Preferential tumor localization in relation to 18F-FDOPA uptake for lower-grade gliomas

PURPOSE

Although tumor localization and 3,4-dihydroxy-6-¹⁸F-fluoro-L-phenylalanine (FDOPA) uptake may have an association, preferential tumor localization in relation to FDOPA uptake is yet to be investigated in lower-grade gliomas (LGGs). This study aimed to identify differences in the frequency of tumor localization between FDOPA hypometabolic and hypermetabolic LGGs using a probabilistic radiographic atlas.

METHODS

Fifty-one patients with newly diagnosed LGG (WHO grade II, 29; III, 22; isocitrate dehydrogenase wild-type, 21; mutant 1p19q non-codeleted,16; mutant codeleted, 14) who underwent FDOPA positron emission tomography (PET) were retrospectively selected. Semiautomated tumor segmentation on FLAIR was performed. Patients with LGGs were separated into two groups (FDOPA hypometabolic and hypermetabolic LGGs) according to the normalized maximum standardized uptake value of FDOPA PET (a threshold of the uptake in the striatum) within the segmented regions. Spatial normalization procedures to build a 3D MRI-based atlas using each segmented region were validated by an analysis of differential involvement statistical mapping.

RESULTS

Superimposition of regions of interest showed a high number of hypometabolic LGGs localized in the frontal lobe, while a high number of hypermetabolic LGGs was localized in the insula, putamen, and temporal lobe. The statistical mapping revealed that hypometabolic LGGs occurred more frequently in the superior frontal gyrus (close to the supplementary motor area), while hypermetabolic LGGs occurred more frequently in the insula.

CONCLUSION

Radiographic atlases revealed preferential frontal lobe localization for FDOPA hypometabolic LGGs, which may be associated with relatively early detection.

Differentiating IDH status in human gliomas using machine learning and multiparametric MR/PET

Background

The purpose of this study was to develop a voxel-wise clustering method of multiparametric magnetic resonance imaging (MRI) and 3,4-dihydroxy-6-[¹⁸F]-fluoro-L-phenylalanine (FDOPA) positron emission tomography (PET) images using an unsupervised, two-level clustering approach followed by support vector machine in order to classify the isocitrate dehydrogenase (IDH) status of gliomas.

Methods

Sixty-two treatment-naïve glioma patients who underwent FDOPA PET and MRI were retrospectively included. Contrast enhanced T1-weighted images, T2-weighted images, fluid-attenuated inversion recovery images, apparent diffusion coefficient maps, and relative cerebral blood volume maps, and FDOPA PET images were used for voxel-wise feature extraction. An unsupervised two-level clustering approach, including a self-organizing map followed by the K-means algorithm was used, and each class label was applied to the original images. The logarithmic ratio of labels in each class within tumor regions was applied to a support vector machine to differentiate IDH mutation status. The area under the curve (AUC) of receiver operating characteristic curves, accuracy, and F1-socore were calculated and used as metrics for performance.

Results

The associations of multiparametric imaging values in each cluster were successfully visualized. Multiparametric images with 16-class clustering revealed the highest classification performance to differentiate IDH status with the AUC, accuracy, and F1-score of 0.81, 0.76, and 0.76, respectively.

Conclusions

Machine learning using an unsupervised two-level clustering approach followed by a support vector machine classified the IDH mutation status of gliomas, and visualized voxel-wise features from multiparametric MRI and FDOPA PET images. Unsupervised clustered features may improve the understanding of prioritizing multiparametric imaging for classifying IDH status.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40644-021-00396-5.

Abstract P348: Perfusion Collateral Index vs. Hypoperfusion Intensity Ratio in Assessment of Angiographic Collateral Scores in Patients With Acute Ischemic Stroke

Background and Purpose: In acute ischemic stroke (AIS), perfusion imaging, while not directly visualizing collateral vessels, can provide important insight into collateral robustness, indexed by perfusion lesion volume and by perfusion lesion heterogeneity. Two proposed perfusion lesion heterogeneity measures indexing collateral status are the Perfusion Collateral Index (PCI) and Hypoperfusion Intensity Ratio (HIR), but their accuracy compared with direct collateral assessment on DSA has been incompletely characterized.

Methods: Consecutive AIS patients with anterior circulation large vessel occlusion who underwent pre-endovascular thrombectomy MRI perfusion imaging were included. MRI measures analyzed were: 1) Perfusion Collateral Index ( PCI) - the volume of moderately hypoperfused tissue (arterial tissue delay time between 2 and 6 seconds: ATD 2-6sec ) multiplied by its corresponding relative cerebral blood volume using Olea software; 2) Hypoperfusion Intensity Ratio (HIR) ratio of moderate TMax >6 s lesion volume versus severe Tmax >10 s lesion volume with the RAPID software program. DSA collateral scores were evaluated by ASITN grading and dichotomized to inadequate (ASTIN <2) vs. adequate (ASTIN ≥3).

Results: Among 48 patients meeting entry criteria, age (mean ± SD) was 70 (± 15.2), 54% were female, and NIHSS (median, IQR) was 15 (10-19). For HIR, there was no significant difference in score values in patients with adequate vs inadequate collaterals: 0.35 ± 0.20 vs 0.39 ± 0.25, p=0.68. ROC analysis using previously described cut-off of 0.4 resulted in an AUC of 0.52 and sensitivity/specificity of 71% / 33%. For PCI, score values were significantly higher in patients with adequate vs inadequate collaterals, 117 ± 61 vs. 57 ± 41, p=0.002. ROC analysis using previously described cut-off of 62 resulted in an AUC of 0.8 and sensitivity/specificity of 84% / 78%.

Conclusion: Collateral status can be accurately assessed on perfusion MRI with the Perfusion Collateral Index, which outperformed the Hypoperfusion Intensity Ratio. MRI-PCI is an informative imaging biomarker of collateral status in patients with AIS.

Abstract P378: Automated Estimation of Ischemic Core Volume on Non-Contrast-Enhanced Computed Tomography via Machine Learning

Background and Purpose: Accurate estimation of ischemic core on baseline imaging has treatment implications in patients with acute ischemic stroke (AIS). Machine learning (ML) algorithms have shown promising results in estimating ischemic core using routine non-contrast CT (NCCT). We used a ML-trained algorithm to quantify ischemic core volume on NCCT and compared the results to concurrent diffusion MRI as the reference standard in patients with AIS.

Methods: We analyzed consecutive anterior circulation AIS patients who had baseline (pretreatment) NCCT and MRI (DWI). Ischemic lesion volume was calculated on MRI-DWI using an automated software (Olea Medical SAS, La Ciotat, France). An automatic segmentation approach using a combination of traditional 3D graphics and statistical methods, and ML classification techniques (Brainomix, Oxford, United Kingdom) was used to identify ischemic core voxels on NCCT. Total ischemic core volumes on ML-NCCT and DWI-MR were quantitatively compared by Bland-Altman plots and Pearson correlation.

Results: A total of 50 patients (27 female, 23 male, mean age 72.6 years) were included. Baseline imaging was performed within 173 ± 143 minutes (mean ± SD) from symptom onset. The mean time difference between MRI and NCCT was 72 min. The baseline NIHSS was 14, 8-21 (Median, IQR). Algorithm-segmented ischemic core volume detected on NCCT was median 12.7 mL, IQR 3.5-26.0 mL. Ischemic core volume on DWI MRI was median 8.8 mL, IQR 3.2-34.0 mL. ML-NCCT core volumes significantly correlated with DWI MRI core volumes, r =0.61, p <0.001. The mean difference between the ML-NCCT and DWI MRI core volumes was 12.4 mL, p =0.81. For the reperfusion treatment threshold of an ischemic core volume within 70 mL, while no patients would have been excluded using our algorithm, five patients would have been incorrectly dichotomized as having an ischemic volume of <70 mL compared to MRI.

Conclusion: This ML-approach accurately quantifies ischemic core volume on NCCT compared to the reference standard of diffusion MRI in patients with AIS.

The Extraordinary Case of Jahi McMath

Jahi McMath was a 13-year-old girl who was declared brain dead on December 12, 2013, after a hemorrhagic complication following complex oropharyngeal surgery. Her case gained international attention as her mother fought a legal battle to keep her on life support. Upon issuance of a death certificate shortly after the declaration of brain death, Jahi was transferred from California to New Jersey, where the law includes a religious exemption from the neurologic determination of death. There she became statutorily resurrected and was treated as a comatose, living patient for the next four and a half years. During that time, she underwent menarche and other aspects of puberty and developed intermittent responsiveness to commands, documented by eyewitness attestations and multiple home videos. Jahi died on June 18, 2018, from abdominal complications. This article summarizes her clinical history over those intervening years, taken directly from her medical records and personal observation. Her case represents an instance of a false-positive diagnosis of brain death, unquestionably made according to both the pediatric and adult guidelines, reinforced by four false-positive EEGs and a false-positive radionuclide blood flow test. The bioethical consequences of a nonnegligible risk of false-positive declaration of death are profound.

Metabolic characterization of human IDH mutant and wild type gliomas using simultaneous pH- and oxygen-sensitive molecular MRI

BACKGROUND

Isocitrate dehydrogenase 1 (IDH1) mutant gliomas are thought to have distinct metabolic characteristics, including a blunted response to hypoxia and lower glycolytic flux. We hypothesized that non-invasive quantification of abnormal metabolic behavior in human IDH1 mutant gliomas could be performed using a new pH- and oxygen-sensitive molecular MRI technique.

METHODS

Simultaneous pH- and oxygen-sensitive MRI was obtained at 3T using amine CEST-SAGE-EPI. The pH-dependent measure of the magnetization transfer ratio asymmetry (MTRasym) at 3 ppm and oxygen-sensitive measure of R2' were quantified in 90 patients with gliomas. Additionally, stereotactic, image-guided biopsies were performed in 20 patients for a total of 52 samples. The association between imaging measurements and hypoxia-inducible factor 1 alpha (HIF1α) expression was identified using Pearson correlation analysis.

RESULTS

IDH1 mutant gliomas exhibited significantly lower MTRasym at 3 ppm, R2', and MTRasymxR2' (P = 0.007, P = 0.003, and P = 0.001, respectively). MTRasymxR2' could identify IDH1 mutant gliomas with a high sensitivity (81.0%) and specificity (81.3%). HIF1α was positively correlated with MTRasym at 3 ppm, R2' and MTRasymxR2' in IDH1 wild type (r = 0.610, P = 0.003; r = 0.667, P = 0.008; r = 0.635, P = 0.006), but only MTRasymxR2' in IDH1 mutant gliomas (r = 0.727, P = 0.039).

CONCLUSIONS

IDH1 mutant gliomas have distinct metabolic and microenvironment characteristics compared with wild type gliomas. An imaging biomarker combining tumor acidity and hypoxia (MTRasymxR2') can differentiate IDH1 mutation status and is correlated with tumor acidity and hypoxia.

NIMG-49. VALIDATION OF SIMULTANEOUS PH- AND O2-WEIGHTED MOLECULAR MRI USING 18F-FDG PET, IMMUNOHISTOCHEMISTRY, AND BIOENERGETICS

This study aimed to investigate the potential of a novel MRI metric related to aerobic glycolysis in patients with diffuse gliomas. All subjects (Study I–III; 7, 26, and 11 subjects, respectively) were scanned on 3-T systems and underwent pH-weighted amine chemical exchange saturation transfer spin-and-gradient-echo echoplanar imaging (CEST-SAGE-EPI) or CEST-EPI, and perfusion imaging. Relative oxygen extraction fraction (rOEF) was calculated by dividing hypoxia-sensitive R2’ by normalized relative cerebral blood volume (nrCBV); a novel metric that characterizes glycolytic status (aerobic glycolytic index, AGI) was calculated by dividing amine CEST contrast by rOEF. Patients in Study I were additionally scanned by 18F-fluorodeoxyglucose (FDG)-PET. Stereotactic image-guided biopsies were performed on patients in Study II and III, and samples were analyzed by immunohistochemistry (IHC) and extracellular flux bioenergetic analysis, respectively. Pairwise correlation between MR metrics and standardized uptake value of 18F-FDG, IHC metrics, or indices of cellular metabolism was calculated using Spearman’s correlation analysis. In Study I, AGI showed very strong significant correlation with 18F-FDG uptake in glioma (correlation coefficient ρ = 0.86, P =0.014). In Study II, AGI was significantly correlated with glucose transporter 3 (ρ = 0.71; P = 0.0041) and hexokinase 2 (ρ = 0.73; P = 0.0029) in IDH wild-type glioma, while it was significantly correlated with monocarboxylate transporter 1 (ρ = 0.59; P = 0.0094) in IDH mutant glioma. This result may reflect the different glycolytic statuses of these gliomas; specifically, the rate-limiting steps in glycolysis. In Study III, a strong significant correlation with cellular AGI derived from the bioenergetic analysis was found for AGI derived from MRI (ρ = 0.79, P = .036). In conclusion, AGI derived from MRI was correlated with FDG, IHC measurements, and cellular AGI. Future studies investigating the clinical utility of AGI in prediction and evaluation of treatment effects are warranted.

NIMG-52. CHARACTERIZATION OF COGNITIVE FUNCTION IN SURVIVORS OF DIFFUSE GLIOMAS USING MORPHOMETRIC CORRELATION NETWORKS (MCN)

Advanced multimodality treatments have led to improved survival and quality of life for patients with diffuse gliomas, and widespread functional reorganization has been reported to be associated with improved cognitive function. However, investigation of morphological alterations in patients with diffuse gliomas has been problematic, largely due to the tumor resection cavities. This pilot study has overcome these challenges and aims to characterize the relationship between structural plasticity and cognitive function in survivors of diffuse gliomas. High-resolution T1 weighted images were collected from 24 patients with diffuse gliomas (mean age 44.5 ± 11.5) who had completed their treatment within the previous ten years. Interregional correlations of cortical thickness were computed to establish morphometric correlation networks (MCN) for twelve cognitively impaired and twelve non-impaired glioma patients, as well as correlated with self-reported cognitive impairment. Our findings demonstrated that both cognitively impaired (σ=1.5979) and non-impaired (σ=1.3683) glioma patients have a small world architecture in disrupted morphological networks. Although the left fusiform (p=0.0409), left inferior (p=0.0209), and temporal (p=0.0173) gyri were observed to be thicker in non-impaired patients, the robustness of their morphological network was weak and easily vulnerable to pathological attacks and neurological deterioration. Furthermore, regions such as the superior temporal gyrus (p=0.0126) and rostral middle frontal gyrus (p=0.0148) were not only identified as predominant nodes in the MCN of patients but were also found to have greater gray matter thickness, which is associated with better FACT-cognitive function. Together, these results support our hypothesis that a widespread morphological network is altered in survivors of diffuse gliomas. Predominant regions obtained by topological analysis may lead to reliable imaging biomarkers that help evaluate patients’ cognition and ability to function.

Relative oxygen extraction fraction (rOEF) MR imaging reveals higher hypoxia in human epidermal growth factor receptor (EGFR) amplified compared with non-amplified gliomas

PURPOSE

Epidermal growth factor receptor (EGFR) amplification promotes gliomagenesis and is linked to lack of oxygen within the tumor microenvironment. Using hypoxia-sensitive spin-and-gradient echo echo-planar imaging and perfusion MRI, we investigated the influence of EGFR amplification on tissue oxygen availability and utilization in human gliomas.

METHODS

This study included 72 histologically confirmed EGFR-amplified and non-amplified glioma patients. Reversible transverse relaxation rate (R₂'), relative cerebral blood volume (rCBV), and relative oxygen extraction fraction (rOEF) were calculated for the contrast-enhancing and non-enhancing tumor regions. Using Student t test or Wilcoxon rank-sum test, median R₂', rCBV, and rOEF were compared between EGFR-amplified and non-amplified gliomas. ROC analysis was performed to assess the ability of imaging characteristics to discriminate EGFR amplification status. Overall survival (OS) was determined using univariate and multivariate cox models. Kaplan-Meier survival curves were plotted and compared using the log-rank test.

RESULTS

EGFR amplified gliomas exhibited significantly higher median R₂' and rOEF than non-amplified gliomas. ROC analysis suggested that R₂' (AUC = 0.7190; P = 0.0048) and rOEF (AUC = 0.6959; P = 0.0156) could separate EGFR status. Patients with EGFR-amplified gliomas had a significantly shorter OS than non-amplified patients. Univariate cox regression analysis determined both R₂' and rOEF significantly influence OS. No significant difference was observed in rCBV between patient cohorts nor was rCBV found to be an effective differentiator of EGFR status.

CONCLUSION

Imaging of tumor oxygen characteristics revealed EGFR-amplified gliomas to be more hypoxic and contribute to shorter patient survival than EGFR non-amplified gliomas.

Compensatory brainstem functional and structural connectivity in patients with degenerative cervical myelopathy by probabilistic tractography and functional MRI

Degenerative cervical myelopathy (DCM) is the most common cause of spinal cord impairment in adults. Previous supraspinal investigations have primarily focused on cortical changes in this patient population. As the nexus between the brain and the spinal cord, the brainstem has been understudied in patients with DCM. The current study examined the structural and functional connectivity between the brainstem and cortex in DCM patients using probabilistic tractography and resting-state functional MRI. A total of 26 study patients and 32 neurologically intact, healthy volunteers (HCs) participated in this prospective analysis. The study cohort included DCM patients (n = 18), as well as neurologically asymptomatic patients with evidence of cervical spine degenerative changes and spinal cord compression (n = 8). Results of the study demonstrated significant differences in fiber density (FD), fiber cross-section (FDC), and the functional connectivity (FC) between the study cohort and HCs. Through seeding the brainstem, the study cohort showed reductions in FD and FDC along the corticospinal tract, including regions extending through the corona radiata and internal capsule. By correlating FD and FDC with the Neck Disability Index (NDI), and the modified Japanese Orthopaedic Association (mJOA), we identified increasing total volume of projections to the thalamus, basal ganglia, and internal capsule, and increased functional connectivity to visual network and the posterior parietal cortices. These results support our hypothesis that DCM patients tend to have long-term FC reorganization not only localized to sensorimotor regions, but also to regulatory and visual processing regions, designed to ultimately preserve neurological function.

Multiparametric MR-PET measurements in hypermetabolic regions reflect differences in molecular status and tumor grade in treatment-naïve diffuse gliomas

PURPOSE

To assess whether hypermetabolically-defined regions of interest (ROIs) on 3,4-dihydroxy-6-[18F]-fluoro-L-phenylalanine (FDOPA) positron emission tomography (PET) could be used to evaluate physiological features and whether there are measurable differences between molecular subtypes and tumor grades.

METHODS

Sixty-eight treatment-naïve glioma patients who underwent FDOPA PET and magnetic resonance imaging (MRI) were retrospectively included. Fluid-attenuated inversion recovery hyperintense regions (FLAIR_ROI) were segmented. FDOPA hypermetabolic regions (FDOPA_ROI, tumor-to-striatum ratios > 1) within FLAIR_ROI were extracted. Normalized maximum standardized uptake value (nSUV_max), volume of each ROI, and median relative cerebral blood volume (rCBV) and apparent diffusion coefficient (ADC) within FLAIR_ROI or FDOPA_ROI were calculated. Imaging metrics were compared using Students t or Mann-Whitney U tests. Area under the curve (AUC) of receiver-operating characteristic curves were used to determine whether imaging metrics within FLAIR_ROI or FDOPA_ROI can discriminate different molecular statuses or grades.

RESULTS

Using either FLAIR_ROI or FDOPA_ROI, the nSUV_max and rCBV were significantly higher and the ADC was lower in isocitrate dehydrogenase (IDH) wild-type than mutant gliomas, and in higher-grade gliomas (HGGs) than lower-grade gliomas (LGGs). The FDOPA_ROI volume was significantly higher in 1p19q codeleted than non-codeleted gliomas, and in HGGs than LGGs. Although not significant, imaging metrics extracted by FDOPA_ROI discriminated molecular status and tumor grade more accurately than those extracted by FLAIR_ROI (AUC of IDH status, 0.87 vs. 0.82; 1p19q status, 0.78 vs. 0.73; grade, 0.87 vs. 0.76).

CONCLUSION

FDOPA hypermetabolic ROI may extract useful imaging features of gliomas, which can illuminate biological differences between different molecular status or tumor grades.

Focal cortical dysplasia imaging discrepancies between MRI and FDG-PET: Unique association with temporal lobe location

PURPOSE

Although magnetic resonance imaging (MRI) and ¹⁸F-2-fluorodeoxyglucose-positron emission tomography (FDG-PET) are used for pre-surgical assessment of focal cortical dysplasia (FCD), they often disagree. This study aimed to identify factors that contribute to discrepancies in FCD imaging between MRI and FDG-PET.

METHODS

Sixty-two patients (mean age, 18.9 years) with a FCD type I or II were retrospectively selected. These patients were visually categorized into two groups: 1) extent of PET abnormality larger than MRI abnormality and 2) vice versa or equivalent. Predictive factors of these two groups were analyzed by multivariate logistic regression. The extent of hypometabolic transient zone surrounding FCDs and their mean standardized uptake values were measured and compared by the Mann-Whitney U-test.

RESULTS

FCDs were detected on MRI and PET in 46 and 55 patients, respectively, whereas no abnormality was detected in 4 patients. The PET hypometabolic areas were larger than the MRI abnormal areas in 26 patients (88 % in the temporal lobe), whereas the PET hypometabolic areas were equivalent or smaller than the MRI abnormal areas in 32 patients (69 % in the frontal lobe). The temporal lobe location was an independent predictor for differentiating the two groups (OR = 35.2, 95 % CI = 6.81-168.0, P < .001). The temporal lobe lesions had significantly wider transient zones and lower standardized uptake values than those in the other lobes (P < .001, both).

CONCLUSION

The discrepancies between MRI and FDG-PET findings of FCD were associated with temporal lobe location.

Voxelwise and Patientwise Correlation of 18F-FDOPA PET, Relative Cerebral Blood Volume, and Apparent Diffusion Coefficient in Treatment-Naïve Diffuse Gliomas with Different Molecular Subtypes

Our purpose was to identify correlations between ¹⁸F-fluorodihydroxyphenylalanine (¹⁸F-FDOPA) uptake and physiologic MRI, including relative cerebral blood volume (rCBV) and apparent diffusion coefficient (ADC), in gliomas with different molecular subtypes and to evaluate their prognostic values. Methods: Sixty-eight treatment-naïve glioma patients who underwent ¹⁸F-FDOPA PET and physiologic MRI were retrospectively selected (36 with isocitrate dehydrogenase wild-type [IDH_wt], 16 with mutant 1p/19q noncodeleted [IDH_m-noncodel], and 16 with mutant codeleted [IDH_m-codel]). Fluid-attenuated inversion recovery hyperintense areas were segmented and used as regions of interest. For voxelwise and patientwise analyses, Pearson correlation coefficients (r _voxelwise and r _patientwise) between the normalized SUV (nSUV), rCBV, and ADC were evaluated. Cox regression analysis was performed to investigate the associations between overall survival and r _voxelwise, maximum or median nSUV, median rCBV, or median ADC. Results: For IDH_wt and IDH_m-noncodel gliomas, nSUV demonstrated significant positive correlations with rCBV (r _voxelwise = 0.25 and 0.31, respectively; r _patientwise = 0.50 and 0.70, respectively) and negative correlations with ADC (r _voxelwise = -0.19 and -0.19, respectively; r _patientwise = -0.58 and -0.61, respectively) in both voxelwise and patientwise analyses. IDH_m-codel gliomas demonstrated a significant positive correlation between nSUV and ADC only in voxelwise analysis (r _voxelwise = 0.18). In Cox regression analysis, r _voxelwise between nSUV and rCBV (hazard ratio, 28.82) or ADC (hazard ratio, 0.085) had significant associations with overall survival for only IDH_wt gliomas. Conclusion: IDH_m-codel gliomas showed distinctive patterns of correlations between amino acid PET and physiologic MRI. Stronger correlations between nSUV and rCBV or ADC may result in a worse prognosis for IDH_wt gliomas.

Reperfusion Into Severely Damaged Brain Tissue Is Associated With Occurrence of Parenchymal Hemorrhage for Acute Ischemic Stroke

Background and Purpose: This study aims to quantify the reperfusion status within severely damaged brain tissue and to evaluate its relationship with high grade of hemorrhagic transformation (HT).

Methods: Pseudo-continuous ASL was performed along with DWI in 102 patients within 24 h post-treatments. The infarction core was identified using ADC values <550 × 10⁻⁶ mm²/s. CBF within the infarction core and its contralateral counterpart were acquired. CBF at the 25th, median, and 75th percentiles of the contralateral counterpart were used as thresholds and the ASL reperfusion volume above the threshold was labeled as vol-25, -50, and -75, respectively. Recanalization was defined according to Thrombolysis in Myocardial Infarction (TIMI) criteria.

Results: Quantified reperfusion within the infarction core differed significantly in patients with complete and incomplete recanalization. In the ROC analysis for the prediction of parenchymal hematoma (PH), ASL reperfusion vol-25 had the highest area under the curve (AUC) when compared with ASL vol-50 and ASL vol-75. ASL reperfusion vol-25 had significantly higher AUC compared with ADC threshold volume in the prediction of PH (0.783 vs. 0.685, P = 0.0036) and PH-2 (0.844 vs. 0.754, P = 0.0035). In stepwise multivariate logistic regression analysis, only ASL reperfusion vol-25 emerged as an independent predictor of PH (OR = 3.51, 95% CI: 1.65–7.45, P < 0.001) and PH-2 (OR = 2.32, 95% CI: 1.13–4.76, P = 0.022).

Conclusions: Increased reperfusion volume within severely damaged brain tissue is associated with the occurrence of higher grade of HT.

Quantitative Analysis of Spinal Canal Areas in the Lumbar Spine: An Imaging Informatics and Machine Learning Study

BACKGROUND AND PURPOSE

Quantitative imaging biomarkers have not been established for the diagnosis of spinal canal stenosis. This work aimed to lay the groundwork to establish such biomarkers by leveraging the developments in machine learning and medical imaging informatics.

MATERIALS AND METHODS

Machine learning algorithms were trained to segment lumbar spinal canal areas on axial views and intervertebral discs on sagittal views of lumbar MRIs. These were used to measure spinal canal areas at each lumbar level (L1 through L5). Machine-generated delineations were compared with 2 sets of human-generated delineations to validate the proposed techniques. Then, we use these machine learning methods to delineate and measure lumbar spinal canal areas in a normative cohort and to analyze their variation with respect to age, sex, and height using a variable-intercept mixed model.

RESULTS

We established that machine-generated delineations are comparable with human-generated segmentations. Spinal canal areas as measured by machine are statistically significantly correlated with height (P < .05) but not with age or sex.

CONCLUSIONS

Our machine learning methodology demonstrates that this important anatomic structure can be accurately detected and quantitatively measured without human input in a manner comparable with that of human raters. Anatomic deviations measured against the normative model established here could be used to flag spinal stenosis in the future.

Diffusion MRI changes in the anterior subventricular zone following chemoradiation in glioblastoma with posterior ventricular involvement

INTRODUCTION

There is growing evidence that the subventricular zone (SVZ) plays a key role in glioblastoma (GBM) tumorigenesis. However, little is known regarding how the SVZ, which is a harbor for adult neural stem cells, may be influenced by chemoradiation. The current diffusion-weighted imaging (DWI) study explored ipsilateral and contralateral alterations in the anterior SVZ in GBM patients with posterior enhancing lesions following chemoradiation.

METHODS

Forty GBM patients with tumor involvement in the posterior SVZ (mean age = 57 ± 10; left-hemisphere N = 25; right-hemisphere N = 15) were evaluated using DWI before and after chemoradiation. Regions-of-interest were drawn on the ipsilesional and contralesional anterior SVZ on apparent diffusion coefficient (ADC) maps for both timepoints. ADC histogram analysis was performed by modeling a bimodal, double Gaussian distribution to obtain ADC_L, defined as the mean of the lower Gaussian distribution.

RESULTS

The ipsilesional SVZ had lower ADC_L values compared to the contralesional SVZ before treatment (mean difference = 0.025 μm²/ms; P = 0.007). Following chemoradiation, these changes were no longer observed (mean difference = 0.0025 μm²/ms; P > 0.5), as ADC_L values of the ipsilesional SVZ increased (mean difference = 0.026 μm²/ms; P = 0.037). An increase in ipsilesional ADC_L was associated with shorter progression-free (P = 0.0119) and overall survival (P = 0.0265).

CONCLUSIONS

These preliminary observations suggest baseline asymmetry as well as asymmetric changes in the SVZ proximal (ipsilesional) to the tumor with respect to contralesional SVZ regions may be present in GBM, potentially implicating this region in tumorigenesis and/or treatment resistance.

Evaluating Radiology Result Communication in the Emergency Department

PURPOSE

To assess the pattern of result communication that occurs between radiologists and referring physicians in the emergency department setting.

METHODS

An institutional review board-approved prospective study was performed at a large academic medical center with 24/7 emergency radiology cover. Emergency radiologists logged information regarding all result-reporting communication events that occurred over a 168-hour period.

RESULTS

A total of 286 independent result communication events occurred during the study period, the vast majority of which occurred via telephone (232/286). Emergency radiologists spent 10% of their working time communicating results. Similar amounts of time were spent discussing negative and positive cross-sectional imaging examinations. In a small minority of communication events, additional information was gathered through communication that resulted in a change of interpretation from a normal to an abnormal study.

CONCLUSIONS

Effective and efficient result communication is critical to care delivery in the emergency department setting. Discussion regarding abnormal cases, both in person and over the phone, is encouraged. However, in the emergency setting, time spent on routine direct communication of negative examination results in advance of the final report may lead to increased disruptions, longer turnaround times, and negatively impact patient care. In very few instances, does the additional information gained from the communication event result in a change of interpretation?

Abstract WMP27: PAT Model Accurately Predicts Aneurysm Enlargement in 16 Growing Aneurysm Cases

Objective: Imaging technology for unruptured intracranial aneurysms (UIA) has improved detection of such aneurysms. However, there is limited information on UIA change over time, and how to predict the rate of enlargement. The objective of this study was to quantify the accuracy of the Predicted Aneurysm Trajectory (PAT) model recently developed by Chien et al. (J Neurosurgery. 2019; Mar 1:1-11).

Methods: Patients diagnosed with UIA were prospectively enrolled at the UCLA Medical Center, and followed through serial imaging. 16 UIA cases exhibiting growth across multiple follow-ups were included in this study. Prior images and medical records were collected. Characteristics relevant to the PAT model (mean ± stdev), including initial UIA size (7.26 ± 6.38), patient age (67.4 ± 9.48 yrs.), sex (4 male), history of smoking (n=5), hypothyroidism (n=4), and follow-up duration (36.5 ± 50.0 mos.) were used to predict UIA size at each follow-up. Predicted and actual UIA sizes at follow-up were compared using symmetric mean absolute percentage error (SMAPE) with percentage error ranging from 0-100%.

Results: The 16 UIA cases were split by initial UIA size. For UIA smaller than 7 mm (10 cases, 23 follow-up), SMAPE = 11.13%. For UIA greater than 7 mm (6 cases, 15 follow-up), SMAPE = 8.07%. For all UIA cases (16 cases, 38 follow-up), SMAPE = 9.92%.

Conclusions: The PAT model predicts the rate of enlargement for UIA, as opposed to whether or not UIA will grow. With this new sample of data, we found the predicted UIA size at follow-up to be quite accurate, deviating in the range of 10% from the actual, measured size. Patient characteristics such as the demographics and behavior included in the model influence the growth of UIA, which allows prediction of growth to optimize treatment and management in future cases.

Rate of change in maximum 18F-FDOPA PET uptake and non-enhancing tumor volume predict malignant transformation and overall survival in low-grade gliomas

PURPOSE

To examine whether the rate of change in maximum ¹⁸F-FDOPA PET uptake and the rate of change in non-enhancing tumor volume could predict malignant transformation and residual overall survival (OS) in low grade glioma (LGG) patients who received serial ¹⁸F-FDOPA PET and MRI scans.

METHODS

27 LGG patients with ≥ 2 ¹⁸F-FDOPA PET and MRI scans between 2003 and 2016 were included. The rate of change in FLAIR volume (uL/day) and maximum normalized ¹⁸F-FDOPA specific uptake value (nSUV_max/month), were compared between histological and molecular subtypes. General linear models (GLMs) were used to integrate clinical information with MR-PET measurements to predict malignant transformation. Cox univariate and multivariable regression analyses were performed to identify imaging and clinical risk factors related to OS.

RESULTS

A GLM using patient age, treatment, the rate of change in FLAIR and ¹⁸F-FDOPA nSUV_max could predict malignant transformation with > 67% sensitivity and specificity (AUC = 0.7556, P = 0.0248). A significant association was observed between OS and continuous rates of change in PET uptake (HR = 1.0212, P = 0.0034). Cox multivariable analysis confirmed that continuous measures of the rate of change in PET uptake was an independent predictor of OS (HR = 1.0242, P = 0.0033); however, stratification of patients based on increasing or decreasing rate of change in FLAIR (HR = 2.220, P = 0.025), PET uptake (HR = 2.148, P = 0.0311), or both FLAIR and PET (HR = 2.354, P = 0.0135) predicted OS.

CONCLUSIONS

The change in maximum normalized ¹⁸F-FDOPA PET uptake, with or without clinical information and rate of change in tumor volume, may be useful for predicting the risk of malignant transformation and estimating residual survival in patients with LGG.

Leptomeningeal Carcinomatosis of Prostate Cancer: A Case Report and Review of the Literature

Leptomeningeal carcinomatosis is a rare complication of prostate cancer. It is likely underdiagnosed as suggested by autopsy studies and is expected to become more prevalent with increasing survival of prostate cancer patients. Prostate cancer leptomeningeal carcinomatosis is associated with rapid functional decline and a median survival of approximately 1 month. Diagnosis is challenging because the clinical manifestations are varied, and no gold-standard diagnostic approach exists. Treatment of prostate cancer leptomeningeal carcinomatosis is not standardized and multiple approaches have been reported, mostly as case studies. Herein we report a case of a 73-year-old patient with metastatic castration-resistant prostate cancer who presented to our clinic with subacute cognitive decline, ataxia, and urinary incontinence, and was found to have leptomeningeal carcinomatosis.

Alterations in Cortical Thickness and Subcortical Volume are Associated With Neurological Symptoms and Neck Pain in Patients With Cervical Spondylosis

BACKGROUND

Advanced cervical spondylosis (CS) can cause structural damage to the spinal cord resulting in long-term neurological impairment including neck pain and motor weakness. We hypothesized long-term structural reorganization within the brain in patients with CS.

OBJECTIVE

To explore the associations between cortical thickness, subcortical volumes, neurological symptoms, and pain severity in CS patients with or without myelopathy and healthy controls (HCs).

METHODS

High-resolution T1-weighted structural magnetic resonance imaging (MRI) scans from 26 CS patients and 45 HCs were acquired. Cortical thickness and subcortical volumes were computed and compared to the modified Japanese Orthopedic Association (mJOA) and the Neck Disability Index (NDI) scores.

RESULTS

Cortical thinning within the superior frontal gyrus, anterior cingulate, precuneus, and reduction in putamen volume were associated with worsening neurological and pain symptoms. Among the strongest associations were cortical thickness within the left precuneus (R2 = 0.34) and left and right putamen (R2 = 0.43, 0.47, respectively) vs mJOA, and the left precuneus (R2 = 0.55), insula (R2 = 0.57), and right putamen (R2 = 0.54) vs NDI (P ≤ .0001 for all). Cortical thickness along Brodmann areas 3a, 4a, and 4p were also moderately associated with mJOA. Preliminary evidence also suggests that patients with CS may undergo cortical atrophy at a faster rate than HCs.

CONCLUSION

Patients with CS appear to exhibit cortical thinning and atrophy with worsening neurological and pain symptoms in specific brain regions associated with sensorimotor and pain processing.

Deep Learning Detection of Penumbral Tissue on Arterial Spin Labeling in Stroke

Background and Purpose- Selection of patients with acute ischemic stroke for endovascular treatment generally relies on dynamic susceptibility contrast magnetic resonance imaging or computed tomography perfusion. Dynamic susceptibility contrast magnetic resonance imaging requires injection of contrast, whereas computed tomography perfusion requires high doses of ionizing radiation. The purpose of this work was to develop and evaluate a deep learning (DL)-based algorithm for assisting the selection of suitable patients with acute ischemic stroke for endovascular treatment based on 3-dimensional pseudo-continuous arterial spin labeling (pCASL). Methods- A total of 167 image sets of 3-dimensional pCASL data from 137 patients with acute ischemic stroke scanned on 1.5T and 3.0T Siemens MR systems were included for neural network training. The concurrently acquired dynamic susceptibility contrast magnetic resonance imaging was used to produce labels of hypoperfused brain regions, analyzed using commercial software. The DL and 6 machine learning (ML) algorithms were trained with 10-fold cross-validation. The eligibility for endovascular treatment was determined retrospectively based on the criteria of perfusion/diffusion mismatch in the DEFUSE 3 trial (Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke). The trained DL algorithm was further applied on twelve 3-dimensional pCASL data sets acquired on 1.5T and 3T General Electric MR systems, without fine-tuning of parameters. Results- The DL algorithm can predict the dynamic susceptibility contrast-defined hypoperfusion region in pCASL with a voxel-wise area under the curve of 0.958, while the 6 ML algorithms ranged from 0.897 to 0.933. For retrospective determination for subject-level endovascular treatment eligibility, the DL algorithm achieved an accuracy of 92%, with a sensitivity of 0.89 and specificity of 0.95. When applied to the GE pCASL data, the DL algorithm achieved a voxel-wise area under the curve of 0.94 and a subject-level accuracy of 92% for endovascular treatment eligibility. Conclusions- pCASL perfusion magnetic resonance imaging in conjunction with the DL algorithm provides a promising approach for assisting decision-making for endovascular treatment in patients with acute ischemic stroke.