Using Machine Learning to Identify COVID-19-Related Stroke
Author Information
Author(s): Eylem Gul Ates, Coban Gokcen, Karakaya Jale
Primary Institution: Hacettepe University
Hypothesis
This study aims to evaluate the impact of COVID-19 on acute ischemic stroke using radiomic features extracted from brain MR images and machine learning methods.
Conclusion
The proposed radiomics-based classification model effectively distinguishes acute ischemic strokes associated with COVID-19 from other causes.
Supporting Evidence
- The study included MRI data from 57 patients diagnosed with acute ischemic stroke.
- Machine learning algorithms were used to classify strokes based on radiomic features.
- The k-NN classifier achieved the best performance with an AUC of 0.882.
Takeaway
Researchers used brain scans to teach a computer how to tell if a stroke was caused by COVID-19 or not, helping doctors make better decisions.
Methodology
This retrospective study analyzed MRI data from 57 patients diagnosed with acute ischemic stroke, using machine learning algorithms to classify COVID-19-related strokes.
Potential Biases
The strict inclusion criteria may have introduced selection bias, affecting the representativeness of the sample.
Limitations
The study had a small sample size and was conducted at a single center, which may limit the generalizability of the findings.
Participant Demographics
The average age of participants was 60.88 years, with 30 COVID-19-positive and 27 COVID-19-negative patients; 66.7% of the COVID-19-negative group were women.
Statistical Information
P-Value
0.05
Confidence Interval
95% CI: 0.806–0.900
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website