Watch time: 4 minutes
“There is growing evidence in our knowledge that machine learning is generally not highly capable of using less complex data to contribute something meaningful. Our results support that.”
The use of systems such as artificial intelligence (AI) has become a popular tool in neurology as they can incorporate deep learning and machine learning algorithms. As stroke continues to be a major public health concern and one of the leading causes of death worldwide, incorporating AI technology has the potential to improve care for these patients.
A study recently published in Journal of American Medical Association We compared the performance of stroke-specific algorithms with pooled cohort equations across different subgroups to assess the value of machine learning.1Michael Pencina, PhD, study lead author, and colleague, From the results, we note that predictive models and algorithmic techniques for assessing risk did not significantly improve the accuracy of identifying new-onset stroke by cohort studies.
in a recent interview with neurology live®, Pencina discussed the generalizability of results based on US cohort data. He shared how the findings compared to the study’s original hypothesis. discussed the potential for future research on inequality in health and how it relates to his findings.
