27 November 2017

Deciphex has filed two provisional patent applications directed to the field of abnormality detection.

Patholytix finds hyperplasia in rat liver

Deciphex has established priority filing dates in the US on two inventions we believe will be critical to the future of Deep Learning in Pathology.

The applications describe novel screening methods for histopathology tissues, that leverage deep learning approaches to detect abnormality in an unsupervised fashion. The techniques are organ and disease agnostic and are highly scalable.

“With the myriad of species, organs and pathologies routinely assessed by pathologists we believe that supervised training of deep learning processes is not scalable for pathology”, said Mark Gregson, CTO, “With the large amount of normal biopsies reviewed in certain applications, we believe there is fundamental value in detection of abnormalities in these applications.”

The company is now well placed to complement the existing deep learning approaches used in the industry to deliver high performing artificial intelligence systems that can be leveraged effectively in pathology.

Deciphex, an Enterprise Ireland High Potential Startup, based in INVENT DCU, Dublin, Ireland, plan to revolutionise pathology through automation of visual examination of tissue. Founded by former Danaher executives Dr Donal O’Shea and Mark Gregson, our mission is to optimise pathology workflows by helping pathologists focus on critical content.