Feature-based PET/MRI radiomics in patients with brain tumors

Area of Science:

• Medical Imaging
• Quantitative Analysis
• Neuro-oncology

Background:

• Radiomics extracts quantitative features from medical images (CT, MRI, PET) for enhanced clinical decision-making.
• Automated feature computation from routine follow-up scans provides cost-effective diagnostic information.
• Radiomics features can be integrated with clinical data to build predictive and prognostic models.

Purpose of the Study:

• To review current research on PET radiomics and combined PET/MRI radiomics in neuro-oncology.
• To highlight the application of radiomics in noninvasively predicting biomarkers, differentiating tumor relapse, and assessing treatment response in glioma patients.
• To discuss the growing body of evidence for amino acid PET in brain tumor diagnostics and its integration with radiomics.

Main Methods:

• Systematic review of studies investigating PET radiomics and combined PET/MRI radiomics in neuro-oncology.
• Analysis of radiomic feature extraction from medical imaging modalities, particularly PET scans.
• Evaluation of the use of radiomic models for diagnostic and prognostic applications in brain tumors.

Main Results:

• Radiomics provides valuable quantitative data from medical images for clinical applications.
• PET and combined PET/MRI radiomics show increasing potential in neuro-oncology for various diagnostic tasks.
• The integration of radiomics with PET imaging is a growing area of research for brain tumor management.

Conclusions:

• Radiomics, especially PET and PET/MRI-based approaches, offers significant potential for advancing neuro-oncology diagnostics.
• Further research is warranted to fully leverage radiomics for personalized treatment strategies in brain tumor patients.
• The combination of radiomics with advanced imaging techniques like PET is crucial for improving patient outcomes.