This site exploits session and third parties cookies. Pressing on 'Accept' button, you accept cookie usage. For further information click Cookie Policy button

  • Banner

An AI Platform integrating imaging data and models, supporting precision care through prostate cancer’s continuum

Funded under: H2020-EU.3.1.5

Project reference: Grant Agreement n. 952159


Start date 1 October 2020     End date 30 September 2024

Keywords: Medical Imaging; Artificial and Computational Intelligence; Prostate cancer; Open image space; Trustworthy AI

Prostate cancer (PCa) is the second most frequent type of cancer in men and the third most lethal in Europe. Current clinical practices suffer from lack of precision, often leading to overdiagnosis and overtreatment of indolent tumours. This calls for advanced AI models to go beyond SoA by deciphering non-intuitive, high-level medical image patterns and increase performance in discriminating indolent from aggressive disease, early predicting recurrence and detecting metastases or predicting effectiveness of therapies. To date, efforts in the field are fragmented, based on single–institution, size-limited and vendor-specific datasets, thus making model generalizability impossible.

The ProCAncer-I project brings together 20 partners, including PCa centres of reference, world leaders in AI and innovative SMEs, with recognized expertise in their respective domains, working to design, develop and sustain a cloud based, secure European Image Infrastructure with tools and services for data handling. The platform will host the largest collection of PCa multi-parametric MRI, anonymized image data worldwide (>17,000 cases), in line with EU legislation through data donorship. Robust AI models will be developed, based on novel ensemble learning methodologies, leading to vendor-specific and vendor-neutral AI models for addressing eight PCa clinical scenarios.

To accelerate clinical translation of PCa AI models, the project will focus on improving the trust of the solutions with respect to safety, accuracy and reproducibility. Metrics to monitor model performance and inner causal relationships will shade lights on model outcomes, also informing decision makers on possible failures and errors. A roadmap for AI models certification will be defined, by interacting with regulatory authorities, thus contributing to a European regulatory roadmap for validating the effectiveness of AI-based models in clinical decision making.

ISTI-CNR’s role in the project is key as the team is involved in the development of robust AI models able to cope with the heterogeneity of imaging data and the biases and confounders this might introduce in the learning models. The team will lead the task related to AI trustworthiness, based on safety, transparency and reproducibility of results as well as on performance monitoring when used in clinical practice.