Aid System for the Detection of Various Pathologies Based on Image Analysis and Artificial Intelligence

CADIA

The CADIA project aims to create an image analysis system based on state-of-the-art artificial intelligence techniques (e.g. deep learning). As a clinical scenario for the piloting of this system, breast cancer screening assistance has been considered, developing algorithms for the analysis of lesions in mammography images and digital tomosynthesis, and breast cancer diagnosis, by means of innovative techniques to help the detection and characterization of tumor areas in digital pathology images.

The project includes retrospective images of mammography and digital tomosynthesis used in screening, and pathological anatomy samples from biopsies for the confirmation of suspicious lesions, in order to generate a database with annotated images that can be used to train algorithms of artificial intelligence. Likewise, the project will develop methodologies for the annotation of these images in a consistent manner, using tools developed for this purpose, which will allow Sergas specialists (radiologists and pathologists) to make the necessary annotations.

The objective is to develop advanced imaging algorithms based on deep learning that ultimately, once validated, allow reducing screening efforts, improving breast cancer detection, especially interval cancer, and improving diagnostic accuracy in digital pathology. , especially when working with full sample. In addition, the project will develop the necessary infrastructure for deploying algorithms in the Sergas that will be integrated with the multi-specialty platform for integrated diagnosis support - ADICAD del Sergas, so that, with the methodology developed, other artificial intelligence models can be developed to other imaging modalities, pathologies or applications. The demonstration of the systems will be carried out through the development of specific applications for specialists in the cases of radiology use (breast cancer screening) and digital pathology (breast cancer diagnosis).

The project is being developed by the UTE formed by the Vicomtech Foundation and the Inycom company. This Innovative Public Procurement project is part of the Code 100 health innovation plan executed within the framework of a collaboration agreement between the Galician Health Service and the Ministry of Science and Innovation, financed by 80% by the ERDF in the framework of the Multiregional Operational Program of Spain (POPE) 2014-2020.

Vicomtech

Parque Científico y Tecnológico de Gipuzkoa,
Paseo Mikeletegi 57,
20009 Donostia / San Sebastián (Spain)

+(34) 943 309 230

close overlay