Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MRI Radiomics for IDH Genotype Prediction in Glioblastoma Diagnosis

Stanislav Kozák

TL;DR

This review assesses MRI radiomics for non-invasive IDH genotype prediction in glioblastoma, detailing a standard radiomic workflow (acquisition, segmentation, pre-processing, feature extraction, feature selection, and modeling) and summarizing three exemplar studies that applied PyRadiomics to multi-sequence MRI. Despite promising performance (AUCs around 0.88–0.96 and MCC up to ~0.68), the studies report little overlap in predictive features and are limited by small, single-institution datasets and varying pipelines. The findings underscore the potential of radiomics to support diagnosis and treatment planning but also emphasize the need for standardized methodologies, larger and more diverse datasets, and exploration of end-to-end deep learning alternatives with attention to interpretability. Overall, radiomic profiling of MRI remains a promising non-invasive tool for IDH genotype prediction, contingent on robust validation and generalizability across scanners, sequences, and patient populations.

Abstract

Radiomics is a relatively new field which utilises automatically identified features from radiological scans. It has found a widespread application, particularly in oncology because many of the important oncological biomarkers are not visible to the naked eye. The recent advent of big data, including in medical imaging, and the development of new ML techniques brought the possibility of faster and more accurate oncological diagnosis. Furthermore, standardised mathematical feature extraction based on radiomics helps to eliminate possible radiologist bias. This paper reviews the recent development in the oncological use of MRI radiomic features. It focuses on the identification of the isocitrate dehydrogenase (IDH) mutation status, which is an important biomarker for the diagnosis of glioblastoma and grade IV astrocytoma.

MRI Radiomics for IDH Genotype Prediction in Glioblastoma Diagnosis

TL;DR

This review assesses MRI radiomics for non-invasive IDH genotype prediction in glioblastoma, detailing a standard radiomic workflow (acquisition, segmentation, pre-processing, feature extraction, feature selection, and modeling) and summarizing three exemplar studies that applied PyRadiomics to multi-sequence MRI. Despite promising performance (AUCs around 0.88–0.96 and MCC up to ~0.68), the studies report little overlap in predictive features and are limited by small, single-institution datasets and varying pipelines. The findings underscore the potential of radiomics to support diagnosis and treatment planning but also emphasize the need for standardized methodologies, larger and more diverse datasets, and exploration of end-to-end deep learning alternatives with attention to interpretability. Overall, radiomic profiling of MRI remains a promising non-invasive tool for IDH genotype prediction, contingent on robust validation and generalizability across scanners, sequences, and patient populations.

Abstract

Radiomics is a relatively new field which utilises automatically identified features from radiological scans. It has found a widespread application, particularly in oncology because many of the important oncological biomarkers are not visible to the naked eye. The recent advent of big data, including in medical imaging, and the development of new ML techniques brought the possibility of faster and more accurate oncological diagnosis. Furthermore, standardised mathematical feature extraction based on radiomics helps to eliminate possible radiologist bias. This paper reviews the recent development in the oncological use of MRI radiomic features. It focuses on the identification of the isocitrate dehydrogenase (IDH) mutation status, which is an important biomarker for the diagnosis of glioblastoma and grade IV astrocytoma.
Paper Structure (21 sections, 1 figure, 3 tables)

This paper contains 21 sections, 1 figure, 3 tables.

Table of Contents

  1. Introduction
  2. Genotype Alterations in Grade IV Gliomas
  3. IDH Mutation
  4. MRI Radiomics Workflow
  5. Image Acquisition
  6. Image Segmentation
  7. Image Pre-processing
  8. Feature Extraction
  9. Feature Selection and Dimension Reduction
  10. Classification Models
  11. Methodology
  12. Image Acquisition
  13. Image Segmentation
  14. Image Pre-processing Techniques
  15. Feature Extraction
  16. ...and 6 more sections

Figures (1)

  • Figure 1: Radiomics pipeline based on traditional ML models.