Intelligent noninvasive meningioma grading with a fully automatic segmentation using interpretable multiparametric deep learning.

OBJECTIVES: To establish a robust interpretable multiparametric deep learning (DL) model for automatic noninvasive grading of meningiomas along with segmentation. METHODS: In total, 257 patients with pathologically confirmed meningiomas (162 low-grade, 95 high-grade) who underwent a preoperative brain MRI, including T2-weighted (T2) and contrast-enhanced T1-weighted images (T1C), were included in the institutional training set. A two-stage DL grading model was constructed for segmentation and classification based on multiparametric three-dimensional U-net and ResNet. The models were validated in the external validation set consisting of 61 patients with meningiomas (46 low-grade, 15 high-grade). Relevance-weighted Class Activation Mapping (RCAM) method was used to interpret the DL features contributing to the prediction of the DL grading model. RESULTS: On external validation, the combined T1C and T2 model showed a Dice coefficient of 0.910 in segmentation and the highest performance for meningioma grading compared to the T2 or T1C only models, with an area under the curve (AUC) of 0.770 (95% confidence interval: 0.644-0.895) and accuracy, sensitivity, and specificity of 72.1%, 73.3%, and 71.7%, respectively. The AUC and accuracy of the combined DL grading model were higher than those of the human readers (AUCs of 0.675-0.690 and accuracies of 65.6-68.9%, respectively). The RCAM of the DL grading model showed activated maps at the surface regions of meningiomas indicating that the model recognized the features at the tumor margin for grading. CONCLUSIONS: An interpretable multiparametric DL model combining T1C and T2 can enable fully automatic grading of meningiomas along with segmentation. KEY POINTS: • The multiparametric DL model showed robustness in grading and segmentation on external validation. • The diagnostic performance of the combined DL grading model was higher than that of the human readers. • The RCAM interpreted that DL grading model recognized the meaningful features at the tumor margin for grading.

[Magnetic Resonance Imaging for Diagnosis of Idiopathic Pediatric Meningiomas].

Objective To analyze the radiological features of idiopathic pediatric meningiomas and explore their relationships with pathological grading,misdiagnoses,and blood loss during surgery.Methods We retrospectively reviewed 29 cases of pathologically confirmed pediatric meningiomas with pre-operative magnetic resonance imaging in Beijing Tiantan Hospital from November 2014 to July 2018.We assessed the imaging features to explore their relationships with pathological grading,misdiagnoses,and blood loss during surgery. Results Among the 29 cases,7 intraparenchymal meningiomas,5 extraparenchymal meningiomas,4 ventricular meningiomas,and 1 transcranial meningioma were misdiagnosed.Tumor location was significantly associated with possibility of misdiagnoses(P=0.021),and intraparenchymal tumors were most likely to be misdiagnosed.Twelve patients had positive dural tail sign,and 4 of them were misdiagnosed;16 patients did not have dural tail sign,and 12 of them were misdiagnosed.Fisher exact test showed that positive dural tail sign was associated with decreased possibility of misdiagnoses(one-sided P=0.034).Univariable regression analysis showed that the feature of tumor surrounding arteries or interfering with veins(P=0.020)and the tumor maximum diameter(P=0.001)had positively linear relationships with blood loss volume during surgery.Combining these two variables,the multivariable regression model showed better fitting performance($R_{ad}^2$=0.468).Conclusions Pediatric meningiomas are extremely rare,with scarce radiological characteristics.They are hard to diagnose,and the intraparenchymal meningiomas are very likely to be misdiagnosed and therefore should be treated with extra caution.Among all the radiological features,tumor surrounding arteries or interfering with veins and tumor maximum diameter were associated with increased blood loss during surgery.

Association of apparent diffusion coefficient with Ki-67 proliferation index, progesterone-receptor status and various histopathological parameters, and its utility in predicting the high grade in meningiomas.

BACKGROUND: Accurate preoperative determination of the histological grade and cellular proliferative potential of meningioma by non-invasive imaging is of paramount importance. PURPOSE: To evaluate the utility of apparent diffusion coefficient (ADC) in determining the histological grade of meningioma, and to investigate the correlation of ADC with Ki-67 proliferation index (PI), progesterone receptor (PR) status, and a number of other histopathological parameters. MATERIAL AND METHODS: Histopathologically confirmed 94 meningioma patients (72 low-grade, 22 high-grade) who had undergone preoperative diffusion-weighted imaging were retrospectively evaluated. ADC values were obtained by manually drawing the regions of interest (ROIs) within the solid components of the tumor. The relationship between ADC and Ki-67 values, PR status, and multiple histopathological parameters were investigated, and the ADC values of high-grade and low-grade meningiomas were compared. Independent sample t-test, Mann-Whitney U test, receiver operating characteristic, Pearson correlation, and multiple logistic regression analysis were used for statistical assessment. RESULTS: All ADC and rADC values were significantly lower in high-grade meningiomas than in low-grade meningiomas (all P < 0.05). ADC values showed significantly negative correlations with Ki-67 and mitotic index (P < 0.001 for each). Numerous ADC parameters were significantly lower in meningiomas demonstrating hypercellularity and necrosis features (P < 0.05). ADC values did not show a significant correlation with PR score (all P > 0.05). CONCLUSION: ADC can be utilized as a reliable imaging biomarker for predicting the proliferative potential and histological grade in meningiomas.

Magnetic Resonance Imaging for Diagnosis of Idiopathic Pediatric Meningiomas / 中国医学科学院学报

Objective To analyze the radiological features of idiopathic pediatric meningiomas and explore their relationships with pathological grading,misdiagnoses,and blood loss during surgery.Methods We retrospectively reviewed 29 cases of pathologically confirmed pediatric meningiomas with pre-operative magnetic resonance imaging in Beijing Tiantan Hospital from November 2014 to July 2018.We assessed the imaging features to explore their relationships with pathological grading,misdiagnoses,and blood loss during surgery. Results Among the 29 cases,7 intraparenchymal meningiomas,5 extraparenchymal meningiomas,4 ventricular meningiomas,and 1 transcranial meningioma were misdiagnosed.Tumor location was significantly associated with possibility of misdiagnoses(

Phosphohistone H3 (PHH3) as a surrogate of mitotic figure count for grading in meningiomas: a comparison of PHH3 (S10) versus PHH3 (S28) antibodies.

Mitotic figure (MF) counting is important in the evaluation of meningioma grading. Nevertheless, mitosis assessment on hematoxylin and eosin (H&E)-stained slides may be problematic because of technical factors and pathologist's experience. Phosphohistone H3 (PHH3) is a mitosis-specific antibody that has proven to facilitate mitotic count in various tumors. However, the antibody performance between PHH3 serine10 (S10) and serine28 (S28) has never been compared in these tumors before. In this study, 48 cases of meningioma (28 grade I, 14 grade II, 6 grade III) were evaluated using immunohistochemical stains for four commercially available PHH3 (S10) and S28 antibodies to identify MFs and validate PHH3 intra- and interobserver reproducibility and agreement. Two pathologists counted MFs on both H&E- and PHH3-stained slides. H&E and PHH3 MFs were highly correlated (Spearman's rho = 0.96 for PHH3 (S10)-Biocare, 0.96 for PHH3 (S10)-CST, 0.91 for PHH3 (S28)-Abcam, and 0.89 for PHH3 (S28)-Santa Cruz. The mean difference between an H&E and PHH3 mitotic count is 0.81 for PHH3 (S10)-Biocare, 0.95 for PHH3 (S10)-CST, - 0.97 for PHH3 (S28)-Abcam, and - 0.97 for PHH3 (S28)-Santa Cruz. For comparison among four PHH3 antibodies, PHH3 mitotic counts had both a good intra- and interobserver reproducibility (p > 0.05). Regarding to World Health Organization (WHO) grade, there was not a significant discrepancy in the stratification of tumor grades for all four PHH3 antibodies in terms of interobserver agreement. The Cohen's kappa coefficient (K) was 0.93 for PHH3 (S10)-Biocare, 0.82 for PHH3 (S10)-CST, 0.76 for PHH3 (S28)-Abcam, and 0.80 for PHH3 (S28)-Santa Cruz. Considering survival analyses, all five proliferation indices were univariately associated with recurrences. Increased PHH3 mitotic indices (MIs) were significantly associated with recurrence-free survival in univariate Cox proportional hazards regression analysis (p < 0.001) and remained an independent predictor in multivariate analysis (p < 0.05). The appropriate prognostic cutoff values for recurrence prediction were 5 or more per 10 high-power fields (HPFs) for PHH3 (S10) and 3 or more per 10 HPFs for PHH3 (S28).

rCBV- and ADC-based Grading of Meningiomas With Glimpse Into Emerging Molecular Diagnostics.

INTRODUCTION: This study was conducted to grade meningiomas based on relative Cerebral Blood Volume (rCBV) and Apparent Diffusion Coefficient (ADC) to help surgeons plan the approach and extent of operation as well as decide on the need of any adjuvant radio/chemo therapy. The current and evolving genomic, proteomic, and spectroscopic technologies are also discussed which can supplement the current radiologic methods and procedures in grading meningiomas. METHODS: A total of 35 patients with meningioma prospectively underwent basic MR sequences (T1W, T2W, T2W/FLAIR) in axial, sagittal and coronal planes followed by Diffusion Weighted (DW) imaging having b value of 1000 (minimum ADC values used for analysis). Then, gadobenate dimeglumine/meglumine gadoterate was administered (0.1 mmol/kg at a rate of 4 mL/s) followed by saline flush (20 mL at a rate of 4 mL/s). Next, T2*W/FFE dynamic images were acquired; dynamics showing maximum fall in intensity was used for creating rCBV and relative Cerebral Blood Flow (rCBF) maps and calculating rCBV. RESULTS: Both maximum rCBV and minimum ADC within the tumor were not significant for differentiating benign from malignant meningiomas. A cut-off maximum rCBV of 2.5 mL/100 g in peritumoral edema was 75% sensitive, 84.6% specific, and 83.3% accurate in differentiating benign from malignant meningiomas. CONCLUSION: Benign and malignant meningiomas can be differentiated based on maximum rCBV in peritumoral edema but ADC values within the tumor are insignificant in differentiating benign and malignant tumors. rCBV values within tumor, however, may be helpful in subtyping meningiomas, especially transitional and meningothelial meningiomas.

Clinical application value of MRI combined with DWI sequence in the diagnosis and follow-up of patients with meningioma / 重庆医学

Objective To explore the performance of meningiomas in MRI dynamic enhancement and DWI sequence,and to improve the accuracy of differential diagnosis of benign and malignant meningiomas.Methods Meningioma MRI data,which were pathology proved by the Guilin medical college affiliated hospital at different levels(Ⅰ,Ⅱ and Ⅲ),were retrospectively analyzed.The enhancement of meningiomas at all levels and the extent of edema were summarized,at the meantime the dynamic enhancement perfusion parameters and ADC values of meningioma MRI meningioma were contrastively analyzed.Results There were 30 cases of meningioma located in the brain falcus,27 cases of meningioma located in the sagittal sinus,10 cases of meningioma located in the cerebellopontine angle and 6 cases of meningioma located in other parts.The difference of ADC value between grade Ⅰ meningiomas and grade Ⅲ meningiomas was statistically significant different[(1.253±0.123)×10-3 mm2/s vs.(0.891±0.103)×10-3 mm2/s,P<0.05].The relative blood flow(rCBF)and relative blood volume(rCBV)of grade Ⅰ meningiomas were significantly lower than those of grade Ⅲ meningiomas(P<0.05).Conclusion MRI dynamic enhancement combined with DWI can accurately identify grade Ⅰ meningiomas and grade Ⅲ meningiomas,and it is necessary that the assessment of its classification should be appropriately considered to the higher level for patients who do not have a characteristic manifestation of meningioma in the preoperative.