A new artificial intelligence technology promoted by the Hospital Clínic of Barcelona will facilitate the diagnosis of prostate cancer based on magnetic resonance imaging, which will have an impact on the treatment of each case. The instrument, easy to use, is available to all professionals.
The early and precise diagnosis is essential to increase survival rates in a type of cancer such as prostate cancer that will present 30,000 new cases in Spain and represents 15% of all tumors diagnosed, according to calculations by the Spanish Society of Medical Oncology.
Starting from this high incidence, the Hospital Clínic is going to tender four more than the prostate tumors fossin incorporated into the European project Incisive, financed by the EU, to create AI services to support the diagnosis of lung and colorectal cancer i from mom.
“The prostate, which causes incidence, mortality and social repercussions is the breast cancer of the home, for the moment it will be included in the Incisive project at our request because we see that otherwise it is producing gender discrimination” , explained Antonio Alcaraz, head of the Clínic’s urology service.
Four minutes later, the instrument is in operation. A repository of federal data has been created for more than 3.7 million medical images, as well as for more than 9,000 patients. AI acts on this basis of data to enhance the analysis of magnetic resonance images related to the diagnosis, prediction and monitoring of prostate cancer.
The developed algorithms focus on patient prioritization, lesion localization and segmentation, cancer diagnosis and staging, and the risk of metastasis.
Access to the information is universal and very easy to use: after registration, the professional sends an MRI image to the system and sends a report indicating whether the patient develops cancer. “This report explains the different modules that the AI has processed and the probability that the patient has a tumor, including whether it is a significant or non-significant tumor, and reports on the aggressiveness of the tumor,” explains the researcher. Lourdes Mengual.
The procedure for diagnosing the tumor involves performing an MRI if it exists – normally detected by PSA levels using blood analysis – and a biopsy on the identified lesions. “But the resonance is not reliable cent per cent. “There are dubious images, images that are not visible because of the smallest measure,” says Carles Nicolau, head of radiology at the Clínic. “This AI technique helps to detect lesions that are not visible to humans, from many images that have a definitive diagnosis.”
Although there are no studies on the effectiveness of the system, in a personal estimate Antonio Alcaraz considers that the diagnosis will be around 20%. According to what he says, it “democratizes patient care at all levels of the world” and is especially useful for radiologists that do not address a specific type but rather a diversity of tumors.
“We are going to start four years ago, when many people thought that AI was ‘fer volar coloms’ (a false expectation). Ara has arrived to stay, it will be a change at least as revolutionary as the Internet will be today. Medicine will change radically, and will arrive at surgery”, emphasizes Alcaraz, head of the Idibaps research group in genetics and urological tumors.
The Incisive project has brought together 26 partners (academics, industrialists, hospital centers) from 9 European countries. In addition to the Clínic, the TIC Social Salut foundation, Barcelona Supercomputing Center and Medtronic Ibérica have participated in Spain, and a financing of 10 million euros has been awarded.
The repository system of millions of medical images and clinical associates that has billed Incisive is incorporated into a larger community project that is in the market, Eucaim (European Initiative for Obtaining Oncological Images), which seeks to promote innovation. and the deployment of digital technologies in cancer treatment and care to achieve more accurate and faster clinical decisions, diagnoses, treatments and predictive medicine.
#pioneer #facilitate #diagnosis #prostate #cancer
