Portuguese develop artificial intelligence-based tool for automatic diagnosis of oncological pathologies
11th August 2020
Innovative solution allows taking a considerable leap in the way the anatomopathological diagnosis of histological samples is currently carried out. With the CADPath.AI project, the diagnostic process may become, starting in 2022, digital, introducing the algorithms as a complement to the work of anatomopathologists in the identification of anomalies.
In the medical specialty of Pathological Anatomy, oncological disease and its diagnosis are, increasingly, a central concern. Knowing that a timely and accurate diagnosis is an essential tool to fight cancer, anatomopathologists will have an important tool based on artificial intelligence from 2022 on. The technology is being developed by the CADPath.AI - Computer Aided Diagnosis in Pathology project consortium.
The high complexity that the oncological diagnosis presents, often recommends the need to obtain a second opinion. It is here that technology becomes an essential ally of doctors. In fact, in the recent years we have seen a proliferation of scanners for digitizing histological slides shows, a new instrument at the service of the so-called digital pathology.
In addition to making it possible to carry out a network diagnosis, aided by other technological tools, digital pathology will allow reducing the time that the pathologist spends on microscopic observation. “Tasks such as, for example, the identification of tumor cells, the counting of mitotic cells, or the identification of invasive growth, as well as their measurement, can now be performed through the use of artificial intelligence”, explains Jaime Cardoso, researcher at INESC TEC and professor at FEUP.
With the possibility of digitizing the histological slides, there was a great scientific interest in the development of image analysis algorithms, which can complement and make the role of pathologists more efficient, particularly in the diagnosis of colorectal and cervical cancers.
“We remain focused on the continuous improvement of the diagnosis, but we intend to go further and provide the market with an automatic diagnosis tool for oncological pathologies; a database, including the digitized slides and respective notes, clinical history and diagnosis; and a platform for generating scientific knowledge, ”says Ana Monteiro, project manager at IMP Diagnostics.
CADPath.AI, an IMP Diagnostics project, in a technological partnership with INESC TEC and the company Leica Biosystems, has a financing of around 70% through the COMPETE2020 program in a total investment of 1M €.