Thursday, January 25, 2024
Researchers at City of Hope, a prominent cancer research and treatment institution in the United States, in collaboration with the Translational Genomics Research Institute (TGen), have successfully developed and tested an innovative machine-learning approach aimed at facilitating the early detection of cancer through smaller blood samples. The findings of this study were recently published in the journal Science Translational Medicine.
The significance of detecting cancer in its early stages cannot be overstated, as late-stage cancer diagnoses often result in higher mortality rates. The novel technology created by City of Hope, TGen, and their collaborators has the potential to bring us closer to a future where individuals undergo annual blood tests to identify cancer at earlier, more treatable stages, possibly leading to a cure.
The machine-learning system demonstrated an impressive capability to identify half of the cancers across 11 types that were studied. Remarkably, the test exhibited high accuracy, with a false positive rate of only 1 in every 100 samples tested. Notably, the majority of the cancer samples were sourced from individuals with early-stage disease, featuring few or no metastatic lesions at the time of diagnosis.
Rather than focusing on specific DNA mutations, the researchers, led by City of Hope in collaboration with colleagues from John Hopkins University, devised a novel method to detect variations in fragmentation patterns within repetitive regions of cancer and normal cell-free DNA (cfDNA).
The Center for Cancer Prevention and Early Detection at City of Hope is dedicated to generating crucial research findings and technologies centered around noninvasive blood tests and imaging. The goal is to detect cancers years before conventional diagnostic methods become effective.
This breakthrough in utilizing machine learning for cancer detection, particularly in its early stages, showcases the potential for advancing medical diagnostics. Although further validation and testing in larger, more diverse populations are necessary, the research paves the way for a more effective and accessible approach to cancer screening and detection.
Source: businesswire.com
