Semantic Technologies and Smart Data Systems

digital-workplace-wearable-sensors-information-automation

Projects | Current Projects | Deep Learning for Medical Image Analysis

Deep Learning for Medical Image Analysis

Computer Vision for supporting clinical decision making

needs-and-functionality-icon

Needs and functionalities

Medical Imaging plays a crucial role in the detection and diagnosis of disease, and in the assessment and decision of the appropriate treatment, but also in the preclinical research and clinical trials required to create new therapies. However, the amount of information in medical images such as X-Ray, CT, MRI, Ultrasound and Histopathology exceeds the capabilities of human experts to process everything that may be relevant in a reasonable time. Computer solutions can augment the human as a second reader, looking for the tiniest signs that could be easily overlooked. They can analyse, quantify images and therefore support decision-making and can be used to correlate image characteristics with disease outcomes.

used-technology

Used technologies

We use deep learning and other machine learning approaches to process images: classification, semantic segmentation, object detection.

We develop techniques to visualize the results from deep neural networks and to make them robust against the variations inherent in medical images.

Konica Minolta Laboratory Europe Solution

Our solutions in the field of digital pathology allow classification of regions in whole-slide tissue images (gigapixel-size images) including cell detection and classification.

By computer-analysis, the image assessment becomes objective and quantified.

Deep-Learning-for-Medical-image-Analysis
application

Applications

Our solutions in digital pathology facilitate the workflow of clinical pathologists as well as preclinical researchers. We automate the cumbersome work and facilitate the pathological assessment through quantification and analysis. For X-Ray analysis, applications include disease detection, supporting the physicians as a second reader, or prioritizing image reading based on the analyzed image content.

more-info

More info

X