Improving tumour heterogeneity MRI assessment with histograms. Br J Cancer. 2014;111:2205-2213. [PMID: 25268373 PMCID: PMC4264439 DOI: 10.1038/bjc.2014.512]

Histogram analysis of stretched-exponential and monoexponential diffusion-weighted imaging models for distinguishing low and intermediate/high gleason scores in prostate carcinoma

BACKGROUND

Noninvasive measures to evaluate the aggressiveness of prostate carcinoma (PCa) may benefit patients.

PURPOSE

To assess the value of stretched-exponential and monoexponential diffusion-weighted imaging (DWI) for predicting the aggressiveness of PCa.

STUDY TYPE

Retrospective study.

SUBJECTS

Seventy-five patients with PCa.

FIELD STRENGTH

3T DWI examinations were performed using b-values of 0, 500, 1000, and 2000 s/mm² .

ASSESSMENT

The research were based on entire-tumor histogram analysis and the reference standard was radical prostectomy.

STATISTICAL TESTS

The correlation analysis was programmed with Spearman's rank-order analysis between the histogram variables and Gleason grade group (GG). Receiver operating characteristic (ROC) regression was used to analyze the ability of these histogram variables to differentiate low-grade (LG) from intermediate/high-grade (HG) PCa.

RESULTS

The percentiles and mean of apparent diffusion coefficient (ADC) and distributed diffusion coefficient (DDC) were correlated with GG (ρ: 0.414-0.593), while there was no significant relation among α value, skewnesses, and kurtosises with GG (ρ:0.034-0.323). HG tumors (ADC:484 ± 136, 592 ± 139, 670 ± 144, 788 ± 146, 895 ± 141 mm² /s; DDC: 410 ± 142, 532 ± 172, 666 ± 193, 786 ± 196, 914 ± 181 mm² /s) had lower values in the 10^th , 25^th , 50^th , 75^th percentiles and means than LG tumors (ADC: 644 ± 779, 737 ± 84, 836 ± 83, 919 ± 82, 997 ± 107 mm² /s; DDC: 552 ± 82, 680 ± 94, 829 ± 112, 931 ± 106, 1045 ± 100 mm² /s). However, there was no difference between LG and HG tumors in α value (0.671 ± 0.041 vs. 0.633 ± 0.114), kurtosises (ADC 0.09 vs. 0.086; DDC -0.033 vs. -0.317), or skewnesses (ADC -0.036 vs. 0.073; DDC -0.063 vs. 0.136). The above statistics were P < 0.01. ADC10 with AUC = 0.840 and DDC10 with AUC = 0.799 were similar in discriminating between LG and HG PCa at P < 0.05.

DATA CONCLUSION

Histogram variables of DDC and ADC may predict the aggressiveness of PCa, while α value does not. The abilities of ADC10 and DDC10 to discriminate LG from HG tumors were similar, and both better than their respective means.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 1 J. MAGN. RESON. IMAGING 2018;48:491-498.

Can histogram analysis of MR images predict aggressiveness in pancreatic neuroendocrine tumors?

OBJECTIVES

To evaluate MRI derived whole-tumour histogram analysis parameters in predicting pancreatic neuroendocrine neoplasm (panNEN) grade and aggressiveness.

METHODS

Pre-operative MR of 42 consecutive patients with panNEN >1 cm were retrospectively analysed. T1-/T2-weighted images and ADC maps were analysed. Histogram-derived parameters were compared to histopathological features using the Mann-Whitney U test. Diagnostic accuracy was assessed by ROC-AUC analysis; sensitivity and specificity were assessed for each histogram parameter.

RESULTS

ADC_entropy was significantly higher in G2-3 tumours with ROC-AUC 0.757; sensitivity and specificity were 83.3 % (95 % CI: 61.2-94.5) and 61.1 % (95 % CI: 36.1-81.7). ADC_kurtosis was higher in panNENs with vascular involvement, nodal and hepatic metastases (p= .008, .021 and .008; ROC-AUC= 0.820, 0.709 and 0.820); sensitivity and specificity were: 85.7/74.3 % (95 % CI: 42-99.2 /56.4-86.9), 36.8/96.5 % (95 % CI: 17.2-61.4 /76-99.8) and 100/62.8 % (95 % CI: 56.1-100/44.9-78.1). No significant differences between groups were found for other histogram-derived parameters (p >.05).

CONCLUSIONS

Whole-tumour histogram analysis of ADC maps may be helpful in predicting tumour grade, vascular involvement, nodal and liver metastases in panNENs. ADC_entropy and ADC_kurtosis are the most accurate parameters for identification of panNENs with malignant behaviour.

KEY POINTS

• Whole-tumour ADC histogram analysis can predict aggressiveness in pancreatic neuroendocrine neoplasms. • ADC entropy and kurtosis are higher in aggressive tumours. • ADC histogram analysis can quantify tumour diffusion heterogeneity. • Non-invasive quantification of tumour heterogeneity can provide adjunctive information for prognostication.

Histogram analysis of apparent diffusion coefficient maps for assessing thymic epithelial tumours: correlation with world health organization classification and clinical staging

OBJECTIVE

To investigate the value of apparent diffusion coefficients (ADCs) histogram analysis for assessing World Health Organization (WHO) pathological classification and Masaoka clinical stages of thymic epithelial tumours.

METHODS

37 patients with histologically confirmed thymic epithelial tumours were enrolled. ADC measurements were performed using hot-spot ROI (ADC_HS-ROI) and histogram-based approach. ADC histogram parameters included mean ADC (ADC_mean), median ADC (ADC_median), 10 and 90 percentile of ADC (ADC₁₀ and ADC₉₀), kurtosis and skewness. One-way ANOVA, independent-sample t-test, and receiver operating characteristic were used for statistical analyses.

RESULTS

There were significant differences in ADC_mean, ADC_median, ADC₁₀, ADC₉₀ and ADC_HS-ROI among low-risk thymoma (type A, AB, B1; n = 14), high-risk thymoma (type B2, B3; n = 9) and thymic carcinoma (type C, n = 14) groups (all p-values <0.05), while no significant difference in skewness (p = 0.181) and kurtosis (p = 0.088). ADC₁₀ showed best differentiating ability (cut-off value, ≤0.689 × 10^-3 mm² s^-1; AUC, 0.957; sensitivity, 95.65%; specificity, 92.86%) for discriminating low-risk thymoma from high-risk thymoma and thymic carcinoma. Advanced Masaoka stages (Stage III and IV; n = 24) tumours showed significant lower ADC parameters and higher kurtosis than early Masaoka stage (Stage I and II; n = 13) tumours (all p-values <0.05), while no significant difference on skewness (p = 0.063). ADC₁₀ showed best differentiating ability (cut-off value, ≤0.689 × 10^-3 mm² s^-1; AUC, 0.913; sensitivity, 91.30%; specificity, 85.71%) for discriminating advanced and early Masaoka stage epithelial tumours.

CONCLUSION

ADC histogram analysis may assist in assessing the WHO pathological classification and Masaoka clinical stages of thymic epithelial tumours. Advances in knowledge: 1. ADC histogram analysis could help to assess WHO pathological classification of thymic epithelial tumours. 2. ADC histogram analysis could help to evaluate Masaoka clinical stages of thymic epithelial tumours. 3. ADC₁₀ might be a promising imaging biomarker for assessing and characterizing thymic epithelial tumours.

Differentiation between malignant and benign breast masses: combination of semi-quantitative analysis on DCE-MRI and histogram analysis of ADC maps

AIM

To investigate the performance of combined semi-quantitative analysis on dynamic contrast enhanced (DCE) magnetic resonance imaging (MRI) and histogram analysis of diffusion-weighted imaging (DWI) for distinguishing malignant from benign breast masses.

MATERIALS AND METHODS

This study included 178 patients with breast masses (benign:malignant=88:9) who underwent both DCE-MRI and DWI. The semi-quantitative parameters, derived from DCE-MRI, included maximum slope of increase (MSI), signal intensity slope (SI_slope), initial percentage of enhancement (E_initial), percentage of peak enhancement (E_peak), early signal enhancement ratio (ESER), and second enhancement percentage (SEP). Histogram parameters derived from apparent diffusion coefficient (ADC) maps included ADC_min, ADC_max, ADC_mean, ADC₁₀, ADC₂₅, ADC₅₀, ADC₇₅, ADC₉₀, skewness, and kurtosis. All parameters were compared between malignant and benign groups, and their differences were tested using independent-samples t-test or Mann-Whitney test. Receiver operating characteristic (ROC) curves were used to determine the diagnostic value of each significant parameter.

RESULTS

Among semi-quantitative parameters, SI_slope exhibited the best diagnostic performance in predicting malignancy (cut-off value, 0.096; ROC, 0.756; sensitivity, 86.7%; specificity, 61.4%). Among histogram parameters, ADC₁₀ exhibited the best diagnostic performance in predicting malignancy (cut-off value, 1.051; ROC, 0.885; sensitivity, 86.7%; specificity, 84.1%). The optimal diagnostic performance of combined ADC₁₀ and SI_slope (area under curve [AUC], 0.888; sensitivity, 82.2%; specificity, 95.5%) was significantly better than SI_slope alone (p<0.001). Moreover, the combination showed higher AUC (0.888 versus 0.885) than ADC₁₀ alone, but the difference was not statistically significant (p=0.914).

CONCLUSION

SI_slope and ADC₁₀ are significant predictors for breast malignancy. The combination of DCE-MRI and DWI improves differentiating performance.

MRI Texture Analysis Reflects Histopathology Parameters in Thyroid Cancer - A First Preliminary Study

OBJECT

Thyroid cancer represents the most frequent malignancy of the endocrine system with an increasing incidence worldwide. Novel imaging techniques are able to further characterize tumors and even predict histopathology features. Texture analysis is an emergent imaging technique to extract extensive data from an radiology images. The present study was therefore conducted to identify possible associations between texture analysis and histopathology parameters in thyroid cancer.

METHODS

The radiological database was retrospectively reviewed for thyroid carcinoma. Overall, 13 patients (3 females, 23.1%) with a mean age of 61.6 years were identified. The MaZda program was used for texture analysis. The T1-precontrast and T2-weighted images were analyzed and overall 279 texture feature for each sequence was investigated. For every patient cell count, Ki67-index and p53 count were investigated.

RESULTS

Several significant correlations between texture features and histopathology were identified. Regarding T1-weighted images, S(0;1)Sum Averg correlated the most with cell count (r=0.82). An inverse correlations with S(5;0)AngScMom, S(5;0)DifVarnc S(5;0), DiffEntrp and GrNonZeros (r=-0.69, -0.66, -0.69 and -0.63, respectively) was also identified. For T2-weighted images, Variance with r=0.63 was the highest coefficient, WavEnLL_S3 correlated inversely with cell count (r=-0.57). WavEnLL_S2 derived from T1-weighted images was the highest coefficient r=-0.80, S(0;5)SumVarnc was positively with r=0.74. Regarding T2-weighted images WavEnHL_s-1 was inverse correlated with Ki67 index (r=-0.77). S(1;0)Correlat was with r=0.75 the best correlation with Ki67 index. For T1-weighed images S(5;0)SumofSqs was the best with r=0.65 with p53 count. For T2-weighted images S(1;-1)SumEntrp was the inverse correlation with r=-0.72, whereas S(0;4)AngScMom correlated positively with r=0.63.

CONCLUSIONS

MRI texture analysis derived from conventional sequences reflects histopathology features in thyroid cancer. This technique might be a novel noninvasive modality to further characterize thyroid cancer in clinical oncology.

Diffusion-weighted MRI and intravoxel incoherent motion model for diagnosis of pediatric solid abdominal tumors

Background

Pediatric retroperitoneal tumors in the renal bed are often large and heterogeneous, and their diagnosis based on conventional imaging alone is not possible. More advanced imaging methods, such as diffusion‐weighted (DW) MRI and the use of intravoxel incoherent motion (IVIM), have the potential to provide additional biomarkers that could facilitate their noninvasive diagnosis.

Purpose

To assess the use of an IVIM model for diagnosis of childhood malignant abdominal tumors and discrimination of benign from malignant lesions.

Study Type

Retrospective.

Population

Forty‐two pediatric patients with abdominal lesions (n = 32 malignant, n = 10 benign), verified by histopathology.

Field Strength/Sequence

1.5T MRI system and a DW‐MRI sequence with six b‐values (0, 50, 100, 150, 600, 1000 s/mm²).

Assessment

Parameter maps of apparent diffusion coefficient (ADC), and IVIM maps of slow diffusion coefficient (D), fast diffusion coefficient (D*), and perfusion fraction (f) were computed using a segmented fitting model. Histograms were constructed for whole‐tumor regions of each parameter.

Statistical Tests

Comparison of histogram parameters of and their diagnostic performance was determined using Kruskal–Wallis, Mann–Whitney U, and receiver‐operating characteristic (ROC) analysis.

Results

IVIM parameters D* and f were significantly higher in neuroblastoma compared to Wilms' tumors (P < 0.05). The ROC analysis showed that the best diagnostic performance was achieved with D* 90^th percentile (area under the curve [AUC] = 0.935; P = 0.002; cutoff value = 32,376 × 10⁻⁶ mm²/s) and f mean values (AUC = 1.00; P < 0.001; cutoff value = 14.7) in discriminating between neuroblastoma (n = 11) and Wilms' tumors (n = 8). Discrimination between tumor types was not possible with IVIM D or ADC parameters. Malignant tumors revealed significantly lower ADC, D, and higher D* values than in benign lesions (all P < 0.05).

Data Conclusion

IVIM perfusion parameters could distinguish between malignant childhood tumor types, providing potential imaging biomarkers for their diagnosis.

Level of Evidence: 4

Technical Efficacy: Stage 2

J. Magn. Reson. Imaging 2018;47:1475–1486.

Histogram analysis of diffusion kurtosis imaging derived maps may distinguish between low and high grade gliomas before surgery

OBJECTIVE

To investigate the value of histogram analysis of diffusion kurtosis imaging (DKI) maps in the evaluation of glioma grading.

METHODS

A total of 39 glioma patients who underwent preoperative magnetic resonance imaging (MRI) were classified into low-grade (13 cases) and high-grade (26 cases) glioma groups. Parametric DKI maps were derived, and histogram metrics between low- and high-grade gliomas were analysed. The optimum diagnostic thresholds of the parameters, area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were achieved using a receiver operating characteristic (ROC).

RESULT

Significant differences were observed not only in 12 metrics of histogram DKI parameters (P<0.05), but also in mean diffusivity (MD) and mean kurtosis (MK) values, including age as a covariate (F=19.127, P<0.001 and F=20.894, P<0.001, respectively), between low- and high-grade gliomas. Mean MK was the best independent predictor of differentiating glioma grades (B=18.934, 22.237 adjusted for age, P<0.05). The partial correlation coefficient between fractional anisotropy (FA) and kurtosis fractional anisotropy (KFA) was 0.675 (P<0.001). The AUC of the mean MK, sensitivity, and specificity were 0.925, 88.5% and 84.6%, respectively.

CONCLUSIONS

DKI parameters can effectively distinguish between low- and high-grade gliomas. Mean MK is the best independent predictor of differentiating glioma grades.

KEY POINTS

• DKI is a new and important method. • DKI can provide additional information on microstructural architecture. • Histogram analysis of DKI may be more effective in glioma grading.

Quantitative Assessment of Heterogeneity in Bladder Tumor MRI Diffusivity: Can Response be Predicted Prior to Neoadjuvant Chemotherapy?

Background:

It is a critical unmet need to predict chemosensitivity in muscle-invasive bladder cancer patients who receive neoadjuvant chemotherapy (NAC). Quantification of tumor heterogeneity has been shown to be useful in the assessment of therapeutic response. Apparent diffusion coefficient (ADC) is derived from diffusion weighted MRI (DWI) to quantify the water diffusivity which characterizes micro-cellularity in tumor tissues.

Objective:

The aim of this study is to assess if a quantitative measurement of ADC heterogeneity in bladder tumors can be a predictor of therapeutic response to NAC.

Materials and Methods:

Twenty patients with pT2 bladder cancer have been included in this study. Patient MRI was performed on a 3T system with DWI prior to NAC. Regions of interest (ROIs) were placed over the whole tumor volume on ADC maps to acquire a data matrix of voxel-wise ADC values for each patient. We performed histogram analysis on each ADC data matrix to calculate uniformity (U) and entropy (E). These quantities were subsequently correlated with the patient’s response to chemotherapy. Statistical significance was found with P < 0.05.

Results:

Fifteen patients were categorized as responders, and five as non-responders. The data showed that tumors of responders were significantly higher in U (P = 0.01) and lower in E (P < 0.01) than non-responders. This finding indicates that resistant tumors were more heterogeneous in their spatial distribution of ADC values. While this difference in ADC heterogeneity was not always visually recognizable, it could be quantified by the data analytics.

Conclusions:

This study demonstrates that the quantitative readout of tumor heterogeneity in micro-cellularity is associated with the patient’s defined response to chemotherapy. Quantification of tumor ADC heterogeneity may provide useful information to enable the prediction of chemotherapeutic response prior to the treatment to improve patient outcomes.

Cancer Metabolism and Tumor Heterogeneity: Imaging Perspectives Using MR Imaging and Spectroscopy

Cancer cells reprogram their metabolism to maintain viability via genetic mutations and epigenetic alterations, expressing overall dynamic heterogeneity. The complex relaxation mechanisms of nuclear spins provide unique and convertible tissue contrasts, making magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) pertinent imaging tools in both clinics and research. In this review, we summarized MR methods that visualize tumor characteristics and its metabolic phenotypes on an anatomical, microvascular, microstructural, microenvironmental, and metabolomics scale. The review will progress from the utilities of basic spin-relaxation contrasts in cancer imaging to more advanced imaging methods that measure tumor-distinctive parameters such as perfusion, water diffusion, magnetic susceptibility, oxygenation, acidosis, redox state, and cell death. Analytical methods to assess tumor heterogeneity are also reviewed in brief. Although the clinical utility of tumor heterogeneity from imaging is debatable, the quantification of tumor heterogeneity using functional and metabolic MR images with development of robust analytical methods and improved MR methods may offer more critical roles of tumor heterogeneity data in clinics. MRI/MRS can also provide insightful information on pharmacometabolomics, biomarker discovery, disease diagnosis and prognosis, and treatment response. With these future directions in mind, we anticipate the widespread utilization of these MR-based techniques in studying in vivo cancer biology to better address significant clinical needs.

Features from Computerized Texture Analysis of Breast Cancers at Pretreatment MR Imaging Are Associated with Response to Neoadjuvant Chemotherapy

Purpose To evaluate whether features from texture analysis of breast cancers were associated with pathologic complete response (pCR) after neoadjuvant chemotherapy and to explore the association between texture features and tumor subtypes at pretreatment magnetic resonance (MR) imaging. Materials and Methods Institutional review board approval was obtained. This retrospective study included 85 patients with 85 breast cancers who underwent breast MR imaging before neoadjuvant chemotherapy between April 10, 2008, and March 12, 2015. Two-dimensional texture analysis was performed by using software at T2-weighted MR imaging and contrast material-enhanced T1-weighted MR imaging. Quantitative parameters were compared between patients with pCR and those with non-pCR and between patients with triple-negative breast cancer and those with non-triple-negative cancer. Multiple logistic regression analysis was used to determine independent parameters. Results Eighteen tumors (22%) were triple-negative breast cancers. pCR was achieved in 30 of the 85 tumors (35%). At univariate analysis, mean pixel intensity with spatial scaling factor (SSF) of 2 and 4 on T2-weighted images and kurtosis on contrast-enhanced T1-weighted images showed a significant difference between triple-negative breast cancer and non-triple-negative breast cancer (P = .009, .003, and .001, respectively). Kurtosis (SSF, 2) on T2-weighted images showed a significant difference between pCR and non-pCR (P = .015). At multiple logistic regression, kurtosis on T2-weighted images was independently associated with pCR in non-triple-negative breast cancer (P = .033). A multivariate model incorporating T2-weighted and contrast-enhanced T1-weighted kurtosis showed good performance for the identification of triple-negative breast cancer (area under the receiver operating characteristic curve, 0.834). Conclusion At pretreatment MR imaging, kurtosis appears to be associated with pCR to neoadjuvant chemotherapy in non-triple-negative breast cancer and may be a promising biomarker for the identification of triple-negative breast cancer. ^© RSNA, 2017.

Apparent Diffusion Coefficient Histograms of Human Papillomavirus-Positive and Human Papillomavirus-Negative Head and Neck Squamous Cell Carcinoma: Assessment of Tumor Heterogeneity and Comparison with Histopathology

BACKGROUND AND PURPOSE

Head and neck squamous cell carcinoma associated with human papillomavirus infection represents a distinct tumor entity. We hypothesized that diffusion phenotypes based on the histogram analysis of ADC values reflect distinct degrees of tumor heterogeneity in human papillomavirus-positive and human papillomavirus-negative head and neck squamous cell carcinomas.

MATERIALS AND METHODS

One hundred five consecutive patients (mean age, 64 years; range, 45-87 years) with primary oropharyngeal (n = 52) and oral cavity (n = 53) head and neck squamous cell carcinoma underwent MR imaging with anatomic and diffusion-weighted sequences (b = 0, b = 1000 s/mm², monoexponential ADC calculation). The collected tumor voxels from the contoured ROIs provided histograms from which position, dispersion, and form parameters were computed. Histogram data were correlated with histopathology, p16-immunohistochemistry, and polymerase chain reaction for human papillomavirus DNA.

RESULTS

There were 21 human papillomavirus-positive and 84 human papillomavirus-negative head and neck squamous cell carcinomas. At histopathology, human papillomavirus-positive cancers were more often nonkeratinizing (13/21, 62%) than human papillomavirus-negative cancers (19/84, 23%; P = .001), and their mitotic index was higher (71% versus 49%; P = .005). ROI-based mean and median ADCs were significantly lower in human papillomavirus-positive (1014 ± 178 × 10^-6 mm²/s and 970 ± 187 × 10^-6 mm²/s, respectively) than in human papillomavirus-negative tumors (1184 ± 168 × 10^-6 mm²/s and 1161 ± 175 × 10^-6 mm²/s, respectively; P < .001), whereas excess kurtosis and skewness were significantly higher in human papillomavirus-positive (1.934 ± 1.386 and 0.923 ± 0.510, respectively) than in human papillomavirus-negative tumors (0.643 ± 0.982 and 0.399 ± 0.516, respectively; P < .001). Human papillomavirus-negative head and neck squamous cell carcinoma had symmetric normally distributed ADC histograms, which corresponded histologically to heterogeneous tumors with variable cellularity, high stromal component, keratin pearls, and necrosis. Human papillomavirus-positive head and neck squamous cell carcinomas had leptokurtic skewed right histograms, which corresponded to homogeneous tumors with back-to-back densely packed cells, scant stromal component, and scattered comedonecrosis.

CONCLUSIONS

Diffusion phenotypes of human papillomavirus-positive and human papillomavirus-negative head and neck squamous cell carcinomas show significant differences, which reflect their distinct degree of tumor heterogeneity.

ADC-derived spatial features can accurately classify adnexal lesions

BACKGROUND

The role of quantitative apparent diffusion coefficient (ADC) maps in differentiating adnexal masses is unresolved.

PURPOSE/HYPOTHESIS

To propose an objective diagnostic method devised based on spatial features for predicting benignity/malignancy of adnexal masses in ADC maps.

STUDY TYPE

Prospective.

POPULATION

In all, 70 women with sonographically indeterminate and histopathologically confirmed adnexal masses (38 benign, 3 borderline, and 29 malignant) were considered for this study.

FIELD STRENGTH/SEQUENCE

Conventional and diffusion-weighted magnetic resonance (MR) images (b-values = 50, 400, 1000 s/mm² ) were acquired on a 3T scanner.

ASSESSMENT

For each patient, two radiologists in consensus manually delineated lesion borders in whole ADC map volumes, which were consequently analyzed using spatial models (first-order histogram [FOH], gray-level co-occurrence matrix [GLCM], run-length matrix [RLM], and Gabor filters). Two independent radiologists were asked to identify the attributed (benign/malignant) classes of adnexal masses based on morphological features on conventional MRI.

STATISTICAL TESTS

Leave-one-out cross-validated feature selection followed by cross-validated classification were applied to the feature space to choose the spatial models that best discriminate benign from malignant adnexal lesions. Two schemes of feature selection/classification were evaluated: 1) including all benign and malignant masses, and 2) scheme 1 after excluding endometrioma, hemorrhagic cysts, and teratoma (14 benign, 29 malignant masses). The constructed feature subspaces for benign/malignant lesion differentiation were tested for classification of benign/borderline/malignant and also borderline/malignant adnexal lesions.

RESULTS

The selected feature subspace consisting of RLM features differentiated benign from malignant adnexal masses with a classification accuracy of ∼92%. The same model discriminated benign, borderline, and malignant lesions with 87% and borderline from malignant with 100% accuracy. Qualitative assessment of the radiologists based on conventional MRI features reached an accuracy of 80%.

DATA CONCLUSION

The spatial quantification methodology proposed in this study, which works based on cellular distributions within ADC maps of adnexal masses, may provide a helpful computer-aided strategy for objective characterization of adnexal masses.

LEVEL OF EVIDENCE

1 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:1061-1071.

Monitoring In-Vivo the Mammary Gland Microstructure during Morphogenesis from Lactation to Post-Weaning Using Diffusion Tensor MRI

Lactation and the return to the pre-conception state during post-weaning are regulated by hormonal induced processes that modify the microstructure of the mammary gland, leading to changes in the features of the ductal / glandular tissue, the stroma and the fat tissue. These changes create a challenge in the radiological workup of breast disorder during lactation and early post-weaning. Here we present non-invasive MRI protocols designed to record in vivo high spatial resolution, T₂-weighted images and diffusion tensor images of the entire mammary gland. Advanced imaging processing tools enabled tracking the changes in the anatomical and microstructural features of the mammary gland from the time of lactation to post-weaning. Specifically, by using diffusion tensor imaging (DTI) it was possible to quantitatively distinguish between the ductal / glandular tissue distention during lactation and the post-weaning involution. The application of the T₂-weighted imaging and DTI is completely safe, non-invasive and uses intrinsic contrast based on differences in transverse relaxation rates and water diffusion rates in various directions, respectively. This study provides a basis for further in-vivo monitoring of changes during the mammary developmental stages, as well as identifying changes due to malignant transformation in patients with pregnancy associated breast cancer (PABC).

Relationship between Glioblastoma Heterogeneity and Survival Time: An MR Imaging Texture Analysis

BACKGROUND AND PURPOSE

The heterogeneity of glioblastoma contributes to the poor and variant prognosis. The aim of this retrospective study was to assess the glioblastoma heterogeneity with MR imaging textures and to evaluate its impact on survival time.

MATERIALS AND METHODS

A total of 133 patients with primary glioblastoma who underwent postcontrast T1-weighted imaging (acquired before treatment) and whose data were filed with the survival times were selected from the Cancer Genome Atlas. On the basis of overall survival, the patients were divided into 2 groups: long-term (≥12 months, n = 67) and short-term (<12 months, n = 66) survival. To measure heterogeneity, we extracted 3 types of textures, co-occurrence matrix, run-length matrix, and histogram, reflecting local, regional, and global spatial variations, respectively. Then the support vector machine classification was used to determine how different texture types perform in differentiating the 2 groups, both alone and in combination. Finally, a recursive feature-elimination method was used to find an optimal feature subset with the best differentiation performance.

RESULTS

When used alone, the co-occurrence matrix performed best, while all the features combined obtained the best survival stratification. According to feature selection and ranking, 43 top-ranked features were selected as the optimal subset. Among them, the top 10 features included 7 run-length matrix and 3 co-occurrence matrix features, in which all 6 regional run-length matrix features emphasizing high gray-levels ranked in the top 7.

CONCLUSIONS

The results suggest that local and regional heterogeneity may play an important role in the survival stratification of patients with glioblastoma.

Textural features of dynamic contrast-enhanced MRI derived model-free and model-based parameter maps in glioma grading

BACKGROUND

Presurgical glioma grading by dynamic contrast-enhanced MRI (DCE-MRI) has unresolved issues.

PURPOSE

The aim of this study was to investigate the ability of textural features derived from pharmacokinetic model-based or model-free parameter maps of DCE-MRI in discriminating between different grades of gliomas, and their correlation with pathological index.

STUDY TYPE

Retrospective.

SUBJECTS

Forty-two adults with brain gliomas.

FIELD STRENGTH/SEQUENCE

3.0T, including conventional anatomic sequences and DCE-MRI sequences (variable flip angle T1-weighted imaging and three-dimensional gradient echo volumetric imaging).

ASSESSMENT

Regions of interest on the cross-sectional images with maximal tumor lesion. Five commonly used textural features, including Energy, Entropy, Inertia, Correlation, and Inverse Difference Moment (IDM), were generated.

RESULTS

All textural features of model-free parameters (initial area under curve [IAUC], maximal signal intensity [Max SI], maximal up-slope [Max Slope]) could effectively differentiate between grade II (n = 15), grade III (n = 13), and grade IV (n = 14) gliomas (P < 0.05). Two textural features, Entropy and IDM, of four DCE-MRI parameters, including Max SI, Max Slope (model-free parameters), vp (Extended Tofts), and vp (Patlak) could differentiate grade III and IV gliomas (P < 0.01) in four measurements. Both Entropy and IDM of Patlak-based K^trans and vp could differentiate grade II (n = 15) from III (n = 13) gliomas (P < 0.01) in four measurements. No textural features of any DCE-MRI parameter maps could discriminate between subtypes of grade II and III gliomas (P < 0.05). Both Entropy and IDM of Extended Tofts- and Patlak-based vp showed highest area under curve in discriminating between grade III and IV gliomas. However, intraclass correlation coefficient (ICC) of these features revealed relatively lower inter-observer agreement. No significant correlation was found between microvascular density and textural features, compared with a moderate correlation found between cellular proliferation index and those features.

DATA CONCLUSION

Textural features of DCE-MRI parameter maps displayed a good ability in glioma grading.

LEVEL OF EVIDENCE

3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:1099-1111.

Statistical Tests for Large Tree-Structured Data

We develop a general statistical framework for the analysis and inference of large tree-structured data, with a focus on developing asymptotic goodness-of-fit tests. We first propose a consistent statistical model for binary trees, from which we develop a class of invariant tests. Using the model for binary trees, we then construct tests for general trees by using the distributional properties of the Continuum Random Tree, which arises as the invariant limit for a broad class of models for tree-structured data based on conditioned Galton-Watson processes. The test statistics for the goodness-of-fit tests are simple to compute and are asymptotically distributed as χ ² and F random variables. We illustrate our methods on an important application of detecting tumour heterogeneity in brain cancer. We use a novel approach with tree-based representations of magnetic resonance images and employ the developed tests to ascertain tumor heterogeneity between two groups of patients.

Lack of robustness of textural measures obtained from 3D brain tumor MRIs impose a need for standardization

Purpose

Textural measures have been widely explored as imaging biomarkers in cancer. However, their robustness under dynamic range and spatial resolution changes in brain 3D magnetic resonance images (MRI) has not been assessed. The aim of this work was to study potential variations of textural measures due to changes in MRI protocols.

Materials and methods

Twenty patients harboring glioblastoma with pretreatment 3D T1-weighted MRIs were included in the study. Four different spatial resolution combinations and three dynamic ranges were studied for each patient. Sixteen three-dimensional textural heterogeneity measures were computed for each patient and configuration including co-occurrence matrices (CM) features and run-length matrices (RLM) features. The coefficient of variation was used to assess the robustness of the measures in two series of experiments corresponding to (i) changing the dynamic range and (ii) changing the matrix size.

Results

No textural measures were robust under dynamic range changes. Entropy was the only textural feature robust under spatial resolution changes (coefficient of variation under 10% in all cases).

Conclusion

Textural measures of three-dimensional brain tumor images are not robust neither under dynamic range nor under matrix size changes. Standards should be harmonized to use textural features as imaging biomarkers in radiomic-based studies. The implications of this work go beyond the specific tumor type studied here and pose the need for standardization in textural feature calculation of oncological images.

From the Cover: MagneticResonance Imaging Reveals Progressive Brain Injury in Rats Acutely Intoxicated With Diisopropylfluorophosphate

Acute intoxication with organophosphates (OPs) can trigger seizures that progress to status epilepticus, and survivors often exhibit chronic neuropathology, cognitive impairment, affective disorders, and/or electroencephalographic abnormalities. Understanding how acute injury transitions to persistent neurological sequelae is critical to developing medical countermeasures for mitigating damage following OP-induced seizures. Here, we used in vivo magnetic resonance imaging (MRI) to monitor the spatiotemporal patterns of neuropathology for 1 month after acute intoxication with diisopropylfluorophosphate (DFP). Adult male Sprague Dawley rats administered pyridostigmine bromide (0.1 mg/kg, im) 30 min prior to successive administration of DFP (4 mg/kg, sc), atropine sulfate (2 mg/kg, im), and 2-pralidoxime (25 mg/kg, im) exhibited moderate-to-severe seizure behavior. T2-weighted and diffusion-weighted MR imaging prior to DFP exposure and at 3, 7, 14, 21, or 28 days postexposure revealed prominent lesions, tissue atrophy, and ventricular enlargement in discrete brain regions. Lesions varied in intensity and/or extent over time, with the overall magnitude of injury strongly influenced by seizure severity. Importantly, lesions detected by MRI correlated spatially and temporally with histological evidence of brain pathology. Analysis of histogram parameters extracted from frequency distributions of regional apparent diffusion coefficient (ADC) values identified the standard deviation and 90th percentile of the ADC as robust metrics for quantifying persistent and progressive neuropathological changes. The interanimal and interregional variations observed in lesion severity and progression, coupled with potential reinjury following spontaneous recurrent seizures, underscore the advantages of using in vivo imaging to longitudinally monitor neuropathology and, ultimately, therapeutic response, following acute OP intoxication.

Quantification of hepatocellular carcinoma heterogeneity with multiparametric magnetic resonance imaging

Tumour heterogeneity poses a significant challenge for treatment stratification. The goals of this study were to quantify heterogeneity in hepatocellular carcinoma (HCC) using multiparametric magnetic resonance imaging (mpMRI), and to report preliminary data correlating quantitative MRI parameters with advanced histopathology and gene expression in a patient subset. Thirty-two HCC patients with 39 HCC lesions underwent mpMRI including diffusion-weighted imaging (DWI), blood-oxygenation-level-dependent (BOLD), tissue-oxygenation-level-dependent (TOLD) and dynamic contrast-enhanced (DCE)-MRI. Histogram characteristics [central tendency (mean, median) and heterogeneity (standard deviation, kurtosis, skewness) MRI parameters] in HCC and liver parenchyma were compared using Wilcoxon signed-rank tests. Histogram data was correlated between MRI methods in all patients and with histopathology and gene expression in 14 patients. HCCs exhibited significantly higher intra-tissue heterogeneity vs. liver with all MRI methods (P < 0.030). Although central tendency parameters showed significant correlations between MRI methods and with each of histopathology and gene expression, heterogeneity parameters exhibited additional complementary correlations between BOLD and DCE-MRI and with histopathologic hypoxia marker HIF1α and gene expression of Wnt target GLUL, pharmacological target FGFR4, stemness markers EPCAM and KRT19 and immune checkpoint PDCD1. Histogram analysis combining central tendency and heterogeneity mpMRI features is promising for non-invasive HCC characterization on the imaging, histologic and genomics levels.

Whole-volume apparent diffusion coefficient-based entropy parameters for assessment of gastric cancer aggressiveness

PURPOSE

To explore the role of whole-volume apparent diffusion coefficient (ADC)-based entropy parameters in the preoperative assessment of gastric cancer's aggressiveness.

MATERIALS AND METHODS

In all, 64 patients with gastric cancers who underwent 3.0T magnetic resonance imaging (MRI) were retrospectively included. Regions of interest were drawn manually using in-house software, around gastric cancer lesions on each slice of the diffusion-weighted images and ADC maps. Entropy-related parameters based on ADC maps were calculated automatically: (1) first-order entropy; (2-5) second-order entropies, including entropy(H)₀ , entropy(H)₄₅ , entropy(H)₉₀ , and entropy(H)₁₃₅ ; (6) entropy(H)_mean ; and (7) entropy(H)_range . Correlations between entropy-related parameters and pathological characteristics were analyzed with the Spearman correlation test. The parameters were compared among different pathological characteristics with independent-samples Kruskal-Wallis or Mann-Whitney U-test. Additionally, diagnostic performances of parameters in differentiating different pathological characteristics were analyzed by receiver operating characteristic (ROC) curve analysis.

RESULTS

All the entropy-related parameters significantly correlated with T, N, and overall stages, especially the first-order entropy (r = 0.588, 0.585, and 0.677, respectively, all P < 0.05). All the entropy-related parameters showed significant differences in gastric cancers at different T, N, and overall stages, as well as at different status of vascular invasion (P < 0.001-0.027). And four parameters, including entropy, entropy(H)₀ , entropy(H)₄₅ , and entropy(H)₉₀ , showed significant differences between gastric cancers with and without perineural invasion (P 0.006-0.040).

CONCLUSION

Entropy-related parameters derived from whole-volume ADC texture analysis could help assess the aggressiveness of gastric cancers via analyzing intratumoral heterogeneity quantitatively, especially the first-order entropy.

LEVEL OF EVIDENCE

2 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2018;47:168-175.

Histogram Analysis of Diffusion Weighted Imaging at 3T is Useful for Prediction of Lymphatic Metastatic Spread, Proliferative Activity, and Cellularity in Thyroid Cancer

Pre-surgical diffusion weighted imaging (DWI) is increasingly important in the context of thyroid cancer for identification of the optimal treatment strategy. It has exemplarily been shown that DWI at 3T can distinguish undifferentiated from well-differentiated thyroid carcinoma, which has decisive implications for the magnitude of surgery. This study used DWI histogram analysis of whole tumor apparent diffusion coefficient (ADC) maps. The primary aim was to discriminate thyroid carcinomas which had already gained the capacity to metastasize lymphatically from those not yet being able to spread via the lymphatic system. The secondary aim was to reflect prognostically important tumor-biological features like cellularity and proliferative activity with ADC histogram analysis. Fifteen patients with follicular-cell derived thyroid cancer were enrolled. Lymph node status, extent of infiltration of surrounding tissue, and Ki-67 and p53 expression were assessed in these patients. DWI was obtained in a 3T system using b values of 0, 400, and 800 s/mm². Whole tumor ADC volumes were analyzed using a histogram-based approach. Several ADC parameters showed significant correlations with immunohistopathological parameters. Most importantly, ADC histogram skewness and ADC histogram kurtosis were able to differentiate between nodal negative and nodal positive thyroid carcinoma. Conclusions: histogram analysis of whole ADC tumor volumes has the potential to provide valuable information on tumor biology in thyroid carcinoma. However, further studies are warranted.

Clinical Applications of Contrast-Enhanced Perfusion MRI Techniques in Gliomas: Recent Advances and Current Challenges

Gliomas possess complex and heterogeneous vasculatures with abnormal hemodynamics. Despite considerable advances in diagnostic and therapeutic techniques for improving tumor management and patient care in recent years, the prognosis of malignant gliomas remains dismal. Perfusion-weighted magnetic resonance imaging techniques that could noninvasively provide superior information on vascular functionality have attracted much attention for evaluating brain tumors. However, nonconsensus imaging protocols and postprocessing analysis among different institutions impede their integration into standard-of-care imaging in clinic. And there have been very few studies providing a comprehensive evidence-based and systematic summary. This review first outlines the status of glioma theranostics and tumor-associated vascular pathology and then presents an overview of the principles of dynamic contrast-enhanced MRI (DCE-MRI) and dynamic susceptibility contrast-MRI (DSC-MRI), with emphasis on their recent clinical applications in gliomas including tumor grading, identification of molecular characteristics, differentiation of glioma from other brain tumors, treatment response assessment, and predicting prognosis. Current challenges and future perspectives are also highlighted.

Semiautomated Workflow for Clinically Streamlined Glioma Parametric Response Mapping

Management of glioblastoma multiforme remains a challenging problem despite recent advances in targeted therapies. Timely assessment of therapeutic agents is hindered by the lack of standard quantitative imaging protocols for determining targeted response. Clinical response assessment for brain tumors is determined by volumetric changes assessed at 10 weeks post-treatment initiation. Further, current clinical criteria fail to use advanced quantitative imaging approaches, such as diffusion and perfusion magnetic resonance imaging. Development of the parametric response mapping (PRM) applied to diffusion-weighted magnetic resonance imaging has provided a sensitive and early biomarker of successful cytotoxic therapy in brain tumors while maintaining a spatial context within the tumor. Although PRM provides an earlier readout than volumetry and sometimes greater sensitivity compared with traditional whole-tumor diffusion statistics, it is not routinely used for patient management; an automated and standardized software for performing the analysis and for the generation of a clinical report document is required for this. We present a semiautomated and seamless workflow for image coregistration, segmentation, and PRM classification of glioblastoma multiforme diffusion-weighted magnetic resonance imaging scans. The software solution can be integrated using local hardware or performed remotely in the cloud while providing connectivity to existing picture archive and communication systems. This is an important step toward implementing PRM analysis of solid tumors in routine clinical practice.

Functional MRI for quantitative treatment response prediction in locally advanced rectal cancer

Despite advances in multimodality treatment strategies for locally advanced rectal cancer and improvements in locoregional control, there is still a considerable variation in response to neoadjuvant chemoradiotherapy (CRT). Accurate prediction of response to neoadjuvant CRT would enable early stratification of management according to good responders and poor responders, in order to adapt treatment to improve therapeutic outcomes in rectal cancer. Clinical studies in diffusion-weighted imaging (DWI) and dynamic contrast-enhanced (DCE) MRI have shown promising results for the prediction of therapeutic response in rectal cancer. DWI allows for assessment of tumour cellularity. DCE-MRI enables evaluation of factors of the tumour microvascular environment and changes in perfusion in response to treatment. Studies have demonstrated that predictors of good response to CRT include lower tumour pre-CRT apparent diffusion coefficient (ADC), greater percentage increase in ADC during and post CRT, and higher pre-CRT K^trans. However, the mean ADC and K^trans values do not adequately reflect tumour heterogeneity. Multiparametric MRI using quantitative DWI and DCE-MRI in combination, and a histogram analysis technique can assess tumour heterogeneity and its response to treatment. This strategy has the potential to improve the accuracy of therapeutic response prediction in rectal cancer and warrants further investigation.

Diffusion MRI in early cancer therapeutic response assessment

Imaging biomarkers for the predictive assessment of treatment response in patients with cancer earlier than standard tumor volumetric metrics would provide new opportunities to individualize therapy. Diffusion-weighted MRI (DW-MRI), highly sensitive to microenvironmental alterations at the cellular level, has been evaluated extensively as a technique for the generation of quantitative and early imaging biomarkers of therapeutic response and clinical outcome. First demonstrated in a rodent tumor model, subsequent studies have shown that DW-MRI can be applied to many different solid tumors for the detection of changes in cellularity as measured indirectly by an increase in the apparent diffusion coefficient (ADC) of water molecules within the lesion. The introduction of quantitative DW-MRI into the treatment management of patients with cancer may aid physicians to individualize therapy, thereby minimizing unnecessary systemic toxicity associated with ineffective therapies, saving valuable time, reducing patient care costs and ultimately improving clinical outcome. This review covers the theoretical basis behind the application of DW-MRI to monitor therapeutic response in cancer, the analytical techniques used and the results obtained from various clinical studies that have demonstrated the efficacy of DW-MRI for the prediction of cancer treatment response. Copyright © 2016 John Wiley & Sons, Ltd.

Whole-lesion histogram analysis metrics of the apparent diffusion coefficient as a marker of breast lesions characterization at 1.5 T

INTRODUCTION

To retrospectively assess the role of whole-lesion apparent diffusion coefficient (ADC) in the characterization of breast tumors by comparing different histogram metrics.

METHODS

49 patients with 53 breast lesions underwent magnetic resonance imaging (MRI). ADC histogram parameters, including the mean, mode, 10th/50th/90th percentile, skewness, kurtosis, and entropy ADCs, were derived for the whole-lesion volume in each patient. Mann-Whitney U-test, area under the receiver-operating characteristic curve (AUC) were used for statistical analysis.

RESULTS

The mean, mode and 10th/50th/90th percentile ADC values were significantly lower in malignant lesions compared with benign ones (all P < 0.0001), while skewness was significantly higher in malignant lesions P = 0.02. However, no significant difference was found between entropy and kurtosis values in malignant lesions compared with benign ones (P = 0.06 and P = 1.00, respectively). Univariate logistic regression showed that 10th and 50th percentile ADC yielded the highest AUC (0.985; 95% confidence interval [CI]: 0.902, 1.000 and 0.982; 95% confidence interval [CI]: 0.896, 1.000 respectively), whereas kurtosis value yielded the lowest AUC (0.500; 95% CI: 0.355, 0.645), indicating that 10th and 50th percentile ADC values may be more accurate for lesion discrimination.

CONCLUSION

Whole-lesion ADC histogram analysis could be a helpful index in the characterization and differentiation between benign and malignant breast lesions with the 10th and 50th percentile ADC be the most accurate discriminators.

Non-small cell lung cancer: Whole-lesion histogram analysis of the apparent diffusion coefficient for assessment of tumor grade, lymphovascular invasion and pleural invasion

PURPOSE

Investigating the diagnostic accuracy of histogram analyses of apparent diffusion coefficient (ADC) values for determining non-small cell lung cancer (NSCLC) tumor grades, lymphovascular invasion, and pleural invasion.

MATERIALS AND METHODS

We studied 60 surgically diagnosed NSCLC patients. Diffusion-weighted imaging (DWI) was performed in the axial plane using a navigator-triggered single-shot, echo-planar imaging sequence with prospective acquisition correction. The ADC maps were generated, and we placed a volume-of-interest on the tumor to construct the whole-lesion histogram. Using the histogram, we calculated the mean, 5th, 10th, 25th, 50th, 75th, 90th, and 95th percentiles of ADC, skewness, and kurtosis. Histogram parameters were correlated with tumor grade, lymphovascular invasion, and pleural invasion. We performed a receiver operating characteristics (ROC) analysis to assess the diagnostic performance of histogram parameters for distinguishing different pathologic features.

RESULTS

The ADC mean, 10th, 25th, 50th, 75th, 90th, and 95th percentiles showed significant differences among the tumor grades. The ADC mean, 25th, 50th, 75th, 90th, and 95th percentiles were significant histogram parameters between high- and low-grade tumors. The ROC analysis between high- and low-grade tumors showed that the 95th percentile ADC achieved the highest area under curve (AUC) at 0.74. Lymphovascular invasion was associated with the ADC mean, 50th, 75th, 90th, and 95th percentiles, skewness, and kurtosis. Kurtosis achieved the highest AUC at 0.809. Pleural invasion was only associated with skewness, with the AUC of 0.648.

CONCLUSIONS

ADC histogram analyses on the basis of the entire tumor volume are able to stratify NSCLCs' tumor grade, lymphovascular invasion and pleural invasion.

Radiomic analysis reveals DCE-MRI features for prediction of molecular subtypes of breast cancer

The purpose of this study was to investigate the role of features derived from breast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and to incorporated clinical information to predict the molecular subtypes of breast cancer. In particular, 60 breast cancers with the following four molecular subtypes were analyzed: luminal A, luminal B, human epidermal growth factor receptor-2 (HER2)-over-expressing and basal-like. The breast region was segmented and the suspicious tumor was depicted on sequentially scanned MR images from each case. In total, 90 features were obtained, including 88 imaging features related to morphology and texture as well as dynamic features from tumor and background parenchymal enhancement (BPE) and 2 clinical information-based parameters, namely, age and menopausal status. An evolutionary algorithm was used to select an optimal subset of features for classification. Using these features, we trained a multi-class logistic regression classifier that calculated the area under the receiver operating characteristic curve (AUC). The results of a prediction model using 24 selected features showed high overall classification performance, with an AUC value of 0.869. The predictive model discriminated among the luminal A, luminal B, HER2 and basal-like subtypes, with AUC values of 0.867, 0.786, 0.888 and 0.923, respectively. An additional independent dataset with 36 patients was utilized to validate the results. A similar classification analysis of the validation dataset showed an AUC of 0.872 using 15 image features, 10 of which were identical to those from the first cohort. We identified clinical information and 3D imaging features from DCE-MRI as candidate biomarkers for discriminating among four molecular subtypes of breast cancer.

Multimodal MRI features predict isocitrate dehydrogenase genotype in high-grade gliomas

BACKGROUND

High-grade gliomas with mutations in the isocitrate dehydrogenase (IDH) gene family confer longer overall survival relative to their IDH-wild-type counterparts. Accurate determination of the IDH genotype preoperatively may have both prognostic and diagnostic value. The current study used a machine-learning algorithm to generate a model predictive of IDH genotype in high-grade gliomas based on clinical variables and multimodal features extracted from conventional MRI.

METHODS

Preoperative MRIs were obtained for 120 patients with primary grades III (n = 35) and IV (n = 85) glioma in this retrospective study. IDH genotype was confirmed for grade III (32/35, 91%) and IV (22/85, 26%) tumors by immunohistochemistry, spectrometry-based mutation genotyping (OncoMap), or multiplex exome sequencing (OncoPanel). IDH1 and IDH2 mutations were mutually exclusive, and all mutated tumors were collapsed into one IDH-mutated cohort. Cases were randomly assigned to either the training (n = 90) or validation cohort (n = 30). A total of 2970 imaging features were extracted from pre- and postcontrast T1-weighted, T2-weighted, and apparent diffusion coefficient map. Using a random forest algorithm, nonredundant features were integrated with clinical data to generate a model predictive of IDH genotype.

RESULTS

Our model achieved accuracies of 86% (area under the curve [AUC] = 0.8830) in the training cohort and 89% (AUC = 0.9231) in the validation cohort. Features with the highest predictive value included patient age as well as parametric intensity, texture, and shape features.

CONCLUSION

Using a machine-learning algorithm, we achieved accurate prediction of IDH genotype in high-grade gliomas with preoperative clinical and MRI features.

Dynamic Contrast-Enhanced MRI Perfusion Parameters as Imaging Biomarkers of Angiogenesis

Hypoxia in the tumor microenvironment is the leading factor in angiogenesis. Angiogenesis can be identified by dynamic contrast-enhanced breast MRI (DCE MRI). Here we investigate the relationship between perfusion parameters on DCE MRI and angiogenic and prognostic factors in patients with invasive ductal carcinoma (IDC). Perfusion parameters (K^trans, k_ep and v_e) of 81 IDC were obtained using histogram analysis. Twenty-fifth, 50th and 75th percentile values were calculated and were analyzed for association with microvessel density (MVD), vascular endothelial growth factor (VEGF) and conventional prognostic factors. Correlation between MVD and v_e50 was positive (r = 0.33). K^trans₅₀ was higher in tumors larger than 2 cm than in tumors smaller than 2 cm. In multivariate analysis, K^trans₅₀ was affected by tumor size and MVD with 12.8% explanation. There was significant association between K^trans₅₀ and tumor size and MVD. Therefore we conclude that DCE MRI perfusion parameters are potential imaging biomarkers for prediction of tumor angiogenesis and aggressiveness.

Whole-Lesion Apparent Diffusion Coefficient-Based Entropy-Related Parameters for Characterizing Cervical Cancers: Initial Findings

RATIONALE AND OBJECTIVES

This study aimed to develop whole-lesion apparent diffusion coefficient (ADC)-based entropy-related parameters of cervical cancer to preliminarily assess intratumoral heterogeneity of this lesion in comparison to adjacent normal cervical tissues.

MATERIALS AND METHODS

A total of 51 women (mean age, 49 years) with cervical cancers confirmed by biopsy underwent 3-T pelvic diffusion-weighted magnetic resonance imaging with b values of 0 and 800 s/mm² prospectively. ADC-based entropy-related parameters including first-order entropy and second-order entropies were derived from the whole tumor volume as well as adjacent normal cervical tissues. Intraclass correlation coefficient, Wilcoxon test with Bonferroni correction, Kruskal-Wallis test, and receiver operating characteristic curve were used for statistical analysis.

RESULTS

All the parameters showed excellent interobserver agreement (all intraclass correlation coefficients  > 0.900). Entropy, entropy(H)₀, entropy(H)₄₅, entropy(H)₉₀, entropy(H)₁₃₅, and entropy(H)_mean were significantly higher, whereas entropy(H)_range and entropy(H)_std were significantly lower in cervical cancers compared to adjacent normal cervical tissues (all P <.0001). Kruskal-Wallis test showed that there were no significant differences among the values of various second-order entropies including entropy(H)_0, entropy(H)₄₅, entropy(H)₉₀, entropy(H)₁₃₅, and entropy(H)_mean. All second-order entropies had larger area under the receiver operating characteristic curve than first-order entropy in differentiating cervical cancers from adjacent normal cervical tissues. Further, entropy(H)₄₅, entropy(H)₉₀, entropy(H)₁₃₅, and entropy(H)_mean had the same largest area under the receiver operating characteristic curve of 0.867.

CONCLUSION

Whole-lesion ADC-based entropy-related parameters of cervical cancers were developed successfully, which showed initial potential in characterizing intratumoral heterogeneity in comparison to adjacent normal cervical tissues.

Exploring the complementarity of THz pulse imaging and DCE-MRIs: Toward a unified multi-channel classification and a deep learning framework

We provide a comprehensive account of recent advances in biomedical image analysis and classification from two complementary imaging modalities: terahertz (THz) pulse imaging and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). The work aims to highlight underlining commonalities in both data structures so that a common multi-channel data fusion framework can be developed. Signal pre-processing in both datasets is discussed briefly taking into consideration advances in multi-resolution analysis and model based fractional order calculus system identification. Developments in statistical signal processing using principal component and independent component analysis are also considered. These algorithms have been developed independently by the THz-pulse imaging and DCE-MRI communities, and there is scope to place them in a common multi-channel framework to provide better software standardization at the pre-processing de-noising stage. A comprehensive discussion of feature selection strategies is also provided and the importance of preserving textural information is highlighted. Feature extraction and classification methods taking into consideration recent advances in support vector machine (SVM) and extreme learning machine (ELM) classifiers and their complex extensions are presented. An outlook on Clifford algebra classifiers and deep learning techniques suitable to both types of datasets is also provided. The work points toward the direction of developing a new unified multi-channel signal processing framework for biomedical image analysis that will explore synergies from both sensing modalities for inferring disease proliferation.

Prognostic value of preoperative dynamic contrast-enhanced MRI perfusion parameters for high-grade glioma patients

Introduction

The prognostic value of the dynamic contrast-enhanced (DCE) MRI perfusion and its histogram analysis-derived metrics is not well established for high-grade glioma (HGG) patients. The aim of this prospective study was to investigate DCE perfusion transfer coefficient (K^trans), vascular plasma volume fraction (v_p), extracellular volume fraction (v_e), reverse transfer constant (k_ep), and initial area under gadolinium concentration time curve (IAUGC) as predictors of progression-free (PFS) and overall survival (OS) in HGG patients.

Methods

Sixty-nine patients with suspected anaplastic astrocytoma or glioblastoma underwent preoperative DCE-MRI scans. DCE perfusion whole tumor region histogram parameters, clinical details, and PFS and OS data were obtained. Univariate, multivariate, and Kaplan–Meier survival analyses were conducted. Receiver operating characteristic (ROC) curve analysis was employed to identify perfusion parameters with the best differentiation performance.

Results

On univariate analysis, v_e and skewness of v_p had significant negative impacts, while k_ep had significant positive impact on OS (P < 0.05). v_e was also a negative predictor of PFS (P < 0.05). Patients with lower v_e and IAUGC had longer median PFS and OS on Kaplan–Meier analysis (P < 0.05). K^trans and v_e could also differentiate grade III from IV gliomas (area under the curve 0.819 and 0.791, respectively).

Conclusions

High v_e is a consistent predictor of worse PFS and OS in HGG glioma patients. v_p skewness and k_ep are also predictive for OS. K^trans and v_e demonstrated the best diagnostic performance for differentiating grade III from IV gliomas.

Whole-lesion apparent diffusion coefficient (ADC) metrics as a marker of breast tumour characterization-comparison between ADC value and ADC entropy

OBJECTIVE

To prospectively assess the role of whole-lesion apparent diffusion coefficient (ADC) metrics in the characterization of breast tumours by comparing ADC value with ADC entropy.

METHODS

49 patients with 53 breast lesions underwent phased-array breast coil 1.5-T MRI. Two radiologists experienced in breast MRI, blinded to the final diagnosis, reviewed the ADC maps and placed a volume of interest on all slices including each lesion on the ADC map to obtain whole-lesion mean ADC value and ADC entropy. The mean ADC value and ADC entropy in benign and malignant lesions were compared by the Mann-Whitney U-test. Receiver-operating characteristic analysis was performed to assess the sensitivity and specificity of the two variables in the characterization of the breast lesions.

RESULTS

The benign (n = 19) and malignant lesions (n = 34) had mean diameters of 20.8 mm (10.1-31.5 mm) and 26.4 mm (10.5-42.3 mm), respectively. The mean ADC value of the malignant lesions was significantly lower than that of the benign ones (0.87 × 10^-3 vs 1.49 × 10^-3 mm² s^-1; p < 0.0001). Malignant ADC entropy was higher than benign entropy, without reaching levels of statistical significance (5.4 vs 5.0; p = 0.064). At a mean ADC cut-off value of 1.16 × 10^-3 mm² s^-1, the sensitivity and specificity for diagnosing malignancy became optimal (97.1% and 93.7, respectively) with an area under the curve (AUC) of 0.975. With regard to ADC entropy, the sensitivity and specificity at a cut-off of 5.18 were 67.6 and 68.7%, respectively, with an AUC of 0.664.

CONCLUSION

Whole-lesion mean ADC could be a helpful index in the characterization of suspicious breast lesions, with higher sensitivity and specificity than ADC entropy. Advances in knowledge: Two separate parameters of the whole-lesion histogram were compared for their diagnostic accuracy in characterizing breast lesions. Mean ADC was found to be able to characterize breast lesions, whereas entropy proved to be unable to differentiate benign from malignant breast lesions. It is, however, likely that entropy may distinguish these two groups if a larger cohort were used, or the fact that this may be influenced by the molecular subtypes of breast cancers included.

MRI Evaluation of Non-Necrotic T2-Hyperintense Foci in Pediatric Diffuse Intrinsic Pontine Glioma

BACKGROUND AND PURPOSE

The conventional MR imaging appearance of diffuse intrinsic pontine glioma suggests intralesional histopathologic heterogeneity, and various distinct lesion components, including T2-hypointense foci, have been described. Here we report the prevalence, conventional MR imaging semiology, and advanced MR imaging features of non-necrotic T2-hyperintense foci in diffuse intrinsic pontine glioma.

MATERIALS AND METHODS

Twenty-five patients with diffuse intrinsic pontine gliomas were included in this study. MR imaging was performed at 3T by using conventional and advanced MR imaging sequences. Perfusion (CBV), vascular permeability (v_e, K^trans), and diffusion (ADC) metrics were calculated and used to characterize non-necrotic T2-hyperintense foci in comparison with other lesion components, namely necrotic T2-hyperintense foci, T2-hypointense foci, peritumoral edema, and normal brain stem. Statistical analysis was performed by using Kruskal-Wallis and Wilcoxon rank sum tests.

RESULTS

Sixteen non-necrotic T2-hyperintense foci were found in 12 tumors. In these foci, ADC values were significantly higher than those in either T2-hypointense foci (P = .002) or normal parenchyma (P = .0002), and relative CBV values were significantly lower than those in either T2-hypointense (P = .0002) or necrotic T2-hyperintense (P = .006) foci. Volume transfer coefficient values in T2-hyperintense foci were lower than those in T2-hypointense (P = .0005) or necrotic T2-hyperintense (P = .0348) foci.

CONCLUSIONS

Non-necrotic T2-hyperintense foci are common, distinct lesion components within diffuse intrinsic pontine gliomas. Advanced MR imaging data suggest low cellularity and an early stage of angioneogenesis with leaky vessels resulting in expansion of the extracellular space. Because of the lack of biopsy validation, the underlying histoarchitectural and pathophysiologic changes remain unclear; therefore, these foci may correspond to a poorly understood biologic event in tumor evolution.

Influence of gray level and space discretization on brain tumor heterogeneity measures obtained from magnetic resonance images

PURPOSE

Tumor heterogeneity in medical imaging is a current research trend due to its potential relationship with tumor malignancy. The aim of this study is to analyze the effect of dynamic range and matrix size changes on the results of different heterogeneity measures.

MATERIALS AND METHODS

Four patients harboring three glioblastomas and one metastasis were considered. Sixteen textural heterogeneity measures were computed for each patient, with a configuration including co-occurrence matrices (CM) features (local heterogeneity) and run-length matrices (RLM) features (regional heterogeneity). The coefficient of variation measured agreement between the textural measures in two types of experiments: (i) fixing the matrix size and changing the dynamic range and (ii) fixing the dynamic range and changing the matrix size.

RESULTS

None of the measures considered were robust under dynamic range changes. The CM Entropy and the RLM high gray-level run emphasis (HGRE) were the outstanding textural features due to their robustness under matrix size changes. Also, the RLM low gray-level run emphasis (LGRE) provided robust results when the dynamic range considered was sufficiently high (more than 8 levels). All of the remaining textural features were not robust.

CONCLUSION

Tumor texture studies based on images with different characteristics (e.g. multi-center studies) should first fix the dynamic range to be considered. For studies involving images of different resolutions either (i) only robust measures should be used (in our study CM entropy, RLM HGRE and/or RLM LGRE) or (ii) images should be resampled to match those of the lowest resolution before computing the textural features.

Non-invasive quantification of tumour heterogeneity in water diffusivity to differentiate malignant from benign tissues of urinary bladder: a phase I study

OBJECTIVES

To quantify the heterogeneity of the tumour apparent diffusion coefficient (ADC) using voxel-based analysis to differentiate malignancy from benign wall thickening of the urinary bladder.

METHODS

Nineteen patients with histopathological findings of their cystectomy specimen were included. A data set of voxel-based ADC values was acquired for each patient's lesion. Histogram analysis was performed on each data set to calculate uniformity (U) and entropy (E). The k-means clustering of the voxel-wised ADC data set was implemented to measure mean intra-cluster distance (MICD) and largest inter-cluster distance (LICD). Subsequently, U, E, MICD, and LICD for malignant tumours were compared with those for benign lesions using a two-sample t-test.

RESULTS

Eleven patients had pathological confirmation of malignancy and eight with benign wall thickening. Histogram analysis showed that malignant tumours had a significantly higher degree of ADC heterogeneity with lower U (P = 0.016) and higher E (P = 0.005) than benign lesions. In agreement with these findings, k-means clustering of voxel-wise ADC indicated that bladder malignancy presented with significantly higher MICD (P < 0.001) and higher LICD (P = 0.002) than benign wall thickening.

CONCLUSIONS

The quantitative assessment of tumour diffusion heterogeneity using voxel-based ADC analysis has the potential to become a non-invasive tool to distinguish malignant from benign tissues of urinary bladder cancer.

KEY POINTS

• Heterogeneity is an intrinsic characteristic of tumoral tissue. • Non-invasive quantification of tumour heterogeneity can provide adjunctive information to improve cancer diagnosis accuracy. • Histogram analysis and k-means clustering can quantify tumour diffusion heterogeneity. • The quantification helps differentiate malignant from benign urinary bladder tissue.

Tumour heterogeneity in glioblastoma assessed by MRI texture analysis: a potential marker of survival

OBJECTIVE:

The main objective of this retrospective work was the study of three-dimensional (3D) heterogeneity measures of post-contrast pre-operative MR images acquired with T₁ weighted sequences of patients with glioblastoma (GBM) as predictors of clinical outcome.

METHODS:

79 patients from 3 hospitals were included in the study. 16 3D textural heterogeneity measures were computed including run-length matrix (RLM) features (regional heterogeneity) and co-occurrence matrix (CM) features (local heterogeneity). The significance of the results was studied using Kaplan-Meier curves and Cox proportional hazards analysis. Correlation between the variables of the study was assessed using the Spearman's correlation coefficient.

RESULTS:

Kaplan-Meyer survival analysis showed that 4 of the 11 RLM features and 4 of the 5 CM features considered were robust predictors of survival. The median survival differences in the most significant cases were of over 6 months.

CONCLUSION:

Heterogeneity measures computed on the post-contrast pre-operative T₁ weighted MR images of patients with GBM are predictors of survival.

ADVANCES IN KNOWLEDGE:

Texture analysis to assess tumour heterogeneity has been widely studied. However, most works develop a two-dimensional analysis, focusing only on one MRI slice to state tumour heterogeneity. The study of fully 3D heterogeneity textural features as predictors of clinical outcome is more robust and is not dependent on the selected slice of the tumour.

DEMARCATE: Density-based magnetic resonance image clustering for assessing tumor heterogeneity in cancer

Tumor heterogeneity is a crucial area of cancer research wherein inter- and intra-tumor differences are investigated to assess and monitor disease development and progression, especially in cancer. The proliferation of imaging and linked genomic data has enabled us to evaluate tumor heterogeneity on multiple levels. In this work, we examine magnetic resonance imaging (MRI) in patients with brain cancer to assess image-based tumor heterogeneity. Standard approaches to this problem use scalar summary measures (e.g., intensity-based histogram statistics) that do not adequately capture the complete and finer scale information in the voxel-level data. In this paper, we introduce a novel technique, DEMARCATE (DEnsity-based MAgnetic Resonance image Clustering for Assessing Tumor hEterogeneity) to explore the entire tumor heterogeneity density profiles (THDPs) obtained from the full tumor voxel space. THDPs are smoothed representations of the probability density function of the tumor images. We develop tools for analyzing such objects under the Fisher-Rao Riemannian framework that allows us to construct metrics for THDP comparisons across patients, which can be used in conjunction with standard clustering approaches. Our analyses of The Cancer Genome Atlas (TCGA) based Glioblastoma dataset reveal two significant clusters of patients with marked differences in tumor morphology, genomic characteristics and prognostic clinical outcomes. In addition, we see enrichment of image-based clusters with known molecular subtypes of glioblastoma multiforme, which further validates our representation of tumor heterogeneity and subsequent clustering techniques.

Tumor Heterogeneity in Lung Cancer: Assessment with Dynamic Contrast-enhanced MR Imaging

Purpose To evaluate histogram and texture parameters on pretreatment dynamic contrast material-enhanced (DCE) magnetic resonance (MR) images in lung cancer in terms of temporal change, optimal time for analysis, and prognostic potential. Materials and Methods This retrospective study was approved by the institutional review board, and the requirement to obtain informed consent was waived. Thirty-eight patients with pathologically proved lung cancer undergoing standard pretreatment DCE MR imaging were included. A fat-suppressed, T1-weighted, volume-interpolated breath-hold MR sequence was performed every 30 seconds for 300 and 480 seconds after contrast material administration. A region of interest was manually drawn in the largest cross-sectional area of the tumor on DCE MR images to extract semiquantitative perfusion, histogram, and texture parameters. Predictability of 2-year progression-free survival (PFS) was analyzed by using the Kaplan-Meier method and Cox regression analysis. Results MR histogram and texture parameters increased rapidly 30-60 seconds after contrast material administration. Standard deviation and entropy then plateaued, whereas skewness and kurtosis rapidly decreased. Univariate Cox regression analysis revealed that standard deviation and entropy were significant predictors of survival; their statistical significance was preserved from 60 to 300 seconds, with the smallest P values (P ≤ .001) occurring from 120 to 180 seconds. At multivariate Cox regression analysis, entropy was the sole significant predictor of 2-year PFS (hazard ratio at 180 seconds, 10.098 [95% confidence interval: 1.579, 64.577], P = .015; hazard ratio at 120 seconds: 11.202 [95% confidence interval: 1.761, 71.260], P = .010). Conclusion Histogram and texture parameter changes varied after contrast material injection. The 120-180-second window after contrast material injection was optimal for MR imaging-derived texture parameter and entropy at DCE MR imaging. (©) RSNA, 2016 Online supplemental material is available for this article.

Texture analysis on parametric maps derived from dynamic contrast-enhanced magnetic resonance imaging in head and neck cancer

AIM: To investigate the merits of texture analysis on parametric maps derived from pharmacokinetic modeling with dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) as imaging biomarkers for the prediction of treatment response in patients with head and neck squamous cell carcinoma (HNSCC).

METHODS: In this retrospective study, 19 HNSCC patients underwent pre- and intra-treatment DCE-MRI scans at a 1.5T MRI scanner. All patients had chemo-radiation treatment. Pharmacokinetic modeling was performed on the acquired DCE-MRI images, generating maps of volume transfer rate (K^trans) and volume fraction of the extravascular extracellular space (v_e). Image texture analysis was then employed on maps of K^trans and v_e, generating two texture measures: Energy (E) and homogeneity.

RESULTS: No significant changes were found for the mean and standard deviation for K^trans and v_e between pre- and intra-treatment (P > 0.09). Texture analysis revealed that the imaging biomarker E of v_e was significantly higher in intra-treatment scans, relative to pretreatment scans (P < 0.04).

CONCLUSION: Chemo-radiation treatment in HNSCC significantly reduces the heterogeneity of tumors.

Radiation treatment monitoring using multimodal functional imaging: PET/CT ((18)F-Fluoromisonidazole & (18)F-Fluorocholine) and DCE-US

Background

This study aims to assess the effect of radiation treatment on the tumour vasculature and its downstream effects on hypoxia and choline metabolism using a multimodal approach in the murine prostate tumour model CWR22. Functional parameters derived from Positron Emission Tomography (PET)/Computer Tomography (CT) with ¹⁸F-Fluoromisonidazole (¹⁸F-FMISO) and ¹⁸F-Fluorocholine (¹⁸F-FCH) as well as Dynamic Contrast-Enhanced Ultrasound (DCE-US) were employed to determine the relationship between metabolic parameters and microvascular parameters that reflect the tumour microenvironment. Immunohistochemical analysis was employed for validation.

Methods

PET/CT and DCE-US were acquired pre- and post-treatment, at day 0 and day 3, respectively. At day 1, radiation treatment was delivered as a single fraction of 10 Gy. Two experimental groups were tested for treatment response with ¹⁸F-FMISO and ¹⁸F-FCH.

Results

The maximum Standardized Uptake Values (SUVmax) and the mean SUV (SUVmean) for the ¹⁸F-FMISO group were decreased after treatment, and the SUVmean of the tumour-to-muscle ratio was correlated to microvessel density (MVD) at day 3. The kurtosis of the amplitude of the contrast uptake A was significantly decreased for the control tumours in the ¹⁸F-FCH group. Furthermore, the eliminating rate constant of the contrast agent from the plasma k_el derived from DCE-US was negatively correlated to the SUVmean of tumour-to-muscle ratio, necrosis and MVD.

Conclusions

The present study suggests that the multimodal approach using ¹⁸F-FMISO PET/CT and DCE-US seems reliable in the assessment of both microvasculature and necrosis as validated by histology. Thus, it has valuable diagnostic and prognostic potential for early non-invasive evaluation of radiotherapy.