“We are very excited about our fast progress in deep learning over the past year and Visiopharm remains committed to investing a lot more in this technology, says Visiopharm CEO Michael Grunkin
Denmark (PRWEB UK) 3 May 2017
Real progress in this field requires a multi-disciplinary approach and therefore Visiopharm has established a broad multidisciplinary strategy to lead further development of Deep Learning in tissue pathology, within several important clinical and research applications. To support this strategy Visiopharm has established an International Consortium for Deep Learning in Tissue Pathology. The consortium involves Academic Medical Centers, Tissue Biobanks, Engineering Universities, Biopharmaceutical companies, and other industrial partners who all share this vision. “We are very excited about our fast progress in deep learning over the past year and Visiopharm remains committed to investing a lot more in this technology. We are currently formalizing the consortium and we look forward to continued collaborations and sharing more details on this initiative soon,” says Michael Grunkin, CEO of Visiopharm.
Visiopharm co-supervises several Masters Students and sponsor/co-supervises industrial PhDs in collaboration with its partners in the consortium, including the Technical University of Denmark and DTU Compute, a department we have worked closely with for many years. Early results of this effort have come from Masters Student, Jeppe Thagaard who is doing his Thesis work on Deep Learning. “We are very excited about these first results! Jeppe is only a few months into the project, and the Deep Learning algorithms developed on the Visiopharm software platform, were ranked 5th out of 23, with a marginal score difference to the winner of the CAMELYON17 competition,” says CTO Johan Doré. “We look forward to a continued collaboration with Jeppe after his Thesis work, and congratulate him on his focused efforts providing such impressive results after only a few months’ work,” continues CTO Johan Doré.
Visiopharm plans to make advantages of this new technology available to current users of the Oncotopix® products for cancer research and diagnostics, but also to include it in new product brands on their way to market.
The goal of the recent CAMELYON17 Deep Learning competition was to evaluate new and existing algorithms for automated detection and classification of breast cancer metastases in whole-slide images of histological lymph node sections, which are of high clinical relevance to pathologists. Full results and details of the CameLyon17 competition can be found at https://camelyon17.grand-challenge.org/results/.