Optimal approaches to analyzing functional MRI data in glioma patients

Area of Science:

• Neuroimaging
• Neuroscience
• Medical Physics

Background:

• Resting-state fMRI (rs-fMRI) is crucial for understanding brain functional organization alterations in glioma patients.
• Existing literature lacks systematic comparisons of different rs-fMRI preprocessing and analysis techniques.
• This variability may affect the reliability of findings regarding glioma's impact on brain networks.

Purpose of the Study:

• To systematically compare the impact of alternative analytical approaches on rs-fMRI studies of glioma patients.
• To identify optimal methods for differentiating glioma patients from healthy controls using rs-fMRI data.
• To provide guidance for technical optimization in glioma neuroimaging research.

Main Methods:

• Conducted a literature survey to identify common rs-fMRI analytical approaches.
• Systematically compared atlas registration (affine vs. non-linear), parcellation schemes (anatomical vs. functional), and graph-theoretical measures.
• Analyzed rs-fMRI data from 59 glioma patients and 163 age-matched healthy controls.

Main Results:

• Non-linear registration enhanced structural alignment and functional connectivity measures compared to affine registration.
• Functionally-derived parcellation schemes yielded greater contrast between glioma patients and controls than anatomically-derived schemes.
• Graph-theoretical measures demonstrated significant sensitivity to parcellation granularity, scheme, and graph density.

Conclusions:

• The evaluation of glioma-induced functional connectome alterations is highly dependent on the chosen analytical methods.
• Non-linear registration and functionally-derived parcellation are recommended for improved sensitivity in glioma rs-fMRI studies.
• This study provides a technical framework for optimizing rs-fMRI analysis in neuro-oncology.