Biopsy Confirmed Glioma Recurrence Predicted by Multi-Modal Neuroimaging Metrics

Glioma Imaging by O-(2-18F-Fluoroethyl)-L-Tyrosine PET and Diffusion-Weighted MRI and Correlation With Molecular Phenotypes, Validated by PET/MR-Guided Biopsies

Gliomas exhibit high intra-tumoral histological and molecular heterogeneity. Introducing stereotactic biopsy, we achieved a superior molecular analysis of glioma using O-(2-18F-fluoroethyl)-L-tyrosine (FET)-positron emission tomography (PET) and diffusion-weighted magnetic resonance imaging (DWI). Patients underwent simultaneous DWI and FET-PET scans. Correlations between biopsy-derived tumor tissue values, such as the tumor-to-background ratio (TBR) and apparent diffusion coefficient (ADC)/exponential ADC (eADC) and histopathological diagnoses and those between relevant genes and TBR and ADC values were determined. Tumor regions with human telomerase reverse transcriptase (hTERT) mutation had higher TBR and lower ADC values. Tumor protein P53 mutation correlated with lower TBR and higher ADC values. α-thalassemia/mental-retardation-syndrome-X-linked gene (ATRX) correlated with higher ADC values. 1p/19q codeletion and epidermal growth factor receptor (EGFR) mutations correlated with lower ADC values. Isocitrate dehydrogenase 1 (IDH1) mutations correlated with higher TBRmean values. No correlation existed between TBRmax/TBRmean/ADC/eADC values and phosphatase and tensin homolog mutations (PTEN) or O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation. Furthermore, TBR/ADC combination had a higher diagnostic accuracy than each single imaging method for high-grade and IDH1-, hTERT-, and EGFR-mutated gliomas. This is the first study establishing the accurate diagnostic criteria for glioma based on FET-PET and DWI.

Machine learning applications to neuroimaging for glioma detection and classification: An artificial intelligence augmented systematic review

Glioma is the most common primary intraparenchymal tumor of the brain and the 5-year survival rate of high-grade glioma is poor. Magnetic resonance imaging (MRI) is essential for detecting, characterizing and monitoring brain tumors but definitive diagnosis still relies on surgical pathology. Machine learning has been applied to the analysis of MRI data in glioma research and has the potential to change clinical practice and improve patient outcomes. This systematic review synthesizes and analyzes the current state of machine learning applications to glioma MRI data and explores the use of machine learning for systematic review automation. Various datapoints were extracted from the 153 studies that met inclusion criteria and analyzed. Natural language processing (NLP) analysis involved keyword extraction, topic modeling and document classification. Machine learning has been applied to tumor grading and diagnosis, tumor segmentation, non-invasive genomic biomarker identification, detection of progression and patient survival prediction. Model performance was generally strong (AUC = 0.87 ± 0.09; sensitivity = 0.87 ± 0.10; specificity = 0.0.86 ± 0.10; precision = 0.88 ± 0.11). Convolutional neural network, support vector machine and random forest algorithms were top performers. Deep learning document classifiers yielded acceptable performance (mean 5-fold cross-validation AUC = 0.71). Machine learning tools and data resources were synthesized and summarized to facilitate future research. Machine learning has been widely applied to the processing of MRI data in glioma research and has demonstrated substantial utility. NLP and transfer learning resources enabled the successful development of a replicable method for automating the systematic review article screening process, which has potential for shortening the time from discovery to clinical application in medicine.

Differentiating high-grade glioma recurrence from pseudoprogression: Comparing diffusion kurtosis imaging and diffusion tensor imaging

PURPOSE

To compare the diagnostic value of DKI and DTI in differentiation of high-grade glioma recurrence and pseudoprogression (PsP).

METHOD

Forty patients with high-grade gliomas who exhibited new enhancing lesions (24 high-grade glioma recurrence and 16 PsP) within 6 months after surgery followed by completion of chemoradiation therapy. All patients underwent repeat surgery or biopsy after routine MRI and DKI (including DTI). They were histologically classified into high-grade glioma recurrence and PsP groups. DKI (mean kurtosis [MK], axial kurtosis [Ka], and radial kurtosis [Kr]) and DTI (mean diffusivity [MD] and fractional anisotropy [FA]) parameters in the enhancing lesions and in the perilesional edema were measured. Inter-group differences between high-grade glioma recurrence and PsP were compared using the Mann-Whitney U test The receiver operating characteristic (ROC) curve was used to assess differential diagnostic efficacy of each parameter, and Z-scores were used to compare the value between DKI and DTI.

RESULTS

Relative MK (rMK) was significantly higher and relative MD (rMD) was significantly lower in the enhancing lesions of high-grade glioma recurrence compared to PsP (P <  0.001, P = 0.006, respectively). The AUC was 0.914 for rMK and 0.760 for rMD, and this difference was significant (P = 0.030). In the perilesional edema, rMK values were significantly higher and rMD values were significantly lower in high-grade glioma recurrence compared to PsP (P <  0.001, P =  0.005).

CONCLUSIONS

DKI had superior performance in differentiating high-grade glioma recurrence from PsP, and rMK appeared to be the best independent predictor.