To develop machine learning models using resting state fMRI data to predict postoperative functional outcomes in adult patients with high-grade gliomas (HGG), thereby enhancing surgical planning and patient counseling.
Key Findings:
Machine learning models can accurately predict postoperative functional outcomes at initial diagnosis, which may improve patient management.
Resting state fMRI provides valuable insights into intrinsic brain connectivity without task performance constraints, offering a practical alternative to task-based fMRI.
Key predictors of functional outcomes include age, resting state functional connectivity, and tumor volume, which should be considered in clinical assessments.
Interpretation:
The study demonstrates the potential of integrating resting state fMRI with machine learning to enhance surgical planning and patient counseling regarding functional outcomes post-surgery for HGG, ultimately improving patient care.
Limitations:
Retrospective design may introduce selection bias; future studies should consider prospective designs.
The study is limited to a single institution, which may affect generalizability; multi-center studies are recommended.
Potential confounding factors not accounted for in the analysis; future research should aim to include a broader range of variables.
Conclusion:
Utilizing resting state fMRI and machine learning can significantly improve the prediction of functional outcomes in HGG patients, aiding in surgical decision-making and patient management, and aligning with current advancements in neuro-oncology.
by Patrick H. Luckett, Michael O. Olufawo, Ki Yun Park, Bidhan Lamichhane, Donna Dierker, Gabriel Trevino Verastegui, John J. Lee, Peter Yang, Albert Kim, Omar H. Butt, Milan G. Chheda, Abraham Z. Snyder, Joshua S. Shimony, Eric C. Leuthardt
