Predicting Motor Function After Brain Tumor Surgery
Author Information
Author(s): Muacevic Alexander, Adler John R, Yuno Takeo, Nakade Yusuke, Nakada Mitsutoshi, Kinoshita Masashi, Nakata Masako, Nakagawa Shiori, Oe Hiroyasu, Mori Mika, Wada Takashi, Kanamori Hajime
Primary Institution: Kanazawa University Hospital
Hypothesis
Can a decision tree model improve the prediction of postoperative motor function using motor evoked potential monitoring data?
Conclusion
The decision tree model is effective in predicting postoperative paralysis after brain tumor resection.
Supporting Evidence
- The decision tree model achieved an accuracy of 0.921 and sensitivity of 0.917 in internal validation.
- The amplitude of the last measured MEP was a significant predictor of postoperative motor function.
- External validation showed the decision tree model outperformed traditional cutoff values from ROC curves.
Takeaway
Doctors can use a special computer model to better guess how well patients will move after brain surgery for tumors.
Methodology
This retrospective cohort study analyzed data from 125 patients who underwent brain tumor resection with MEP monitoring, using decision tree and ROC curve analyses.
Potential Biases
Potential biases include the retrospective design and variability in postoperative assessment timing.
Limitations
The study is retrospective, conducted at a single institution, and has a small external validation sample size.
Participant Demographics
The study included 125 patients, with four pediatric patients aged between six and 15 years.
Statistical Information
P-Value
0.0344
Confidence Interval
0.9746-0.9990
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website