Can I Find Blood Clots Using Artificial Intelligence? Scientists at Mount Sinai’s Icahn School of Medicine have demonstrated artificial intelligence a
Can I Find Blood Clots Using Artificial Intelligence? Scientists at Mount Sinai’s Icahn School of Medicine have demonstrated artificial intelligence algorithms that can detect signs of pulmonary embolisms in electrocardiograms (EKGs).[1]
Can I Use Artificial Intelligence To Find Blood Clots?
Pulmonary embolisms occur when blood clots form, break away, and clog the arteries in the lungs. These clots have the potential to be fatal or to cause long-term lung damage.
‘The fusion artificial intelligence model outperformed its parent algorithms and was also better than other tests at identifying specific pulmonary embolism cases.’
Although some patients may experience shortness of breath or chest pain, these symptoms may also indicate other issues unrelated to blood clots, making it difficult for doctors to properly diagnose and treat cases.
Furthermore, current official diagnoses are based on computed tomography pulmonary angiograms (CTPAs), which are time-consuming chest scans that can only be performed at a few hospitals and expose patients to potentially dangerous levels of radiation.
Researchers have spent more than 20 years working to advance computer programs, or algorithms, designed to assist doctors in determining whether at-risk patients are experiencing pulmonary embolisms. However, the results have been mixed.
“Algorithms that use EHRs, for example, have a wide range of success rates for accurately detecting clots and can be labor-intensive. Meanwhile, the more accurate ones rely heavily on CTPA data “According to the press release.
Researchers later discovered that combining algorithms that rely on EKG and EHR data could be an effective approach because EKGs are widely available and relatively simple to administer.
.
The team used data from 21,183 Mount Sinai Health System patients who showed moderate to highly suspicious signs of pulmonary embolisms to develop and test a variety of algorithms for the study.
Some algorithms were designed to screen for pulmonary embolisms using EKG data, while others were designed to use EHR data.
In each case, the algorithm learned to identify a pulmonary embolism case by comparing EKG or EHR data to corresponding CTPA results.
Furthermore, a third fusion AI algorithm was developed by combining the best-performing EKG algorithm with the best-performing EHR algorithm.
When screening for cases of acute embolism, the researchers estimated that the fusion model was 15% to 30% more accurate. Furthermore, regardless of whether race or gender was tested as a factor, the model performed best at predicting the most severe cases and remained consistent.
The team intends to further develop and test the algorithms in preparation for future clinical use.
ALSO READ: Children With Autism: Do They Find It Hard To Listen To Their Mothers? – New Study Explains
Source: Sound Health and Lasting Wealth
