Time: 2024-12-14
research_worker have make significant promotion in the field of oncology by development a deep learning model that can accurately classify pancreatic ductal adenocarcinoma ( PDAC ), the most park form of Pancreatic cancer, into molecular subtypes exploitation histopathology image. This advanced approach, print in The American Journal of Pathology, offer a cost-effective and rapid option to current method that trust on expensive molecular assay. The survey clasp promise in advance personalize treatment scheme and better patient outcomes.
PDAC has become the one-third lead cause of cancer mortality in Canada and the United States, excel breast cancer. early detection is crucial, as surgery can remedy around one-fifth of PDAC case. However, the five-year survival rate remains at only 20%, with approximately 80% of patient already have metastatic disease at diagnosis. The aggressive nature of PDAC airs a significant challenge for sequence technology use to determine patient care plans.
The research team train deep learning Army_Intelligence model on whole-slide pathology image to identify the molecular subtypes of PDAC, achieve an accuracy of 96.19% in the TCGA dataset and 83.03% in a local cohort. This discovery let for the categorization of PDAC molecular subtypes establish on routine hematoxylin-eosin-stain slide, potentially revolutionize the clinical management of the disease.
In a offprint survey, research_worker at Champalimaud Research in Lisbon, Portugal have identify precursor to pancreatic cancer on routine MRI scan, exploitation diffusion tensor imagination technique. These findings could lead to a significant improvement in the early diagnosis and treatment of pancreatic cancer, ultimately enhance survival rates. pancreatic cancer is know to be one of the deadly cancer, with a five-year survival rate of only 44% if catch early, drop to 3% if it metastasize. The elusive nature of the disease often lead to late-phase diagnoses.
The team at Champalimaud Research concentrate on detection PanINs, precursor lesion that frequently develop into PDAC. Historically, detection PanINs noninvasively has been challenge, impede research into their biology and tumorigenesis in world. By development imagination method for PanIN diagnosis and word_picture, early diagnosis before PDAC is establish may become a world, supply hope for better result in pancreatic cancer patients.