Two studies have come out regarding a new method that “significantly improves the reliability and accuracy of artificial intelligence (AI) for many applications.”
One study reports on “MIGHT (Multidimensional Informed Generalized Hypothesis Testing), an AI method that the researchers created to meet the high level of confidence needed for AI tools used in clinical decision making. To illustrate the benefits of MIGHT, they used it to develop a test for early cancer detection using circulating cell-free DNA (ccfDNA) —fragments of DNA circulating in the blood. A companion study found that ccfDNA fragmentation patterns used to detect cancer also appear in patients with autoimmune and vascular diseases.”
MIGHT is able to “fine-tune itself using real data and checks its accuracy on different subsets of the data…and can be applied to any field employing big data. … It is particularly effective for the analysis of biomedical datasets with many variables but relatively few patient samples.” It “consistently outperformed other AI methods in both sensitivity and consistency” in tests using patient data. A companion study also discovered that “ccfDNA fragmentation signatures previously believed to be specific to individuals with cancer also occur in patients with other diseases, including autoimmune conditions such as lupus, systemic sclerosis, and dermatomyositis, and vascular diseases like venous thromboembolism.”
Researchers also identified “eight key barriers to bringing AI into routine clinical care,” including “the false expectation that AI tools need to be flawless before they’re considered useful; the need to present results as probabilities rather than simple yes-or-no answers; [and] making sure AI predictions match real-world probabilities.”
