AI Article Synopsis
- The study developed a machine learning model to forecast patients' recovery after microsurgery for unruptured intracranial aneurysms, using data from 615 patients.
- Various machine learning techniques, including random forest, logistic regression, and support vector machine, were tested with the support vector machine achieving the highest accuracy (F1-score = 0.904).
- The findings suggest that machine learning could serve as an effective decision support tool for surgical outcomes in intracranial aneurysm treatments.
Article Abstract
Our study aimed to create a machine learning model to predict patients' functional outcomes after microsurgical treatment of unruptured intracranial aneurysms (UIA). Data on 615 microsurgically treated patients with UIA were collected retrospectively from the Electronic Health Records at N.N. Burdenko Neurosurgery Center (Moscow, Russia). The dichotomized modified Rankin Scale (mRS) at the discharge was used as a target variable. Several machine learning models were utilized: a random forest upon decision trees (RF), logistic regression (LR), support vector machine (SVM). The best result with F1-score metric = 0.904 was produced by the SVM model with a label-encode method. The predictive modeling based on machine learning might be promising as a decision support tool in intracranial aneurysm surgery.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.3233/SHTI220503 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!