Goal: The INTELCOR project will develop the best non-invasive colorectal cancer (CRC) diagnosis, early detection, differential diagnosis (with benign adenoma), and progression prediction tests. The last two tests will be developed for the first time. To be considered the best, the tests should be the most accurate (95%-100%), robust, and transparent. The tests are based on circulating microRNA (non-invasive) and advanced artificial intelligence methods (the highest robust accuracy). They will be also white-box or transparent models, discovered from microarray data, and their relationship with the hallmarks of cancer will be revealed, using knowledge mining in various knowledge bases.

These tests will take as inputs the levels of circulating microRNA biomarkers discovered by us, and will output a diagnosis – normal, CRC, or adenoma – or a prognosis – progressive or non-progressive CRC. To transform these molecular and cellular models into real clinical decision support (intelligent) systems, they will be embedded in an automated CRC clinical workflow, created for the first time. The tests will be also implemented as Software as a Service (SaaS) and in hardware (via FPGA), to further increase their commercial potential. This will be done for the first time for a medical test.

INTELCOR is a highly interdisciplinary project. Thus, UMF Cluj (P1) will enroll and investigate the CRC patients clinically, pathologically, imagistic, and will produce the corresponding microarray data. SAIA will develop a bioinformatics data analysis workflow and use it to develop the tests, which will be functionally analyzed together with P1. Rivel Market (P2) will develop and implement a top level CRC Electronic Medical Records for data integration, and the first automated CRC clinical workflow for tests integration into clinical practice. They will also implement the tests as software packages, as SaaS, and in hardware, for commercial reasons.

As validation on large cohort of patients is known to improve the credibility and the marketability of omics tests, more data will be acquired from specialized companies. All available data will be integrated by CO in a multicenter database and a meta-analysis will be performed. As in any project combining high throughput data with artificial intelligence, the number of microarray samples and the computational power are important, but costly. INTELCOR is carefully designed to satisfy the financial and time constraints and reach not only the best published results, but also the best CRC omics tests on the market, by properly combining the human resources and facilities of the 3 partners with the project budget.

Leave a Reply

Your email address will not be published. Required fields are marked *