Glioma: application of whole-tumor texture analysis of diffusion-weighted imaging for the evaluation of tumor heterogeneity

Quantifying intra-tumoral genetic heterogeneity of glioblastoma toward precision medicine using MRI and a data-inclusive machine learning algorithm

BACKGROUND AND OBJECTIVE

Glioblastoma (GBM) is one of the most aggressive and lethal human cancers. Intra-tumoral genetic heterogeneity poses a significant challenge for treatment. Biopsy is invasive, which motivates the development of non-invasive, MRI-based machine learning (ML) models to quantify intra-tumoral genetic heterogeneity for each patient. This capability holds great promise for enabling better therapeutic selection to improve patient outcome.

METHODS

We proposed a novel Weakly Supervised Ordinal Support Vector Machine (WSO-SVM) to predict regional genetic alteration status within each GBM tumor using MRI. WSO-SVM was applied to a unique dataset of 318 image-localized biopsies with spatially matched multiparametric MRI from 74 GBM patients. The model was trained to predict the regional genetic alteration of three GBM driver genes (EGFR, PDGFRA and PTEN) based on features extracted from the corresponding region of five MRI contrast images. For comparison, a variety of existing ML algorithms were also applied. Classification accuracy of each gene were compared between the different algorithms. The SHapley Additive exPlanations (SHAP) method was further applied to compute contribution scores of different contrast images. Finally, the trained WSO-SVM was used to generate prediction maps within the tumoral area of each patient to help visualize the intra-tumoral genetic heterogeneity.

RESULTS

WSO-SVM achieved 0.80 accuracy, 0.79 sensitivity, and 0.81 specificity for classifying EGFR; 0.71 accuracy, 0.70 sensitivity, and 0.72 specificity for classifying PDGFRA; 0.80 accuracy, 0.78 sensitivity, and 0.83 specificity for classifying PTEN; these results significantly outperformed the existing ML algorithms. Using SHAP, we found that the relative contributions of the five contrast images differ between genes, which are consistent with findings in the literature. The prediction maps revealed extensive intra-tumoral region-to-region heterogeneity within each individual tumor in terms of the alteration status of the three genes.

CONCLUSIONS

This study demonstrated the feasibility of using MRI and WSO-SVM to enable non-invasive prediction of intra-tumoral regional genetic alteration for each GBM patient, which can inform future adaptive therapies for individualized oncology.

Radiomics analysis with three-dimensional and two-dimensional segmentation to predict survival outcomes in pancreatic cancer

BACKGROUND

Radiomics can assess prognostic factors in several types of tumors, but considering its prognostic ability in pancreatic cancer has been lacking.

AIM

To evaluate the performance of two different radiomics software in assessing survival outcomes in pancreatic cancer patients.

METHODS

We retrospectively reviewed pretreatment contrast-enhanced dual-energy computed tomography images from 48 patients with biopsy-confirmed pancreatic ductal adenocarcinoma who later underwent neoadjuvant chemoradiation and surgery. Tumors were segmented using TexRad software for 2-dimensional (2D) analysis and MIM software for 3D analysis, followed by radiomic feature extraction. Cox proportional hazard modeling correlated texture features with overall survival (OS) and progression-free survival (PFS). Cox regression was used to detect differences in OS related to pretreatment tumor size and residual tumor following treatment. The Wilcoxon test was used to show the relationship between tumor volume and the percent of residual tumor. Kaplan-Meier analysis was used to compare survival in patients with different tumor densities in Hounsfield units for both 2D and 3D analysis.

RESULTS

3D analysis showed that higher mean tumor density [hazard ratio (HR) = 0.971, P = 0.041)] and higher median tumor density (HR = 0.970, P = 0.037) correlated with better OS. 2D analysis showed that higher mean tumor density (HR = 0.963, P = 0.014) and higher mean positive pixels (HR = 0.962, P = 0.014) correlated with better OS; higher skewness (HR = 3.067, P = 0.008) and higher kurtosis (HR = 1.176, P = 0.029) correlated with worse OS. Higher entropy correlated with better PFS (HR = 0.056, P = 0.036). Models determined that patients with increased tumor size greater than 1.35 cm were likely to have a higher percentage of residual tumors of over 10%.

CONCLUSION

Several radiomics features can be used as prognostic tools for pancreatic cancer. However, results vary between 2D and 3D analyses. Mean tumor density was the only variable that could reliably predict OS, irrespective of the analysis used.

Performance comparison of different medical image fusion algorithms for clinical glioma grade classification with advanced magnetic resonance imaging (MRI)

Non-invasive glioma grade classification is an exciting area in neuroimaging. The primary purpose of this study is to investigate the performance of different medical image fusion algorithms for glioma grading purposes by fusing advanced Magnetic Resonance Imaging (MRI) images. Ninety-six subjects underwent an Apparent diffusion coefficient (ADC) map and Susceptibility-weighted imaging (SWI) MRI scan. After preprocessing, the different medical image fusion methods used to fuse ADC maps and SWI were Principal Component Analysis (PCA), Structure-Aware, Discrete Cosine Harmonic Wavelet Transform (DCHWT), Deep-Convolutional Neural network (DNN), Dual-Discriminator conditional generative adversarial network (DDcGAN), and Laplacian Re-Decomposition (LRD). The Entropy, standard deviation (STD), peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and Relative Signal Contrast (RSC) were calculated for qualitative and quantitative analysis. We found high fused image quality with LRD and DDcGAN methods. Further quantitative analysis showed that RSCs in fused images in Low-Grade glioma (LGG) were significantly higher than RSCs in High-Grade glioma (HGG) with PCA, DCHWT, LRD, and DDcGAN. The Receiver Operating Characteristic (ROC) curve test highlighted that LRD and DDcGAN have the highest performance for glioma grade classification. Our work suggests using the DDcGAN and LRD networks for glioma grade classification by fusing ADC maps and SWI images.

Diagnostic value of MRI radiomics in differentiating high‑grade glioma from low‑grade glioma: A meta‑analysis

No clear conclusions have yet been reached regarding the accuracy of magnetic resonance imaging (MRI) radiomics in distinguishing high-grade glioma (HGG) from low-grade glioma (LGG). In the present study, a meta-analysis was conducted to determine the diagnostic value of MRI radiomics in differentiating between HGG and LGG, in order to guide their clinical diagnosis. PubMed, Embase and the Cochrane Library databases were searched up to November 2022. The search included studies in which true positive, false positive, true negative and false negative values for the differentiation of HGG from LGG were reported or could be calculated by retrograde extrapolation. Duplicate publications, research without full text, studies with incomplete information or unextractable data, animal studies, reviews and systematic reviews were excluded. STATA 15.1 was used to analyze the data. The meta-analysis included 15 studies, which comprised a total of 1,124 patients, of which 701 had HGG and 423 had LGG. The pooled sensitivity and specificity of the studies overall were 0.92 (95% CI: 0.89-0.95) and 0.89 (95% CI: 0.85-0.92), respectively. The positive and negative likelihood ratios of the studies overall were 7.89 (95% CI: 6.01-10.37) and 0.09 (95% CI: 0.07-0.12), respectively. The pooled diagnostic odds ratio of the studies was 85.20 (95% CI: 54.52-133.14). The area under the summary receiver operating characteristic curve was 0.91. These findings indicate that radiomics may be an accurate tool for the differentiation of glioma grades. However, further research is needed to verify the most appropriate of these technologies.

Texture feature analysis of MRI-ADC images to differentiate glioma grades using machine learning techniques

Apparent diffusion coefficient (ADC) of magnetic resonance imaging (MRI) is an indispensable imaging technique in clinical neuroimaging that quantitatively assesses the diffusivity of water molecules within tissues using diffusion-weighted imaging (DWI). This study focuses on developing a robust machine learning (ML) model to predict the aggressiveness of gliomas according to World Health Organization (WHO) grading by analyzing patients' demographics, higher-order moments, and grey level co-occurrence matrix (GLCM) texture features of ADC. A population of 722 labeled MRI-ADC brain image slices from 88 human subjects was selected, where gliomas are labeled as glioblastoma multiforme (WHO-IV), high-grade glioma (WHO-III), and low-grade glioma (WHO I-II). Images were acquired using 3T-MR systems and a region of interest (ROI) was delineated manually over tumor areas. Skewness, kurtosis, and statistical texture features of GLCM (mean, variance, energy, entropy, contrast, homogeneity, correlation, prominence, and shade) were calculated using ADC values within ROI. The ANOVA f-test was utilized to select the best features to train an ML model. The data set was split into training (70%) and testing (30%) sets. The train set was fed into several ML algorithms and selected most promising ML algorithm using K-fold cross-validation. The hyper-parameters of the selected algorithm were optimized using random grid search technique. Finally, the performance of the developed model was assessed by calculating accuracy, precision, recall, and F1 values reported for the test set. According to the ANOVA f-test, three attributes; patient gender (1.48), GLCM energy (9.48), and correlation (13.86) that performed minimum scores were excluded from the dataset. Among the tested algorithms, the random forest classifier(0.8772 ± 0.0237) performed the highest mean-cross-validation score and selected to build the ML model which was able to predict tumor categories with an accuracy of 88.14% over the test set. The study concludes that the developed ML model using the above features except for patient gender, GLCM energy, and correlation, has high prediction accuracy in glioma grading. Therefore, the outcomes of this study enable to development of advanced tumor classification applications that assist in the decision-making process in a real-time clinical environment.

Using Intratumor Heterogeneity of Immunohistochemistry Biomarkers to Classify Laryngeal and Hypopharyngeal Tumors Based on Histologic Features

Haralick texture features are used to quantify the spatial distribution of signal intensities within an image. In this study, the heterogeneity of proliferation (Ki-67 expression) and immune cells (CD45 expression) within tumors was quantified and used to classify histologic characteristics of larynx and hypopharynx carcinomas. Of 21 laryngectomy specimens, 74 whole-mount tumor slides were scored on histologic characteristics. Ki-67 and CD45 immunohistochemistry was performed, and all sections were digitized. The tumor area was annotated in QuPath. Haralick features independent of the diaminobenzidine intensity were extracted from the isolated diaminobenzidine signal to quantify intratumor heterogeneity. Haralick features from both Ki-67 and CD45 were used as input for a principal component analysis. A linear support vector machine was fitted to the first 4 principal components for classification and validated with a leave-one-patient-out cross-validation method. Significant differences in individual Haralick features were found between cohesive and noncohesive tumors for CD45 (angular second motion: P =.03, inverse difference moment: P =.009, and entropy: P =.02) and between the larynx and hypopharynx tumors for both CD45 (angular second motion: P =.03, inverse difference moment: P =.007, and entropy: P =.005) and Ki-67 (correlation: P =.003). Therefore, these features were used for classification. The linear classifier resulted in a classification accuracy of 85% for site of origin and 81% for growth pattern. A leave-one-patient-out cross-validation resulted in an error rate of 0.27 and 0.35 for both classifiers, respectively. In conclusion, we show a method to quantify intratumor heterogeneity of immunohistochemistry biomarkers using Haralick features. This study also shows the feasibility of using these features to classify tumors by histologic characteristics. The classifiers created in this study are a proof of concept because more data are needed to create robust classifiers, but the method shows potential for automated tumor classification.

Three dimensional computed tomography texture analysis of pulmonary lesions: Does radiomics allow differentiation between carcinoma, neuroendocrine tumor and organizing pneumonia?

PURPOSE

To investigate whether CT texture analysis allows differentiation between adenocarcinomas, squamous cell carcinomas, carcinoids, small cell lung cancers and organizing pneumonia and between carcinomas and neuroendocrine tumors.

METHOD

This retrospective study included patients 133 patients (30 patients with organizing pneumonia, 30 patients with adenocarcinoma, 30 patients with squamous cell carcinoma, 23 patients with small cell lung cancer, 20 patients with carcinoid), who underwent CT-guided biopsy of the lung and had a corresponding histopathologic diagnosis. Pulmonary lesions were segmented in consensus by two radiologists with and without a threshold of -50HU in three dimensions. Groupwise comparisons were performed to assess for differences between all five above-listed entities and between carcinomas and neuroendocrine tumors.

RESULTS

Pairwise comparisons of the five entities revealed 53 statistically significant texture features when using no HU-threshold and 6 statistically significant features with a threshold of -50HU. The largest AUC (0.818 [95%CI 0.706-0.930]) was found for the feature wavelet-HHH_glszm_SmallAreaEmphasis for discrimination of carcinoid from the other entities when using no HU-threshold. In differentiating neuroendocrine tumors from carcinomas, 173 parameters proved statistically significant when using no HU threshold versus 52 parameters when using a -50HU-threshold. The largest AUC (0.810 [95%CI 0.728-0,893]) was found for the parameter original_glcm_Correlation for discrimination of neuroendocrine tumors from carcinomas when using no HU-threshold.

CONCLUSIONS

CT texture analysis revealed features that differed significantly between malignant pulmonary lesions and organizing pneumonia and between carcinomas and neuroendocrine tumors of the lung. Applying a HU-threshold for segmentation substantially influenced the results of texture analysis.

Histogram analysis of dynamic contrast-enhanced magnetic resonance imaging to predict extramural venous invasion in rectal cancer

BACKGROUND

To explore the potential of histogram analysis (HA) of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in the identification of extramural venous invasion (EMVI) in rectal cancer patients.

METHODS

This retrospective study included preoperative images of 194 rectal cancer patients at our hospital between May 2019 and April 2022. The postoperative histopathological examination served as the reference standard. The mean values of DCE-MRI quantitative perfusion parameters (Ktrans, Kep and Ve) and other HA features calculated from these parameters were compared between the pathological EMVI-positive and EMVI-negative groups. Multivariate logistic regression analysis was performed to establish the prediction model for pathological EMVI-positive status. Diagnostic performance was assessed and compared using the receiver operating characteristic (ROC) curve. The clinical usefulness of the best prediction model was further measured with patients with indeterminate MRI-defined EMVI (mrEMVI) score 2(possibly negative) and score 3 (probably positive).

RESULTS

The mean values of Ktrans and Ve in the EMVI-positive group were significantly higher than those in the EMVI-negative group (P = 0.013 and 0.025, respectively). Significant differences in Ktrans skewness, Ktrans entropy, Ktrans kurtosis, and Ve maximum were observed between the two groups (P = 0.001,0.002, 0.000, and 0.033, respectively). The Ktrans kurtosis and Ktrans entropy were identified as independent predictors for pathological EMVI. The combined prediction model had the highest area under the curve (AUC) at 0.926 for predicting pathological EMVI status and further reached the AUC of 0.867 in subpopulations with indeterminate mrEMVI scores.

CONCLUSIONS

Histogram Analysis of DCE-MRI Ktrans maps may be useful in preoperative identification of EMVI in rectal cancer, particularly in patients with indeterminate mrEMVI scores.

CT texture analysis reliability in pulmonary lesions: the influence of 3D vs. 2D lesion segmentation and volume definition by a Hounsfield-unit threshold

OBJECTIVE

Reproducibility problems are a known limitation of radiomics. The segmentation of the target lesion plays a critical role in texture analysis variability. This study's aim was to compare the interobserver reliability of manual 2D vs. 3D lung lesion segmentation with and without pre-definition of the volume using a threshold of - 50 HU.

METHODS

Seventy-five patients with histopathologically proven lung lesions (15 patients each with adenocarcinoma, squamous cell carcinoma, small cell lung cancer, carcinoid, and organizing pneumonia) who underwent an unenhanced CT scan of the chest were included. Three radiologists independently segmented each lesion manually in 3D and 2D with and without pre-segmentation volume definition by a HU threshold, and shape parameters and original, Laplacian of Gaussian-filtered, and wavelet-based texture features were derived. To assess interobserver reliability and identify the most robust texture features, intraclass correlation coefficients (ICCs) for different segmentation settings were calculated.

RESULTS

Shape parameters had high reliability (64-79% had excellent and good ICCs). Texture features had weak reliability levels, with the highest ICCs (38% excellent or good) found for original features in 3D segmentation without the use of a HU threshold. A small proportion (4.3-11.5%) of texture features had excellent or good ICC values at all segmentation settings.

CONCLUSION

Interobserver reliability of texture features from CT scans of a heterogeneous collection of manually segmented lung lesions was low with a small proportion of features demonstrating high reliability independent of the segmentation settings. These results indicate a limited applicability of texture analysis and the need to define robust texture features in patients with lung lesions.

KEY POINTS

• Our study showed a low reproducibility of texture features when 3 radiologists independently segmented lung lesions in CT images, which highlights a serious limitation of texture analysis. • Interobserver reliability of texture features was low regardless of whether the lesion was segmented in 2D and 3D with or without a HU threshold. • In contrast to texture features, shape parameters showed a high interobserver reliability when lesions were segmented in 2D vs. 3D with and without a HU threshold of - 50.

Grading the Severity of Damage to the Perijunctional Actomyosin Ring and Zonula Occludens-1 of the Corneal Endothelium by Ensemble Learning Methods

Purpose: In many epithelia, including the corneal endothelium, intracellular/extracellular stresses break down the perijunctional actomyosin ring (PAMR) and zonula occludens-1 (ZO-1) at the apical junctions. This study aims to grade the severity of damage to PAMR and ZO-1 through machine learning. Methods: Immunocytochemical images of PAMR and ZO-1 were drawn from recent studies on the corneal endothelium subjected to hypothermia and oxidative stress. The images were analyzed for their morphological (e.g., Hu moments) and textural features (based on gray-level co-occurrence matrix [GLCM] and Gabor filters). The extracted features were ranked by SHapley analysis and analysis of variance. Then top features were used to grade the severity of damage using a suite of ensemble classifiers, including random forest, bagging classifier (BC), AdaBoost, extreme gradient boosting, and stacking classifier. Results: A partial set of features from GLCM, along with Hu moments and the number of hexagons, enabled the classification of damage to PAMR into Control, Mild, Moderate, and Severe with the area under the receiver operating characteristics curve (AUC) = 0.92 and F1 score = 0.77 with BC. In contrast, a bank of Gabor filters provided a partial set of features that could be combined with Hu moments, branch length, and sharpness for the classification of ZO-1 images into four levels with AUC = 0.95 and F1 score of 0.8 with BC. Conclusions: We have developed a workflow that enables the stratification of damage to PAMR and ZO-1. The approach can be applied to similar data during drug discovery or pathophysiological studies of epithelia.

The value of convolutional neural networks-based deep learning model in differential diagnosis of space-occupying brain diseases

Objectives

It is still a challenge to differentiate space-occupying brain lesions such as tumefactive demyelinating lesions (TDLs), tumefactive primary angiitis of the central nervous system (TPACNS), primary central nervous system lymphoma (PCNSL), and brain gliomas. Convolutional neural networks (CNNs) have been used to analyze complex medical data and have proven transformative for image-based applications. It can quickly acquire diseases' radiographic features and correct doctors' diagnostic bias to improve diagnostic efficiency and accuracy. The study aimed to assess the value of CNN-based deep learning model in the differential diagnosis of space-occupying brain diseases on MRI.

Methods

We retrospectively analyzed clinical and MRI data from 480 patients with TDLs (n = 116), TPACNS (n = 64), PCNSL (n = 150), and brain gliomas (n = 150). The patients were randomly assigned to training (n = 240), testing (n = 73), calibration (n = 96), and validation (n = 71) groups. And a CNN-implemented deep learning model guided by clinical experts was developed to identify the contrast-enhanced T1-weighted sequence lesions of these four diseases. We utilized accuracy, sensitivity, specificity, and area under the curve (AUC) to evaluate the performance of the CNN model. The model's performance was then compared to the neuroradiologists' diagnosis.

Results

The CNN model had a total accuracy of 87% which was higher than senior neuroradiologists (74%), and the AUC of TDLs, PCNSL, TPACNS and gliomas were 0.92, 0.92, 0.89 and 0.88, respectively.

Conclusion

The CNN model can accurately identify specific radiographic features of TDLs, TPACNS, PCNSL, and gliomas. It has the potential to be an effective auxiliary diagnostic tool in the clinic, assisting inexperienced clinicians in reducing diagnostic bias and improving diagnostic efficiency.

Prediction of Ki-67 of Invasive Ductal Breast Cancer Based on Ultrasound Radiomics Nomogram

PURPOSE

The objective of this research was to develop and validate an ultrasound-based radiomics nomogram for the pre-operative assessment of Ki-67 in breast cancer (BC).

MATERIALS AND METHODS

From December 2016 to December 2018, 515 patients with invasive ductal breast cancer who received two-dimensional (2D) ultrasound and Ki-67 examination were studied and analyzed retrospectively. The dataset was distributed at random into a training cohort (n = 360) and a test cohort (n = 155) in the ratio of 7:3. Each tumor region of interest was defined based on 2D ultrasound images and radiomics features were extracted. ANOVA, maximum correlation minimum redundancy (mRMR) algorithm, and minimum absolute shrinkage and selection operator (LASSO) were performed to pick features, and independent clinical predictors were integrated with radscore to construct the nomogram for predicting Ki-67 index by univariate and multivariate logistic regression analysis. The performance and utility of the models were evaluated by plotting receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration curves.

RESULTS

In the testing cohort, the area under the receiver characteristic curve (AUC) of the nomogram was 0.770 (95% confidence interval, 0.690-0.860). In both cohorts, the nomogram outperformed both the clinical model and the radiomics model (P < .05 according to the DeLong test). The analysis of DCA proved that the model has clinical utility.

CONCLUSIONS

The nomogram based on 2D ultrasound images offered an approach for predicting Ki-67 in BC.

Novel 3D magnetic resonance fingerprinting radiomics in adult brain tumors: a feasibility study

OBJECTIVES

To test the feasibility of using 3D MRF maps with radiomics analysis and machine learning in the characterization of adult brain intra-axial neoplasms.

METHODS

3D MRF acquisition was performed on 78 patients with newly diagnosed brain tumors including 33 glioblastomas (grade IV), 6 grade III gliomas, 12 grade II gliomas, and 27 patients with brain metastases. Regions of enhancing tumor, non-enhancing tumor, and peritumoral edema were segmented and radiomics analysis with gray-level co-occurrence matrices and gray-level run-length matrices was performed. Statistical analysis was performed to identify features capable of differentiating tumors based on type, grade, and isocitrate dehydrogenase (IDH1) status. Receiver operating curve analysis was performed and the area under the curve (AUC) was calculated for tumor classification and grading. For gliomas, Kaplan-Meier analysis for overall survival was performed using MRF T1 features from enhancing tumor region.

RESULTS

Multiple MRF T1 and T2 features from enhancing tumor region were capable of differentiating glioblastomas from brain metastases. Although no differences were identified between grade 2 and grade 3 gliomas, differentiation between grade 2 and grade 4 gliomas as well as between grade 3 and grade 4 gliomas was achieved. MRF radiomics features were also able to differentiate IDH1 mutant from the wild-type gliomas. Radiomics T1 features for enhancing tumor region in gliomas correlated to overall survival (p < 0.05).

CONCLUSION

Radiomics analysis of 3D MRF maps allows differentiating glioblastomas from metastases and is capable of differentiating glioblastomas from metastases and characterizing gliomas based on grade, IDH1 status, and survival.

KEY POINTS

• 3D MRF data analysis using radiomics offers novel tissue characterization of brain tumors. • 3D MRF with radiomics offers glioma characterization based on grade, IDH1 status, and overall patient survival.

Diagnostic performance of gliomas grading and IDH status decoding A comparison between 3D amide proton transfer APT and four diffusion-weighted MRI models

BACKGROUND

The focus of neuro-oncology research has changed from histopathologic grading to molecular characteristics, and medical imaging routinely follows this change.

PURPOSE

To compare the diagnostic performance of amide proton transfer (APT) and four diffusion models in gliomas grading and isocitrate dehydrogenase (IDH) genotype.

STUDY TYPE

Prospective.

POPULATION

A total of 62 participants (37 males, 25 females; mean age, 52 ± 13 years) whose IDH genotypes were mutant in 6 of 14 grade II gliomas, 8 of 20 of grade III gliomas, and 4 of 28 grade IV gliomas.

FIELD STRENGTH/SEQUENCE

APT imaging using sampling perfection with application optimized contrasts by using different flip angle evolutions (SPACE) and DWI with q-space Cartesian grid sampling were acquired at 3 T.

ASSESSMENT

The ability of diffusion kurtosis imaging, diffusion kurtosis imaging, neurite orientation dispersion and density imaging (NODDI), mean apparent propagator (MAP), and APT imaging for glioma grade and IDH status were assessed, with histopathological grade and genetic testing used as a reference standard. Regions of interest (ROIs) were drawn by two neuroradiologists after consensus.

STATISTICAL TESTS

T-test and Mann-Whitney U test; one-way analysis of variance (ANOVA); receiver operating curve (ROC) and area under the curve (AUC); DeLong test. P value < 0.05 was considered statistically significant.

RESULTS

Compared with IDH-mutant gliomas, IDH-wildtype gliomas showed a significantly higher mean, 5th-percentile (APT₅ ), and 95th-percentile from APTw, the 95th-percentile value of axial, mean, and radial diffusivity from DKI, and 95th-percentile value of isotropic volume fraction from NODDI, and no significantly different parameters from DTI and MAP (P = 0.075-0.998). The combined APT model showed a significantly wider area under the curve (AUC 0.870) for IDH status, when compared with DKI and NODDI. APT₅ was significantly different between two of the three groups (glioma II vs. glioma III vs. glioma IV: 1.35 ± 0.75 vs. 2.09 ± 0.93 vs. 2.71 ± 0.81).

DATA CONCLUSION

APT has higher diagnostic accuracy than DTI, DKI, MAP, and NODDI in glioma IDH genotype. APT₅ can effectively identify both tumor grading and IDH genotyping, making it a promising biomarker for glioma classification.

EVIDENCE LEVEL

1 TECHNICAL EFFICACY: Stage 2.

Assessment of Early Response to Lung Cancer Chemotherapy by Semiquantitative Analysis of Dynamic Contrast-Enhanced MRI

Objective

To evaluate the early chemotherapy response in patients with lung cancer using semiquantitative analysis of dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI).

Methods

Twenty-two patients with lung cancer treated with chemotherapy were subjected to DCE-MRI at two time points: before starting treatment and after one week of therapy. The image data were collected by DCE-MRI, and the semiquantitative parameters including positive enhancement integral (PEI), signal enhancement ratio (SER), maximum slope of increase (MSI), and time to peak (TTP) were calculated. After chemotherapy, the parameters and relevant variations between the responders and nonresponders were compared with Mann–Whitney U tests. Student's t-test for paired samples was used to evaluate the temporal changes between pre- and posttreatment images.

Results

The patients were categorized as 13 responders and 9 nonresponders based on the tumor response evaluation. After chemotherapy, the PEI, SER, and MSI were significantly increased in responders compared with the pretreatment values (P < 0.05), while no obvious decrease in TTP was observed (P > 0.05). However, 9 nonresponders showed no significant changes in PEI, SER, MSI, and TTP values, as compared with those of pretreatment (P > 0.05). Moreover, the increase of PEI was more dramatically in responders than in nonresponders (P < 0.05), but no significantly differences were observed in SER, MSI, and TTP (P > 0.05).

Conclusion

Semiquantitative analysis of DCE-MRI could provide a reliable noninvasive method for assessing early chemotherapy response in lung cancer patients.

ADC textural features in patients with single brain metastases improve clinical risk models

AIMS

In this retrospective study we performed a quantitative textural analysis of apparant diffusion coefficient (ADC) images derived from diffusion weighted MRI (DW-MRI) of single brain metastases (BM) patients from different primary tumors and tested whether these imaging parameters may improve established clinical risk models.

METHODS

We identified 87 patients with single BM who had a DW-MRI at initial diagnosis. Applying image segmentation, volumes of contrast-enhanced lesions in T1 sequences, hyperintense T2 lesions (peritumoral border zone (T2PZ)) and tumor-free gray and white matter compartment (GMWMC) were generated and registered to corresponding ADC maps. ADC textural parameters were generated and a linear backward regression model was applied selecting imaging features in association with survival. A cox proportional hazard model with backward regression was fitted for the clinical prognostic models (diagnosis-specific graded prognostic assessment score (DS-GPA) and the recursive partitioning analysis (RPA)) including these imaging features.

RESULTS

Thirty ADC textural parameters were generated and linear backward regression identified eight independent imaging parameters which in combination predicted survival. Five ADC texture features derived from T2PZ, the volume of the T2PZ, the normalized mean ADC of the GMWMC as well as the mean ADC slope of T2PZ. A cox backward regression including the DS-GPA, RPA and these eight parameters identified two MRI features which improved the two risk scores (HR = 1.14 [1.05;1.24] for normalized mean ADC GMWMC and HR = 0.87 [0.77;0.97]) for ADC 3D kurtosis of the T2PZ.) CONCLUSIONS: Textural analysis of ADC maps in patients with single brain metastases improved established clinical risk models. These findings may aid to better understand the pathogenesis of BM and may allow selection of patients for new treatment options.

Texture appearance model, a new model-based segmentation paradigm, application on the segmentation of lung nodule in the CT scan of the chest

Lung cancer causes more than one million deaths worldwide each year. Averages of 5-year survival rate of patients with Non-small cell lung cancer (NSCLC), which is the most common type of lung cancer, is 15%. Computer-Aided Detection (CAD) is a very important tool for identifying lung lesions in medical imaging. In general, the process line of a CAD system can be divided into four main stages: preprocessing, localization, feature extraction, and classification. As segmentation is required for localization in computer vision and medical image analysis, this step has become a major and challenging problem, and much research has been done on new segmentation techniques. In recent years, interest in model-based segmentation methods has increased, and the reason for this is even if some object information is lost, such gaps can be filled by using the previous information in the model. This paper proposed Texture Appearance Model (TAM), which is a new model-based method and segments all types of nodule areas accurately and efficiently, including juxta-pleural nodules, without separating the lung from the surrounding area in a CT scan of the lung. In this method, Texture Representation of Image (TRI) is obtained using tissue texture feature extraction and feature selection algorithms. The proposed method has been evaluated in 85 nodules of the dataset, received from the Iranian hospital, in which the ground-truth annotation by physicians and CT imaging data were provided. The results show that the proposed algorithm has an encouraging performance for distinguishing different types of nodules, including pleural, cavity and non-solid nodules, achieving an average dice similarity coefficient (DSC) of 84.75%.

Development of computer-aided model to differentiate COVID-19 from pulmonary edema in lung CT scan: EDECOVID-net

The efforts made to prevent the spread of COVID-19 face specific challenges in diagnosing COVID-19 patients and differentiating them from patients with pulmonary edema. Although systemically administered pulmonary vasodilators and acetazolamide are of great benefit for treating pulmonary edema, they should not be used to treat COVID-19 as they carry the risk of several adverse consequences, including worsening the matching of ventilation and perfusion, impaired carbon dioxide transport, systemic hypotension, and increased work of breathing. This study proposes a machine learning-based method (EDECOVID-net) that automatically differentiates the COVID-19 symptoms from pulmonary edema in lung CT scans using radiomic features. To the best of our knowledge, EDECOVID-net is the first method to differentiate COVID-19 from pulmonary edema and a helpful tool for diagnosing COVID-19 at early stages. The EDECOVID-net has been proposed as a new machine learning-based method with some advantages, such as having simple structure and few mathematical calculations. In total, 13 717 imaging patches, including 5759 COVID-19 and 7958 edema images, were extracted using a CT incision by a specialist radiologist. The EDECOVID-net can distinguish the patients with COVID-19 from those with pulmonary edema with an accuracy of 0.98. In addition, the accuracy of the EDECOVID-net algorithm is compared with other machine learning methods, such as VGG-16 (Acc = 0.94), VGG-19 (Acc = 0.96), Xception (Acc = 0.95), ResNet101 (Acc = 0.97), and DenseNet20l (Acc = 0.97).

Prediction of Pancreatic Neuroendocrine Tumor Grading Risk Based on Quantitative Radiomic Analysis of MR

Background

Pancreatic neuroendocrine tumors (PNETs) grade is very important for treatment strategy of PNETs. The present study aimed to find the quantitative radiomic features for predicting grades of PNETs in MR images.

Materials and Methods

Totally 48 patients but 51 lesions with a pathological tumor grade were subdivided into low grade (G1) group and intermediate grade (G2) group. The ROI was manually segmented slice by slice in 3D-T1 weighted sequence with and without enhancement. Statistical differences of radiomic features between G1 and G2 groups were analyzed using the independent sample t-test. Logistic regression analysis was conducted to find better predictors in distinguishing G1 and G2 groups. Finally, receiver operating characteristic (ROC) was constructed to assess diagnostic performance of each model.

Results

No significant difference between G1 and G2 groups (P > 0.05) in non-enhanced 3D-T1 images was found. Significant differences in the arterial phase analysis between the G1 and the G2 groups appeared as follows: the maximum intensity feature (P = 0.021); the range feature (P = 0.039). Multiple logistic regression analysis based on univariable model showed the maximum intensity feature (P=0.023, OR = 0.621, 95% CI: 0.433-0.858) was an independent predictor of G1 compared with G2 group, and the area under the curve (AUC) was 0.695.

Conclusions

The maximum intensity feature of radiomic features in MR images can help to predict PNETs grade risk.

Grading of meningioma tumors based on analyzing tumor volumetric histograms obtained from conventional MRI and apparent diffusion coefficient images

Background

Our purpose was to evaluate the application of volumetric histogram parameters obtained from conventional MRI and apparent diffusion coefficient (ADC) images for grading the meningioma tumors.

Results

Tumor volumetric histograms of preoperative MRI images from 45 patients with the diagnosis of meningioma at different grades were analyzed to find the histogram parameters. Kruskal-Wallis statistical test was used for comparison between the parameters obtained from different grades. Multi-parametric regression analysis was used to find the model and parameters with high predictive value for the classification of meningioma. Mode; standard deviation on post-contrast T1WI, T2-FLAIR, and ADC images; kurtosis on post-contrast T1WI and T2-FLAIR images; mean and several percentile values on ADC; and post-contrast T1WI images showed significant differences among different tumor grades (P < 0.05). The multi-parametric linear regression showed that the ADC histogram parameters model had a higher predictive value, with cutoff values of 0.212 (sensitivity = 79.6%, specificity = 84.3%) and 0.180 (sensitivity = 70.9%, specificity = 80.8%) for differentiating the grade I from II, and grade II from III, respectively.

Conclusions

The multi-parametric model of volumetric histogram parameters in some of the conventional MRI series (i.e., post-contrast T1WI and T2-FLAIR images) along with the ADC images are appropriate for predicting the meningioma tumors’ grade.

Using of Laplacian Re-decomposition image fusion algorithm for glioma grading with SWI, ADC, and FLAIR images

Introduction: Based on the tumor’s growth potential and aggressiveness, glioma is most often classified into low or high-grade groups. Traditionally, tissue sampling is used to determine the glioma grade. The aim of this study is to evaluate the efficiency of the Laplacian Re-decomposition (LRD) medical image fusion algorithm for glioma grading by advanced magnetic resonance imaging (MRI) images and introduce the best image combination for glioma grading.

Material and methods: Sixty-one patients (17 low-grade and 44 high-grade) underwent Susceptibility-weighted image (SWI), apparent diffusion coefficient (ADC) map, and Fluid attenuated inversion recovery (FLAIR) MRI imaging. To fuse different MRI image, LRD medical image fusion algorithm was used. To evaluate the effectiveness of LRD in the classification of glioma grade, we compared the parameters of the receiver operating characteristic curve (ROC).

Results: The average Relative Signal Contrast (RSC) of SWI and ADC maps in high-grade glioma are significantly lower than RSCs in low-grade glioma. No significant difference was detected between low and high-grade glioma on FLAIR images. In our study, the area under the curve (AUC) for low and high-grade glioma differentiation on SWI and ADC maps were calculated at 0.871 and 0.833, respectively.

Conclusions: By fusing SWI and ADC map with LRD medical image fusion algorithm, we can increase AUC for low and high-grade glioma separation to 0.978. Our work has led us to conclude that, by fusing SWI and ADC map with LRD medical image fusion algorithm, we reach the highest diagnostic accuracy for low and high-grade glioma differentiation and we can use LRD medical fusion algorithm for glioma grading.

Deep Convolutional Radiomic Features on Diffusion Tensor Images for Classification of Glioma Grades

The grading of glioma has clinical significance in determining a treatment strategy and evaluating prognosis to investigate a novel set of radiomic features extracted from the fractional anisotropy (FA) and mean diffusivity (MD) maps of brain diffusion tensor imaging (DTI) sequences for computer-aided grading of gliomas. This retrospective study included 108 patients who had pathologically confirmed brain gliomas and DTI scanned during 2012-2018. This cohort included 43 low-grade gliomas (LGGs; all grade II) and 65 high-grade gliomas (HGGs; grade III or IV). We extracted a set of radiomic features, including traditional texture, morphological, and novel deep features derived from pre-trained convolutional neural network models, in the manually-delineated tumor regions. We employed support vector machine and these radiomic features for two classification tasks: LGGs vs HGGs, and grade III vs IV. The area under the receiver operating characteristic (ROC) curve (AUC), accuracy, sensitivity, and specificity was reported as the performance metrics using the leave-one-out cross-validation method. When combining FA+MD, AUC = 0.93, accuracy = 0.94, sensitivity = 0.98, and specificity = 0.86 in classifying LGGs from HGGs, while AUC = 0.99, accuracy = 0.98, sensitivity = 0.98, and specificity = 1.00 in classifying grade III from IV. The AUC and accuracy remain close when features were extracted from only the solid tumor or additionally including necrosis, cyst, and peritumoral edema. Still, the effects in terms of sensitivity and specificity are mixed. Deep radiomic features derived from pre-trained convolutional neural networks showed higher prediction ability than the traditional texture and shape features in both classification experiments. Radiomic features extracted on the FA and MD maps of brain DTI images are useful for noninvasively classification/grading of LGGs vs HGGs, and grade III vs IV.

Magnetic Resonance Relaxometry for Tumor Cell Density Imaging for Glioma: An Exploratory Study via 11C-Methionine PET and Its Validation via Stereotactic Tissue Sampling

Simple Summary

To test the hypothesis that quantitative magnetic resonance relaxometry reflects glioma tumor load within tissue and that it can be an imaging surrogate for visualizing non-contrast-enhancing tumors, we investigated the correlation between T1- and T2-weighted relaxation times, apparent diffusion coefficient (ADC) on magnetic resonance imaging, and ¹¹C-methionine (MET) on positron emission tomography (PET). Moreover, we compared T1- and T2-relaxation times and ADC with tumor cell density (TCD) findings obtained via stereotactic image-guided tissue sampling. A T1-relaxation time of >1850 ms but <3200 ms or a T2-relaxation time of >115 ms but <225 ms under 3 T indicated high MET uptake. The stereotactic tissue sampling findings confirmed that the T1-relaxation time of 1850–3200 ms significantly indicated higher TCD while the T2-relaxation time and ADC did not significantly correlate with the stereotactic tissue sampling findings. However, synthetically synthesized tumor load images from the T1- and T2-relaxation maps were able to visualize MET uptake presented on PET.

Abstract

One of the most crucial yet challenging issues for glioma patient care is visualizing non-contrast-enhancing tumor regions. In this study, to test the hypothesis that quantitative magnetic resonance relaxometry reflects glioma tumor load within tissue and that it can be an imaging surrogate for visualizing non-contrast-enhancing tumors, we investigated the correlation between T1- and T2-weighted relaxation times, apparent diffusion coefficient (ADC) on magnetic resonance imaging, and ¹¹C-methionine (MET) on positron emission tomography (PET). Moreover, we compared the T1- and T2-relaxation times and ADC with tumor cell density (TCD) findings obtained via stereotactic image-guided tissue sampling. Regions that presented a T1-relaxation time of >1850 ms but <3200 ms or a T2-relaxation time of >115 ms but <225 ms under 3 T indicated a high MET uptake. In addition, the stereotactic tissue sampling findings confirmed that the T1-relaxation time of 1850–3200 ms significantly indicated a higher TCD (p = 0.04). However, ADC was unable to show a significant correlation with MET uptake or with TCD. Finally, synthetically synthesized tumor load images from the T1- and T2-relaxation maps were able to visualize MET uptake presented on PET.

Radiomics-Based Machine Learning Classification for Glioma Grading Using Diffusion- and Perfusion-Weighted Magnetic Resonance Imaging

OBJECTIVE

The aim of this study was to evaluate various radiomics-based machine learning classification models using the apparent diffusion coefficient (ADC) and cerebral blood flow (CBF) maps for differentiating between low-grade gliomas (LGGs) and high-grade gliomas (HGGs).

METHODS

Fifty-two glioma patients, including 18 LGGs (grade II) and 34 HGGs (grade III/IV), were examined using a 3.0-T magnetic resonance scanner. The ADC and CBF maps were obtained from diffusion-weighted imaging and pseudo-continuous arterial spin labeling perfusion-weighted imaging, respectively. A total of 91 radiomic features were extracted from each of the tumor volume on the ADC and CBF maps. We constructed 4 types of machine learning classifiers based on (1) least absolute shrinkage and selection operator regularized logistic regression (LASSO-LR), (2) random forest (RF), (3) support vector machine (SVM) with the radial basis function kernel (SVM-RBF), and (4) SVM with the linear kernel (SVM-L). A training set with 36 gliomas (70%) was used to select the important radiomic features and train each model using 5-fold cross-validation. The remaining 16 gliomas (30%) were used as a test set. Receiver operating characteristic analysis was performed to evaluate the model performance.

RESULTS

A radiomic feature, ADC first-order-based skewness, was selected as an important variable in all classification models. According to the receiver operating characteristic analysis, the areas under the curve of the LASSO-LR, RF, SVM-RBF, and SVM-L models for the training set were 0.965, 1.000, 0.979, and 0.969, respectively. For the test set, the areas under the curve of the LASSO-LR, RF, SVM-RBF, and SVM-L models were 0.883, 0.917, 0.717, and 0.917, respectively. All classification models showed sufficient diagnostic performance on the test set.

CONCLUSIONS

Radiomics-based machine learning classifiers using the quantitative ADC and CBF maps are useful for differentiating HGGs from LGGs.

Texture Analysis of DCE-MRI Intratumoral Subregions to Identify Benign and Malignant Breast Tumors

Purpose

To evaluate the potential of the texture features extracted from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) intratumoral subregions to distinguish benign from malignant breast tumors.

Materials and Methods

A total of 299 patients with pathologically verified breast tumors who underwent breast DCE-MRI examination were enrolled in this study, including 124 benign cases and 175 malignant cases. The whole tumor area was semi-automatically segmented on the basis of subtraction images of DCE-MRI in Matlab 2018b. According to the time to peak of the contrast agent, the whole tumor area was partitioned into three subregions: early, moderate, and late. A total of 467 texture features were extracted from the whole tumor area and the three subregions, respectively. Patients were divided into training (n = 209) and validation (n = 90) cohorts by different MRI scanners. The least absolute shrinkage and selection operator (LASSO) method was used to select the optimal feature subset in the training cohort. The Kolmogorov-Smirnov test was first performed on texture features selected by LASSO to test whether the samples followed a normal distribution. Two machine learning methods, decision tree (DT) and support vector machine (SVM), were used to establish classification models with a 10-fold cross-validation method. The performance of the classification models was evaluated with receiver operating characteristic (ROC) curves.

Results

In the training cohort, the areas under the ROC curve (AUCs) for the DT_Whole model and SVM_Whole model were 0.744 and 0.806, respectively. In contrast, the AUCs of the DT_Early model (P = 0.004), DT_Late model (P = 0.015), SVM_Early model (P = 0.002), and SVM_Late model (P = 0.002) were significantly higher: 0.863 (95% CI, 0.808-0.906), 0.860 (95% CI, 0.806-0.904), 0.934 (95% CI, 0.891-0.963), and 0.921 (95% CI, 0.876-0.954), respectively. The SVM_Early model and SVM_Late model achieved better performance than the DT_Early model and DT_Late model (P = 0.003, 0.034, 0.008, and 0.026, respectively). In the validation cohort, the AUCs for the DT_Whole model and SVM_Whole model were 0.670 and 0.708, respectively. In comparison, the AUCs of the DT_Early model (P = 0.006), DT_Late model (P = 0.043), SVM_Early model (P = 0.001), and SVM_Late model (P = 0.007) were significantly higher: 0.839 (95% CI, 0.747-0.908), 0.784 (95% CI, 0.601-0.798), 0.890 (95% CI, 0.806-0.946), and 0.865 (95% CI, 0.777-0.928), respectively.

Conclusion

The texture features from intratumoral subregions of breast DCE-MRI showed potential in identifying benign and malignant breast tumors.

Performance optimisation of deep learning models using majority voting algorithm for brain tumour classification

BACKGROUND

Although biopsy is the gold standard for tumour grading, being invasive, this procedure also proves fatal to the brain. Thus, non-invasive methods for brain tumour grading are urgently needed. Here, a magnetic resonance imaging (MRI)-based non-invasive brain tumour grading method has been proposed using deep learning (DL) and machine learning (ML) techniques.

METHOD

Four clinically applicable datasets were designed. The four datasets were trained and tested on five DL-based models (convolutional neural networks), AlexNet, VGG16, ResNet18, GoogleNet, and ResNet50, and five ML-based models, Support Vector Machine, K-Nearest Neighbours, Naïve Bayes, Decision Tree, and Linear Discrimination using five-fold cross-validation. A majority voting (MajVot)-based ensemble algorithm has been proposed to optimise the overall classification performance of five DL and five ML-based models.

RESULTS

The average accuracy improvement of four datasets using the DL-based MajVot algorithm against AlexNet, VGG16, ResNet18, GoogleNet, and ResNet50 models was 2.02%, 1.11%, 1.04%, 2.67%, and 1.65%, respectively. Further, a 10.12% improvement was seen in the average accuracy of four datasets using the DL method against ML. Furthermore, the proposed DL-based MajVot algorithm was validated on synthetic face data and improved the male versus female face image classification accuracy by 2.88%, 0.71%, 1.90%, 2.24%, and 0.35% against AlexNet, VGG16, ResNet18, GoogleNet, and ResNet50, respectively.

CONCLUSION

The proposed MajVot algorithm achieved promising results for brain tumour classification and is able to utilise the combined potential of multiple models.

Multispectral co-occurrence of wavelet coefficients for malignancy assessment of brain tumors

Brain tumor is not most common, but truculent type of cancer. Therefore, correct prediction of its aggressiveness nature at an early stage would influence the treatment strategy. Although several diagnostic methods based on different modalities exist, a pre-operative method for determining tumor malignancy state still remains as an active research area. In this regard, the paper presents a new method for the assessment of tumor grades using conventional MR sequences namely, T1, T1 with contrast enhancement, T2 and FLAIR. The proposed method for tumor gradation is mainly based on feature extraction using multiresolution image analysis and classification using support vector machine. Since the wavelet features of different tumor subregions, obtained from single MR sequence, do not carry equally important information, a wavelet fusion technique is proposed based on the texture information content of each voxel. The concept of texture gradient, used in the proposed algorithm, fuses the wavelet coefficients of the given MR sequences. The feature vector is then derived from the co-occurrence of fused wavelet coefficients. As each wavelet subband contains distinct detail information, a novel concept of multispectral co-occurrence of wavelet coefficients is introduced to capture the spatial correlation among different subbands. It enables to convey more informative features to characterize the tumor type. The effectiveness of the proposed method is analyzed, with respect to six classification performance indices, on BRATS 2012 and BRATS 2014 data sets. The classification accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and area under curve assessed by the ten-fold cross-validation are 91.3%, 96.8%, 66.7%, 92.4%, 88.4%, and 92.0%, respectively, on real brain MR data.

Texture Analysis Using Semiquantitative Kinetic Parameter Maps from DCE-MRI: Preoperative Prediction of HER2 Status in Breast Cancer

Objective

To evaluate whether texture features derived from semiquantitative kinetic parameter maps based on breast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can determine human epidermal growth factor receptor 2 (HER2) status of patients with breast cancer.

Materials and Methods

This study included 102 patients with histologically confirmed breast cancer, all of whom underwent preoperative breast DCE-MRI and were enrolled retrospectively. This cohort included 48 HER2-positive cases and 54 HER2-negative cases. Seven semiquantitative kinetic parameter maps were calculated on the lesion area. A total of 55 texture features were extracted from each kinetic parameter map. Patients were randomly divided into training (n = 72) and test (n = 30) sets. The least absolute shrinkage and selection operator (LASSO) was used to select features in the training set, and then, multivariate logistic regression analysis was conducted to establish the prediction models. The classification performance was evaluated by receiver operating characteristic (ROC) analysis.

Results

Among the seven prediction models, the model with features extracted from the early signal enhancement ratio (ESER) map yielded an area under the ROC curve (AUC) of 0.83 in the training set (sensitivity of 70.59%, specificity of 92.11%, and accuracy of 81.94%), and the highest AUC of 0.83 in the test set (sensitivity of 57.14%, specificity of 100.00%, and accuracy of 80.00%). The model with features extracted from the slope of signal intensity (SI_slope) map yielded the highest AUC of 0.92 in the training set (sensitivity of 82.35%, specificity of 97.37%, and accuracy of 90.28%), and an AUC of 0.79 in the test set (sensitivity of 92.86%, specificity of 68.75%, and accuracy of 80.00%).

Conclusions

Texture features derived from kinetic parameter maps, calculated based on breast DCE-MRI, have the potential to be used as imaging biomarkers to distinguish HER2-positive and HER2-negative breast cancer.

A Heterogeneity Radiomic Nomogram for Preoperative Differentiation of Primary Gastric Lymphoma From Borrmann Type IV Gastric Cancer

OBJECTIVE

This study aimed to preoperatively differentiate primary gastric lymphoma from Borrmann type IV gastric cancer by heterogeneity nomogram based on routine contrast-enhanced computed tomographic images.

METHODS

We enrolled 189 patients from 2 hospitals (90 in the training cohort and 99 in the validation cohort). Subjective findings, including high-enhanced mucosal sign, high-enhanced serosa sign, nodular or an irregular outer layer of the gastric wall, and perigastric fat infiltration, were assessed to construct a subjective finding model. A deep learning model was developed to segment tumor areas, from which 1680 three-dimensional heterogeneity radiomic parameters, including first-order entropy, second-order entropy, and texture complexity, were extracted to build a heterogeneity signature by least absolute shrinkage and selection operator logistic regression. A nomogram that integrates heterogeneity signature and subjective findings was developed by multivariate logistic regression. The diagnostic performance of the nomogram was assessed by discrimination and clinical usefulness.

RESULTS

High-enhanced serosa sign and nodular or an irregular outer layer of the gastric wall were identified as independent predictors for building the subjective finding model. High-enhanced serosa sign and heterogeneity signature were significant predictors for differentiating the 2 groups (all, P < 0.05). The area under the curve with heterogeneity nomogram was 0.932 (95% confidence interval, 0.863-0.973) in the validation cohort. Decision curve analysis and stratified analysis confirmed the clinical utility of the heterogeneity nomogram.

CONCLUSIONS

The proposed heterogeneity radiomic nomogram on contrast-enhanced computed tomographic images may help differentiate primary gastric lymphoma from Borrmann type IV gastric cancer preoperatively.

Value of Apparent Diffusion Coefficient Histogram Analysis in the Differential Diagnosis of Nasopharyngeal Lymphoma and Nasopharyngeal Carcinoma Based on Readout-Segmented Diffusion-Weighted Imaging

Background

This study aims to explore the utility of whole-lesion apparent diffusion coefficient (ADC) histogram analysis for differentiating nasopharyngeal lymphoma (NPL) from nasopharyngeal carcinoma (NPC) following readout-segmented echo-planar diffusion-weighted imaging (RESOLVE sequence).

Methods

Thirty-eight patients with NPL and 62 patients with NPC, who received routine head-and-neck MRI and RESOLVE (b-value: 0 and 1,000 s/mm²) examinations, were retrospectively evaluated as derivation cohort (February 2015 to August 2018); another 23 patients were analyzed as validation cohort (September 2018 to December 2019). The RESOLVE data were obtained from the MAGNETOM Skyra 3T MR system (Siemens Healthcare, Erlangen, Germany). Fifteen parameters derived from the whole-lesion histogram analysis (ADC_mean, variance, skewness, kurtosis, ADC₁, ADC₁₀, ADC₂₀, ADC₃₀, ADC₄₀, ADC₅₀, ADC₆₀, ADC₇₀, ADC₈₀, ADC₉₀, and ADC₉₉) were calculated for each patient. Then, statistical analyses were performed between the two groups to determine the statistical significance of each histogram parameter. A receiver operating characteristic curve (ROC) analysis was conducted to assess the diagnostic performance of each histogram parameter for distinguishing NPL from NPC and further tested in the validation cohort; calibration of the selected parameter was tested with Hosmer-Lemeshow test.

Results

NPL exhibited significantly lower ADC_mean, variance, ADC₁, ADC₁₀, ADC₂₀, ADC₃₀, ADC₄₀, ADC₅₀, ADC₆₀, ADC₇₀, ADC₈₀, ADC₉₀ and ADC₉₉, when compared to NPC (all, P < 0.05), while no significant differences were found on skewness and kurtosis. Furthermore, ADC₉₉ revealed the highest diagnostic efficiency, followed by ADC₁₀ and ADC₂₀. Optimal diagnostic performance (AUC = 0.790, sensitivity = 91.9%, and specificity = 63.2%) could be achieved when setting ADC₉₉ = 1,485.0 × 10^-6 mm²/s as the threshold value. The predictive performance was maintained in the validation cohort (AUC = 0.817, sensitivity = 94.6%, and specificity = 56.2%).

Conclusion

Whole-lesion ADC histograms based on RESOLVE are effective in differentiating NPC from NPL.

Texture Analysis in Brain Tumor MR Imaging

Texture analysis, as well as its broader category radiomics, describes a variety of techniques for image analysis that quantify the variation in surface intensity or patterns, including some that are imperceptible to the human visual system. Cerebral gliomas have been most rigorously studied in brain tumors using MR-based texture analysis (MRTA) to determine the correlation of various clinical measures with MRTA features. Promising results in cerebral gliomas have been shown in the previous MRTA studies in terms of the correlation with the World Health Organization grades, risk stratification in gliomas, and the differentiation of gliomas from other brain tumors. Multiple MRTA studies in gliomas have repeatedly shown high performance of entropy, a measure of the randomness in image intensity values, of either histogram- or gray-level co-occurrence matrix parameters. Similarly, researchers have applied MRTA to other brain tumors, including meningiomas and pediatric posterior fossa tumors.

However, the value of MRTA in the clinical use remains undetermined, probably because previous studies have shown only limited reproducibility of the result in the real world. The low-to-modest generalizability may be attributed to variations in MRTA methods, sampling bias that originates from single-institution studies, and overfitting problems to a limited number of samples.

To enhance the reliability and reproducibility of MRTA studies, researchers have realized the importance of standardizing methods in the field of radiomics. Another advancement is the recent development of a comprehensive assessment system to ensure the quality of a radiomics study. These two-way approaches will secure the validity of upcoming MRTA studies. The clinical use of texture analysis in brain MRI will be accelerated by these continuous efforts.

Association between Texture Analysis Parameters and Molecular Biologic KRAS Mutation in Non-Mucinous Rectal Cancer

Purpose

To evaluate the association between magnetic resonance imaging (MRI)-based texture parameters and Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation in patients with non-mucinous rectal cancer.

Materials and Methods

Seventy-nine patients who had pathologically confirmed rectal non-mucinous adenocarcinoma with or without KRAS-mutation and had undergone rectal MRI were divided into a training (n = 46) and validation dataset (n = 33). A texture analysis was performed on the axial T2-weighted images. The association was statistically analyzed using the Mann-Whitney U test. To extract an optimal cut-off value for the prediction of KRAS mutation, a receiver operating characteristic curve analysis was performed. The cut-off value was verified using the validation dataset.

Results

In the training dataset, skewness in the mutant group (n = 22) was significantly higher than in the wild-type group (n = 24) (0.221 ± 0.283; -0.006 ± 0.178, respectively, p = 0.003). The area under the curve of the skewness was 0.757 (95% confidence interval, 0.606 to 0.872) with a maximum accuracy of 71%, a sensitivity of 64%, and a specificity of 78%. None of the other texture parameters were associated with KRAS mutation (p > 0.05). When a cut-off value of 0.078 was applied to the validation dataset, this had an accuracy of 76%, a sensitivity of 86%, and a specificity of 68%.

Conclusion

Skewness was associated with KRAS mutation in patients with non-mucinous rectal cancer.

Predicting the Local Response of Metastatic Brain Tumor to Gamma Knife Radiosurgery by Radiomics With a Machine Learning Method

Purpose

The current study proposed a model to predict the response of brain metastases (BMs) treated by Gamma knife radiosurgery (GKRS) using a machine learning (ML) method with radiomics features. The model can be used as a decision tool by clinicians for the most desirable treatment outcome.

Methods and Material

Using MR image data taken by a FLASH (3D fast, low-angle shot) scanning protocol with gadolinium (Gd) contrast-enhanced T1-weighting, the local response (LR) of 157 metastatic brain tumors was categorized into two groups (Group I: responder and Group II: non-responder). We performed a radiomics analysis of those tumors, resulting in more than 700 features. To build a machine learning model, first, we used the least absolute shrinkage and selection operator (LASSO) regression to reduce the number of radiomics features to the minimum number of features useful for the prediction. Then, a prediction model was constructed by using a neural network (NN) classifier with 10 hidden layers and rectified linear unit activation. The training model was evaluated with five-fold cross-validation. For the final evaluation, the NN model was applied to a set of data not used for model creation. The accuracy and sensitivity and the area under the receiver operating characteristic curve (AUC) of the prediction model of LR were analyzed. The performance of the ML model was compared with a visual evaluation method, for which the LR of tumors was predicted by examining the image enhancement pattern of the tumor on MR images.

Results

By the LASSO analysis of the training data, we found seven radiomics features useful for the classification. The accuracy and sensitivity of the visual evaluation method were 44 and 54%. On the other hand, the accuracy and sensitivity of the proposed NN model were 78 and 87%, and the AUC was 0.87.

Conclusions

The proposed NN model using the radiomics features can help physicians to gain a more realistic expectation of the treatment outcome than the traditional method.

Texture analysis of apparent diffusion coefficient (ADC) map for glioma grading: Analysis of whole tumoral and peri-tumoral tissue

PURPOSE

To prospectively investigate the capabilities of texture analysis (TA) based on apparent diffusion coefficient (ADC) map of the entire tumor volume and the whole volume of peri-tumoral edema, in discriminating between high-grade glioma (HGG) and low-grade glioma (LGG).

MATERIALS AND METHODS

A total of 33 patients with histopathological proven glioma were prospectively included. There were 20 men and 13 women with a mean age of 54.5±14.7 (standard deviation [SD]) years (range: 34-75years). TA parameters of whole tumor and peri-tumoral edema were extracted from the ADC map obtained with diffusion-weighted spin-echo echo-planar magnetic resonance imaging at 1.5-T. TA variables of HGG were compared to those of LGG. The optimum cut-off values of TA variables and their corresponding sensitivity, specificity and accuracy for differentiating between LGG and HGG were calculated using receiver operating characteristic curve analysis.

RESULTS

Mean and median tumoral ADC of HGG were significantly lower than those of LGG, at 1.23×10^-3 mm²/s and 1.21×10^-3 mm²/s cut-off values, yielding 70% sensitivity each (95% CI: 59-82% and 61-80%, respectively), 80% (95% CI: 79-98%) and 90% (95% CI: 82-97%) specificity, and 73% (95% CI: 66-91%) and 76% (95% CI: 72-90%) accuracy, respectively. Significant differences in tumoral and peri-tumoral kurtosis were found between HGG and LGG at 1.60 and 0.314 cut-off values yielding sensitivities of 74% (95% CI: 58-83%) and 70% (95% CI: 59-84%), specificities of 90% (95% CI: 80-95%) and 70% (95% CI: 64-83%) and accuracies of 79% (95% CI: 69-89%) and 70% (95% CI: 64-77%), respectively.

CONCLUSION

Measurements of whole tumoral and peri-tumoral TA, based on ADC maps, provide useful information that helps distinguish between HGG and LGG.

Prediction of the Ki-67 marker index in hepatocellular carcinoma based on CT radiomics features

The noninvasive detection of tumor proliferation is of great value and the Ki-67 is a biomarker of tumor proliferation. We hypothesized that radiomics characteristics may be related to tumor proliferation. To evaluate whether computed tomography (CT) radiomics feature analyses could aid in assessing the Ki-67 marker index in hepatocellular carcinoma (HCC), we retrospectively analyzed preoperative CT findings of 74 patients with HCC. The texture feature calculations were computed from MaZda 4.6 software, and the sequential forward selection algorithm was used as the selection method. The correlation between radiomics features and the Ki-67 marker index, as well as the difference between low Ki-67 (<10%) and high Ki-67 (≥10%) groups were evaluated. A simple logistic regression model was used to evaluate the associations between texture features and high Ki-67, and receiver operating characteristic analysis was performed on important parameters to assess the ability of radiomics characteristics to distinguish the high Ki-67 group from the low Ki-67 group. Contrast, correlation, and inverse difference moment (IDM) were significantly different (P < 0.001) between the low and high Ki-67 groups. Contrast (odds ratio [OR] = 0.957; 95% confidence interval [CI]: 0.926-0.990, P = 0.01) and correlation (OR = 2.5☆105; 95% CI: 7.560-8.9☆109; P = 0.019) were considered independent risk factors for combined model building with logistic regression. Angular second moment (r = -0.285, P = 0.014), contrast (r = -0.449, P < 0.001), correlation (r = 0.552, P < 0.001), IDM (r = 0.458, P < 0.001), and entropy (r = 0.285, P = 0.014) strongly correlated with the Ki-67 scores. Contrast, correlation, and the combined predictor were predictive of Ki-67 status (P < 0.001), with areas under the curve ranging from 0.777 to 0.836. The radiomics characteristics of CT have potential as biomarkers for predicting Ki-67 status in patients with HCC. These findings suggest that the radiomics features of CT might be used as a noninvasive measure of cellular proliferation in HCC.

Texture analysis on bi-parametric MRI for evaluation of aggressiveness in patients with prostate cancer

PURPOSE

To evaluate the association between texture parameters based on bi-parametric MRI and Gleason score (GS) in patients with prostate cancer (PCa) and to evaluate diagnostic performance of any significant parameter for discriminating clinically significant cancer (CSC, GS ≥ 7) from non-CSC.

METHODS

A total of 116 patients who had been confirmed as prostate adenocarcinoma by radical prostatectomy or biopsy were divided into a training (n = 65) and a validation dataset (n = 51). All of the patients underwent preoperative 3T-MRI. Texture analysis was performed on axial T2WI and ADC maps (generated from b values, 0 and 1000 s/mm²) using dedicated software to cover the whole tumor volume. The correlation coefficient was calculated to evaluate the association between texture parameters and GS, and subsequent multiple regression analyses were applied for the significant parameters. To extract an optimal cut-off value for prediction of CSC, ROC curve analysis was performed.

RESULTS

In the training dataset, gray-level co-occurrence matrix (GLCM) entropy on ADC map was the only significant indicator for GS (coefficient of determination R², 0.4227, P = 0.0034). The AUC of GLCM entropy on ADC map was 0.825 (95% CI 0.711-0.907) with a maximum accuracy of 82%, a sensitivity of 86%, a specificity of 71%. When a cut-off value of 2.92 was applied to the validation dataset, it showed an accuracy of 92%, a sensitivity of 98%, and a specificity of 70%.

CONCLUSION

GLCM entropy on ADC map was associated with GS in patients with PCa and its estimated accuracy for discriminating CSC from non-CSC was 82%.

Spatiotemporal Heterogeneity in Multiparametric Physiologic MRI Is Associated with Patient Outcomes in IDH-Wildtype Glioblastoma

PURPOSE

Heterogeneity in glioblastomas is associated with poorer outcomes, and physiologic heterogeneity can be quantified with noninvasive imaging. We developed spatial habitats based on multiparametric physiologic MRI and evaluated associations between temporal changes in these habitats and progression-free survival (PFS) after concurrent chemoradiotherapy (CCRT) in patients with glioblastoma.

EXPERIMENTAL DESIGN

Ninety-seven patients with isocitrate dehydrogenase (IDH)-wildtype glioblastoma were enrolled and two serial MRI examinations after CCRT were analyzed. Cerebral blood volumes and apparent diffusion coefficients were grouped using k-means clustering into three spatial habitats. Associations between temporal changes in spatial habitats and PFS were investigated using Cox proportional hazard modeling. The performance of significant predictors for PFS and overall survival (OS) was measured using a discrete increase of habitat (habitat risk score) in a temporal validation set from a prospective registry (n = 53, ClinicalTrials.gov NCT02619890). The site of progression was matched with the spatiotemporal habitats.

RESULTS

Three spatial habitats of hypervascular cellular, hypovascular cellular, and nonviable tissue were identified. A short-term increase in the hypervascular cellular habitat (HR, 40.0; P = 0.001) and hypovascular cellular habitat was significantly associated with shorter PFS (HR, 3.78; P < 0.001) after CCRT. Combined with clinical predictors, the habitat risk score showed a C-index of 0.79 for PFS and 0.74 for OS and stratified patients with short, intermediate, and long PFS (P = 0.016). An increase in the hypovascular cellular habitat predicted tumor progression sites.

CONCLUSIONS

Hypovascular cellular habitats derived from multiparametric physiologic MRIs may be useful predictors of clinical outcomes in patients with posttreatment glioblastoma.

Radiomic analysis of magnetic resonance fingerprinting in adult brain tumors

PURPOSE

This is a radiomics study investigating the ability of texture analysis of MRF maps to improve differentiation between intra-axial adult brain tumors and to predict survival in the glioblastoma cohort.

METHODS

Magnetic resonance fingerprinting (MRF) acquisition was performed on 31 patients across 3 groups: 17 glioblastomas, 6 low-grade gliomas, and 8 metastases. Using regions of interest for the solid tumor and peritumoral white matter on T1 and T2 maps, second-order texture features were calculated from gray-level co-occurrence matrices and gray-level run length matrices. Selected features were compared across the three tumor groups using Wilcoxon rank-sum test. Receiver operating characteristic curve analysis was performed for each feature. Kaplan-Meier method was used for survival analysis with log rank tests.

RESULTS

Low-grade gliomas and glioblastomas had significantly higher run percentage, run entropy, and information measure of correlation 1 on T1 than metastases (p < 0.017). The best separation of all three tumor types was seen utilizing inverse difference normalized and homogeneity values for peritumoral white matter in both T1 and T2 maps (p < 0.017). In solid tumor T2 maps, lower values in entropy and higher values of maximum probability and high-gray run emphasis were associated with longer survival in glioblastoma patients (p < 0.05). Several texture features were associated with longer survival in glioblastoma patients on peritumoral white matter T1 maps (p < 0.05).

CONCLUSION

Texture analysis of MRF-derived maps can improve our ability to differentiate common adult brain tumors by characterizing tumor heterogeneity, and may have a role in predicting outcomes in patients with glioblastoma.

Enhanced Computed Tomography-Based Radiomics Signature Combined With Clinical Features in Evaluating Nuclear Grading of Renal Clear Cell Carcinoma

OBJECTIVE

The aim of the study was to explore the value of enhanced computed tomography (CT)-based radiomics signature combined with clinical features in evaluating nuclear grading of clear cell renal cell carcinoma (ccRCC).

METHODS

One hundred one patients with ccRCC were classified into low- and high-grade group, and the data were divided into training set and verification set. Radiomics signatures were constructed in the training set in enhanced 3 stages and the combination of them. The predictive nomogram was constructed. The classification efficiency and the clinical practicability of the integrated radiomics model were evaluated.

RESULTS

The classification efficiency of enhanced 3-stage integrated histology model was higher than that of each single-phase model. The predictive nomogram incorporated the best radiomics signature, and the independent clinical risk factors showed good performance. A decision curve analysis curve shows that the net benefit of the combined model.

CONCLUSIONS

It is feasible to evaluate the nuclear grading of ccRCC based on enhanced CT radiomics signature combined with clinical features.

Differentiating High-Grade Gliomas from Brain Metastases at Magnetic Resonance: The Role of Texture Analysis of the Peritumoral Zone

High-grade gliomas (HGGs) and solitary brain metastases (BMs) have similar imaging appearances, which often leads to misclassification. In HGGs, the surrounding tissues show malignant invasion, while BMs tend to displace the adjacent area. The surrounding edema produced by the two cannot be differentiated by conventional magnetic resonance (MRI) examinations. Forty-two patients with pathology-proven brain tumors who underwent conventional pretreatment MRIs were retrospectively included (HGGs, n = 16; BMs, n = 26). Texture analysis of the peritumoral zone was performed on the T2-weighted sequence using dedicated software. The most discriminative texture features were selected using the Fisher and the probability of classification error and average correlation coefficients. The ability of texture parameters to distinguish between HGGs and BMs was evaluated through univariate, receiver operating, and multivariate analyses. The first percentile and wavelet energy texture parameters were independent predictors of HGGs (75–87.5% sensitivity, 53.85–88.46% specificity). The prediction model consisting of all parameters that showed statistically significant results at the univariate analysis was able to identify HGGs with 100% sensitivity and 66.7% specificity. Texture analysis can provide a quantitative description of the peritumoral zone encountered in solitary brain tumors, that can provide adequate differentiation between HGGs and BMs.

Measurement of proptosis using computed tomography based three-dimensional reconstruction software in patients with Graves' orbitopathy

The evaluation of proptosis is essential for the diagnosis of orbital disease. We have developed a computed tomography (CT)-based three-dimensional (3D) reconstruction software to measure the degree of proptosis. To verify clinical usefulness and reliability, the degree of proptosis was measured in 126 patients with Graves’ orbitopathy (GO) with 3D reconstruction software and compared with those obtained with Hertel exophthalmometer and CT. The proptosis values measured by 3D reconstruction software, CT, and Hertel exophthalmometer were closely related to each other, but showed significant differences (p < 0.001). In contrast, the amount of change in proptosis after orbital decompression were not different among the three measurements (p = 0.153). The intra-observer correlation coefficients of the 3D reconstruction software, CT, and Hertel exophthalmometer measurements were 0.997, 0.942, and 0.953, respectively. In patients with strabismus, the intra-observer correlation coefficient of CT and Hertel exophthalmometer decreased to 0.895 and 0.920, respectively, but the intra-observer correlation coefficient of the 3D reconstruction software did not change to 0.996. The inter-observer correlation coefficients of CT and 3D reconstruction software for three different ophthalmologists were 0.742 and 0.846, respectively. In conclusion, the measurement of proptosis by 3D reconstruction software seems to be a reliable method, especially in the presence of eyeball deviation.

Application of Texture Analysis Based on Sagittal Fat-Suppression and Oblique Axial T2-Weighted Magnetic Resonance Imaging to Identify Lymph Node Invasion Status of Rectal Cancer

Objective: To investigate the value of texture features derived from T2-weighted magnetic resonance imaging (T2WI) for predicting preoperative lymph node invasion (N stage) in rectal cancer.

Materials and Methods: One hundred and eighty-two patients with histopathologically confirmed rectal cancer and preoperative magnetic resonance imaging were retrospectively analyzed, who were divided into high (N1-2) and low N stage (N0). Texture features were calculated from histogram, gray-level co-occurrence matrix, and gray-level run-length matrix from sagittal fat-suppression and oblique axial T2WI. Independent sample t-test or Mann-Whitney U-test were used for statistical analysis. Multivariate logistic regression analysis was conducted to build the predictive models. Predictive performance was evaluated by receiver operating characteristic (ROC) analysis.

Results: Energy (ENE), entropy (ENT), information correlation (INC), long-run emphasis (LRE), and short-run low gray-level emphasis (SRLGLE) extracted from sagittal fat-suppression T2WI, and ENE, ENT, INC, low gray-level run emphasis (LGLRE), and SRLGLE from oblique axial T2WI were significantly different between stage N0 and stage N1-2 tumors. The multivariate analysis for features from sagittal fat-suppression T2WI showed that higher SRLGLE and lower ENE were independent predictors of lymph node invasion. The model reached an area under ROC curve (AUC) of 0.759. The analysis for features from oblique axial T2WI showed that higher INC and SRLGLE were independent predictors. The model achieved an AUC of 0.747. The analysis for all extracted features showed that lower ENE from sagittal fat-suppression T2WI and higher INC and SRLGLE from oblique axial T2WI were independent predictors. The model showed an AUC of 0.772.

Conclusions: Texture features derived from T2WI could provide valuable information for identifying the status of lymph node invasion in rectal cancer.

Differentiating glioblastoma multiforme from cerebral lymphoma: application of advanced texture analysis of quantitative apparent diffusion coefficients

PURPOSE

The purpose of this study was to differentiate glioblastoma multiforme from primary central nervous system lymphoma using the customised first and second-order histogram features derived from apparent diffusion coefficients.Methods and materials: A total of 82 patients (57 with glioblastoma multiforme and 25 with primary central nervous system lymphoma) were included in this study. The axial T1 post-contrast and fluid-attenuated inversion recovery magnetic resonance images were used to delineate regions of interest for the tumour and peritumoral oedema. The regions of interest were then co-registered with the apparent diffusion coefficient maps, and the first and second-order histogram features were extracted and compared between glioblastoma multiforme and primary central nervous system lymphoma groups. Receiver operating characteristic curve analysis was performed to calculate a cut-off value and its sensitivity and specificity to differentiate glioblastoma multiforme from primary central nervous system lymphoma.

RESULTS

Based on the tumour regions of interest, apparent diffusion coefficient mean, maximum, median, uniformity and entropy were higher in the glioblastoma multiforme group than the primary central nervous system lymphoma group (P ≤ 0.001). The most sensitive first and second-order histogram feature to differentiate glioblastoma multiforme from primary central nervous system lymphoma was the maximum of 2.026 or less (95% confidence interval (CI) 75.1-99.9%), and the most specific first and second-order histogram feature was smoothness of 1.28 or greater (84.0% CI 70.9-92.8%). Based on the oedema regions of interest, most of the first and second-order histogram features were higher in the glioblastoma multiforme group compared to the primary central nervous system lymphoma group (P ≤ 0.015). The most sensitive first and second-order histogram feature to differentiate glioblastoma multiforme from primary central nervous system lymphoma was the 25th percentile of 0.675 or less (100% CI 83.2-100%) and the most specific first and second-order histogram feature was the median of 1.28 or less (85.9% CI 66.3-95.8%).

CONCLUSIONS

Texture analysis using first and second-order histogram features derived from apparent diffusion coefficient maps may be helpful in differentiating glioblastoma multiforme from primary central nervous system lymphoma.

Advances and potential pitfalls of oncolytic viruses expressing immunomodulatory transgene therapy for malignant gliomas

Glioblastoma (GBM) is an immunosuppressive, lethal brain tumor. Despite advances in molecular understanding and therapies, the clinical benefits have remained limited, and the life expectancy of patients with GBM has only been extended to ~15 months. Currently, genetically modified oncolytic viruses (OV) that express immunomodulatory transgenes constitute a research hot spot in the field of glioma treatment. An oncolytic virus is designed to selectively target, infect, and replicate in tumor cells while sparing normal tissues. Moreover, many studies have shown therapeutic advantages, and recent clinical trials have demonstrated the safety and efficacy of their usage. However, the therapeutic efficacy of oncolytic viruses alone is limited, while oncolytic viruses expressing immunomodulatory transgenes are more potent inducers of immunity and enhance immune cell-mediated antitumor immune responses in GBM. An increasing number of basic studies on oncolytic viruses encoding immunomodulatory transgene therapy for malignant gliomas have yielded beneficial outcomes. Oncolytic viruses that are armed with immunomodulatory transgenes remain promising as a therapy against malignant gliomas and will undoubtedly provide new insights into possible clinical uses or strategies. In this review, we summarize the research advances related to oncolytic viruses that express immunomodulatory transgenes, as well as potential treatment pitfalls in patients with malignant gliomas.

MRI of tumor T cell infiltration in response to checkpoint inhibitor therapy

BACKGROUND

Immune checkpoint inhibitors, the most widespread class of immunotherapies, have demonstrated unique response patterns that are not always adequately captured by traditional response criteria such as the Response Evaluation Criteria in Solid Tumors or even immune-specific response criteria. These response metrics rely on monitoring tumor growth, but an increase in tumor size and/or appearance after starting immunotherapy does not always represent tumor progression, but also can be a result of T cell infiltration and thus positive treatment response. Therefore, non-invasive and longitudinal monitoring of T cell infiltration are needed to assess the effects of immunotherapies such as checkpoint inhibitors. Here, we proposed an innovative concept that a sufficiently large influx of tumor infiltrating T cells, which have a smaller diameter than cancer cells, will change the diameter distribution and decrease the average size of cells within a volume to a degree that can be quantified by non-invasive MRI.

METHODS

We validated our hypothesis by studying tumor response to combination immune-checkpoint blockade (ICB) of anti-PD-1 and anti-CTLA4 in a mouse model of colon adenocarcinoma (MC38). The response was monitored longitudinally using Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion (IMPULSED), a diffusion MRI-based method which has been previously shown to non-invasively map changes in intracellular structure and cell sizes with the spatial resolution of MRI, in cell cultures and in animal models. Tumors were collected for immunohistochemical and flow cytometry analyzes immediately after the last imaging session.

RESULTS

Immunohistochemical analysis revealed that increased T cell infiltration of the tumors results in a decrease in mean cell size (eg, a 10% increase of CD3+ T cell fraction results a ~1 µm decrease in the mean cell size). IMPULSED showed that the ICB responders, mice with tumor volumes were less than 250 mm3 or had tumors with stable or decreased volumes, had significantly smaller mean cell sizes than both Control IgG-treated tumors and ICB non-responder tumors.

CONCLUSIONS

IMPULSED-derived cell size could potentially serve as an imaging marker for differentiating responsive and non-responsive tumors after checkpoint inhibitor therapies, a current clinical challenge that is not solved by simply monitoring tumor growth.

Characterization of parotid gland tumors using T2 mapping imaging: initial findings

Background

Preoperative accurate characterization of parotid gland tumors in different histologic types is crucial. T2 mapping has been proven to be useful for improving the accuracy of tumor characterization.

Purpose

To evaluate the ability of T2 mapping imaging in the characterization of parotid gland tumors.

Material and Methods

T2 mapping imaging was scanned in 74 patients (56 benign, 18 malignant) with pathologically confirmed parotid gland tumors. Mean T2 relaxation time was calculated and compared between benign and malignant group, and among malignant tumors, Warthin’s tumors, and pleomorphic adenomas. Independent-samples t test, one-way analysis of variance test, and receiver operating characteristic curve analyses were used for statistical analyses.

Results

The malignant group showed significantly lower T2 relaxation times than the benign group ( P = 0.001). Using a relaxation time of 91.5 ms as the cut-off value, optimal diagnostic performance could be achieved (area under the curve [AUC] 0.679, sensitivity 46.4%, specificity 94.4%). Pleomorphic adenomas showed significantly higher T2 relaxation times than malignant tumors ( P = 0.003) and Warthin’s tumors ( P = 0.001). However, no significant difference was found on the T2 relaxation times between Warthin’s tumors and malignant tumors ( P = 0.435). Optimal diagnostic performance could be achieved (AUC 0.783, sensitivity 58.1%, specificity 94.4%), when setting a T2 value of 92.0 ms as the threshold value for differentiating pleomorphic adenomas from malignant tumors. Meanwhile, optimal AUC, sensitivity, and specificity were 0.892, 87.1%, and 83.3%, respectively, when setting a T2 value of 80.5 ms as the threshold value for differentiating pleomorphic adenomas from Warthin’s tumors.

Conclusion

T2 mapping imaging could serve as an incremental imaging biomarker for characterizing parotid gland tumors.

Radiogenomics predicts the expression of microRNA-1246 in the serum of esophageal cancer patients

Radiogenomics is a new field that provides clinically useful prognostic predictions by linking molecular characteristics such as the genetic aberrations of malignant tumors with medical images. The abnormal expression of serum microRNA-1246 (miR-1246) has been reported as a prognostic factor of esophageal squamous cell carcinoma (ESCC). To evaluate the power of the miR-1246 level predicted with radiogenomics techniques as a predictor of the prognosis of ESCC patients. The real miR-1246 expression (miR-1246_real) was measured in 92 ESCC patients. Forty-five image features (IFs) were extracted from tumor regions on contrast-enhanced computed tomography. A prediction model for miR-1246_real was constructed using linear regression with selected features identified in a correlation analysis of miR-1246_real and each IF. A threshold to divide the patients into two groups was defined according to a receiver operating characteristic analysis for miR-1246_real. Survival analyses were performed between two groups. Six IFs were correlated with miR-1246_real and were included in the prediction model. The survival curves of high and low groups of miR-1246_real and miR-1246_pred showed significant differences (p = 0.001 and 0.016). Both miR-1246_real and miR-1246_pred were independent predictors of overall survival (p = 0.030 and 0.035). miR-1246_pred produced by radiogenomics had similar power to miR-1246_real for predicting the prognosis of ESCC.

Accuracy of ADC derived from DWI for differentiating high-grade from low-grade gliomas: Systematic review and meta-analysis

OBJECTIVE

Quantitative apparent diffusion coefficient (ADC) values of diffusion weighted imaging (DWI) could be applied to grade gliomas. This meta-analysis was conducted to assess the accuracy of ADC analysis in differentiating high-grade (HGGs) from low-grade gliomas (LGGs).

METHODS

PubMed, Cochrane library, Science Direct, and Embase were searched to identify suitable studies up to September 1, 2018. The quality of studies was evaluated by the quality assessment of diagnostic accuracy studies (QUADAS 2). We estimated the pooled sensitivity, specificity, positive and negative likelihood ratios (LR), diagnostic accuracy ratio (DOR) with 95% confidence intervals (CI), and determined the accuracy of the data by using the summary receiver operating characteristic (SROC) and calculating the area under the curve (AUC) to identity the accuracy of ADC analysis in grading gliomas.

RESULTS

Eighteen studies including 1172 patients were included and analyzed. The pooled sensitivity, specificity, PLR, NLR, DOR, and AUC with 95% CIs of DWI with b values of 1000 s/mm for separating HGGs from LGGs were 0.81 (95% CI 0.75-0.86), 0.87 (95% CI 0.81-0.91), 6.1 (95% CI 4.2-8.9), 0.22 (95% CI 0.17-0.29), 28 (95% CI 17-45), and 0.91 (95% CI 0.88-0.93), respectively. DWI with b values of 3000 s/mm showed slightly higher accuracy than that of 1000 (sensitivity 0.80, specificity 0.90 and AUC 0.92). Meta-regression analyses showed that field strengths and b values had significant impacts on diagnostic efficacy. Deeks testing confirmed no significant publication bias in all studies.

CONCLUSIONS

This meta-analysis suggested that ADC analysis of DWI have high accuracy in differentiating HGGs from LGGs. Standardized methodology is warranted to guide the use of this technique for clinical decision-making.