Descripción del proyecto
According to the WHO, prostate cancer (PCa) is the second most common cause of cancer worldwide. In Europe specifically, PCa is a significant public health concern since it is the most commonly diagnosed non-cutaneous cancer and the second cause of death in men. About 1,300,000 citizens of the European Union are estimated to have had a prostate cancer diagnosis in the last five years. The average estimated number of deaths from PCa was of 38 deaths per 100 000 male inhabitants.There are several tests to diagnose PCa with different levels of accuracy: blood-based prostate-specific antigen (PSA) tests, digital rectal examination (DRE), and blind biopsy were the most common. They are, however, often inaccurate, inconclusive and most importantly fail to assess accurately the malignancy of the tumour, i.e., the risk of metastasis. In the last years MRI has been proven to have a clear benefit if performed before the biopsy (prebiopsy MRI) by reducing the number of cilinders per biopsy, increasing the sensitivity and avoiding unnecessary biopsies. This assessment is critical in PCa because this particular cancer can remain indolent for years, whereas the median survival for non-operable patients with metastatic PCa is only 24 months. In PCa therefore, there is a great risk of overtreatment (aggressive biopsies, removing the entire prostate of an indolent tumour, chemotherapy, etc.) with all the burden and irreparable damage caused to patients, but also undertreatment of tumours that end up becoming rapidly fatal. Today, with the tests available, the discrimination between indolent and aggressive PCa is still problematic, resulting in inaccurate risk stratification. There is therefore a great need for new tools capable of predicting accurately the outcome of the disease to enable physicians to make the best clinical decisions for the benefit of patients, preventing both overtreatment and undertreatment.The aim of this project is to develop a computational tool to create a 4D digital twin of the entire prostate of a patient. Novel AI-based magnetic resonance imaging segmentation algorithms will be applied to extract not only patient-specific prostate anatomy (transitional zone, peripheral zone, seminal vesicles and neurovascular bundle) but also to detect PCa. The digital twin will incorporate in silico models considering the behavior of cells and tissues, to predict the effects of different types of oncological treatments not only on the tumor but also on the entire prostate, as well as to predict the efficacy of these treatments and the possible evolution of the disease. To this end, it will be necessary to compare the results of pathology and their corresponding MRI scans, as well as to analyze all the information available on the treatments applied in real PCa patients along the patient journey, in order to establish the key parameters for the design of the 4D digital twin.