Nikdokht Farid

Background: Differentiating recurrent tumor from post-treatment changes remains a major challenge in glioblastoma (GBM) patients. In this work, we compared the performance of 2 different MR perfusion techniques, dynamic susceptibility contrast (DSC), and arterial spin labeling (ASL) to differentiate recurrent tumor and post-treatment changes from the volume of cellular tumor segmented from combined Deep Learning and multimodal MRI measurements, including multishell diffusion and perfusion. Methods: In this retrospective study, 137 MRIs from 107 patients with GBM were analyzed. Cellular tumor maps were segmented by 2 radiologists based on imaging, clinical history, and pathology. Multimodal MRI with perfusion and multishell diffusion were inputted into 5 nnU-Net Deep Learning models using either DSC or ASL with combination of multishell diffusion and standard MRI sequences to segment cellular tumor. Models with DSC and ASL were compared using segmentation performance (Dice score) and accuracy to detect recurrent tumor from post-treatment changes (area under the curve [AUC] under the receiver operating characteristic curve). Results: Segmentation performances were similar in both cases, with a median Dice score of 0.75 (IQR: 0.53-0.84) for ASL and 0.76 (IQR: 0.57-0.84). AUC was 0.88 (CI 0.82-0.94) for ASL and 0.86 (CI, 0.80-0.92) for DSC, and this difference was statistically significant (P < .05, n = 10 000 permutation test). In 11 individual cases, recurring disease was detected with ASL but missed with cerebral blood volume, including recurring tumor in the vicinity of a surgical cavity (n = 5), close to the skull base (n = 1), and adjacent to an Ommaya reservoir (n = 2). Conclusions: Our results demonstrate the utility of ASL in regions where susceptibility artifacts decrease the quality of DSC images.

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To develop and validate a deep learning (DL) method to detect and segment enhancing and nonenhancing cellular tumor on pre- and posttreatment MRI scans of patients with glioblastoma and to predict overall survival (OS) and progression-free survival (PFS). Materials and Methods This retrospective study included 1397 MRIs in 1297 patients with glioblastoma, including an internal cohort of 243 MRIs (January 2010-June 2022) for model training and cross-validation and four external test cohorts. Cellular tumor maps were segmented by two radiologists based on imaging, clinical history, and pathology. Multimodal MRI with perfusion and multishell diffusion imaging were inputted into a nnU-Net DL model to segment cellular tumor. Segmentation performance (Dice score) and performance in detecting recurrent tumor from posttreatment changes (area under the receiver operating characteristic curve [AUC]) were quantified. Model performance in predicting OS and PFS was assessed using Cox multivariable analysis. Results A cohort of 178 patients (mean age, 56 years ± [SD]13; 121 male, 57 female) with 243 MRI timepoints, as well as four external datasets with 55, 70, 610 and 419 MRI timepoints, respectively, were evaluated. The median Dice score was 0.79 (IQR:0.53-0.89) and the AUC for detecting residual/recurrent tumor was 0.84 (95% CI:0.79- 0.89). In the internal test set, estimated cellular tumor volume was significantly associated with OS (hazard ratio [HR] = 1.04/mL, P < .001) and PFS (HR = 1.04/mL, P < .001) when adjusting for age, sex and gross total resection status. In the external test sets, estimated cellular tumor volume was significantly associated with OS (HR = 1.01/mL, P < .001) when adjusting for age, sex and gross total resection status. Conclusion A DL model incorporating advanced imaging could accurately segment enhancing and nonenhancing cellular tumor, classify recurrent/residual tumor from posttreatment changes, and predict OS and PFS in patients with glioblastoma. ©RSNA, 2024.

Evaluating longitudinal changes in gliomas is a time-intensive process with significant inter-rater variability. Automated segmentation could reduce inter-rater variability and increase workflow efficiency for assessment of treatment response. We sought to evaluate whether neural networks would be comparable to expert assessment of pre- and post-treatment diffuse gliomas tissue subregions including resection cavities. A retrospective cohort of 647 MRIs of patients with diffuse gliomas (average 55.1 years; 29%/36%/34% female/male/unknown; 396 pre-treatment and 251 post-treatment, median 237 days post-surgery) from 7 publicly available repositories in The Cancer Imaging Archive were split into training (536) and test/generalization (111) samples. T1, T1-post-contrast, T2 and FLAIR images were used as input into a 3D nnU-Net, to predict three tumor subregions and resection cavities. We evaluated the performance of networks trained on pre-treatment training cases (Pre-Rx network), post-treatment training cases (Post-Rx network), and both pre- and post-treatment cases (Combined networks). Segmentation performance was as good as or better than inter-rater reliability with median dice scores for main tumor subregions ranging from 0.82-0.94 and strong correlations between manually segmented and predicted total lesion volumes (0.942 values<0.98). The Combined network performed similarly to the Pre-Rx network on pre-treatment cases and the Post-Rx network on post-treatment cases with fewer false positive resection cavities (7% vs 59%). Neural networks that accurately segment pre- and post-treatment diffuse gliomas have the potential to improve response assessment in clinical trials and reduce provider burden and errors in measurement.

BACKGROUND AND PURPOSE: Treatment with bevacizumab is standard of care for recurrent high-grade gliomas; however, monitoring response to treatment following bevacizumab remains a challenge. The purpose of this study was to determine whether quantifying the sharpness of the fluid-attenuated inversion recovery hyperintense border using a measure derived from texture analysis—edge contrast—improves the evaluation of response to bevacizumab in patients with high-grade gliomas. MATERIALS AND METHODS: MRIs were evaluated in 33 patients with high-grade gliomas before and after the initiation of bevacizumab. Volumes of interest within the FLAIR hyperintense region were segmented. Edge contrast magnitude for each VOI was extracted using gradients of the 3D FLAIR images. Cox proportional hazards models were generated to determine the relationship between edge contrast and progression-free survival/overall survival using age and the extent of surgical resection as covariates. RESULTS: After bevacizumab, lower edge contrast of the FLAIR hyperintense region was associated with poorer progression-free survival (P = .009) and overall survival (P = .022) among patients with high-grade gliomas. Kaplan-Meier curves revealed that edge contrast cutoff significantly stratified patients for both progression-free survival (log-rank χ² = 8.3, P = .003) and overall survival (log-rank χ² = 5.5, P = .019). CONCLUSIONS: Texture analysis using edge contrast of the FLAIR hyperintense region may be an important predictive indicator in patients with high-grade gliomas following treatment with bevacizumab. Specifically, low FLAIR edge contrast may partially reflect areas of early tumor infiltration. This study adds to a growing body of literature proposing that quantifying features may be important for determining outcomes in patients with high-grade gliomas.

SUMMARY: Brain AVMs treated with stereotactic radiosurgery typically demonstrate a minimum latency period of 1–3 years between treatment and nidus obliteration. Assessment of treatment response is usually limited to evaluation of AVM nidus structural changes using conventional MR imaging and MRA techniques. This report describes the use of 4D Flow MRI to also measure radiation-induced hemodynamic changes in a Spetzler-Martin grade III AVM, which were detectable as early as 6 months after treatment.

BACKGROUND AND PURPOSE: ADC as a marker of tumor cellularity has been promising for evaluating the response to therapy in patients with glioblastoma but does not successfully stratify patients according to outcomes, especially in the upfront setting. Here we investigate whether restriction spectrum imaging, an advanced diffusion imaging model, performed after an operation but before radiation therapy, could improve risk stratification in patients with newly diagnosed glioblastoma relative to ADC. MATERIALS AND METHODS: Pre-radiation therapy diffusion-weighted and structural imaging of 40 patients with glioblastoma were examined retrospectively. Restriction spectrum imaging and ADC-based hypercellularity volume fraction (restriction spectrum imaging–FLAIR volume fraction, restriction spectrum imaging–contrast-enhanced volume fraction, ADC-FLAIR volume fraction, ADC–contrast-enhanced volume fraction) and intensities (restriction spectrum imaging—FLAIR 90th percentile, restriction spectrum imaging–contrast-enhanced 90th percentile, ADC-FLAIR 10th percentile, ADC–contrast-enhanced 10th percentile) within the contrast-enhanced and FLAIR hyperintensity VOIs were calculated. The association of diffusion imaging metrics, contrast-enhanced volume, and FLAIR hyperintensity volume with progression-free survival and overall survival was evaluated by using Cox proportional hazards models. RESULTS: Among the diffusion metrics, restriction spectrum imaging–FLAIR volume fraction was the strongest prognostic metric of progression-free survival (P = .036) and overall survival (P = .007) in a multivariate Cox proportional hazards analysis, with higher values indicating earlier progression and shorter survival. Restriction spectrum imaging—FLAIR 90th percentile was also associated with overall survival (P = .043), with higher intensities, indicating shorter survival. None of the ADC metrics were associated with progression-free survival/overall survival. Contrast-enhanced volume exhibited a trend toward significance for overall survival (P = .063). CONCLUSIONS: Restriction spectrum imaging–derived cellularity in FLAIR hyperintensity regions may be a more robust prognostic marker than ADC and conventional imaging for early progression and poorer survival in patients with glioblastoma. However, future studies with larger samples are needed to explore its predictive ability.