into Outlook Calendar