Imaging Biomarkers

Extracting biomarkers from medical images requires solid knowledge and know-how in algorithmic sciences applied to image processing.

MEDIAN has developed a strong scientific and technological expertise in this field as well as a large technology core that has been continuously enhanced and enriched over the last fifteen years. Expertise comes from an active internal research and from constant collaborations with worldwide scientific institutes involved in Image Processing and Medical Physics research. Collaborations with luminary institutes initiated in the early development of the company are pursued, including collaborations with major international research laboratories.

As of today, MEDIAN’s technological core is built from this expertise and is continuously evolving. MEDIAN’s core provides capabilities for an efficient management of medical images based on a wide set of algorithms for complex image analysis:

• Automated detection of lesions in images
• Organ and lesion delineation
• Automated and semi-automated organ and lesion segmentation
• Automated calculation of quantitative lesion parameters such as surface, volume, density and texture
• Non-rigid Image registration & Automated lesion pairing
• Artificial intelligence through data analysis and classifiers.

Algorithms are utilized to detect, quantify and follow up lesions or abnormalities identified in medical images. They are very well- adapted to the management of chronic diseases – including cancers but also other chronic and degenerative diseases such as Alzheimer diseases and cardiovascular pathologies-  for which images are compared over time to evaluate disease evolution and treatment efficacy.

Combination of those capabilities with MEDIAN’s technology platform enable the design, implementation and validation of advanced imaging biomarkers and of the associated diagnostic or monitoring tests with appropriate characteristics such as specificity, sensitivity, inter and intra user variability.