Dynamic contrast-enhanced texture analysis of the liver: initial assessment in colorectal cancer

Diagnostic CT of colorectal cancer with artificial intelligence iterative reconstruction: A clinical evaluation

OBJECTIVES

To investigate the clinical value of a novel deep-learning based CT reconstruction algorithm, artificial intelligence iterative reconstruction (AIIR), in diagnostic imaging of colorectal cancer (CRC).

METHODS

This study retrospectively enrolled 217 patients with pathologically confirmed CRC. CT images were reconstructed with the AIIR algorithm and compared with those originally obtained with hybrid iterative reconstruction (HIR). Objective image quality was evaluated in terms of the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Subjective image quality was graded on the conspicuity of tumor margin and enhancement pattern as well as the certainty in diagnosing organ invasion and regional lymphadenopathy. In patients with surgical pathology (n = 116), the performance of diagnosing visceral peritoneum invasion was characterized using receiver operating characteristic (ROC) analysis. Changes of diagnostic thinking in diagnosing hepatic metastases were assessed through lesion classification confidence.

RESULTS

The SNRs and CNRs on AIIR images were significantly higher than those on HIR images (all p < 0.001). The AIIR was scored higher for all subjective metrics (all p < 0.001) except for the certainty of diagnosing regional lymphadenopathy (p = 0.467). In diagnosing visceral peritoneum invasion, higher area under curve (AUC) of the ROC was found for AIIR than HIR (0.87 vs 0.77, p = 0.001). In assessing hepatic metastases, AIIR was found capable of correcting the misdiagnosis and improving the diagnostic confidence provided by HIR (p = 0.01).

CONCLUSIONS

Compared to HIR, AIIR offers better image quality, improves the diagnostic performance regarding CRC, and thus has the potential for application in routine abdominal CT.

A Review of Radiomics in Predicting Therapeutic Response in Colorectal Liver Metastases: From Traditional to Artificial Intelligence Techniques

An early evaluation of colorectal cancer liver metastasis (CRCLM) is crucial in determining treatment options that ultimately affect patient survival rates and outcomes. Radiomics (quantitative imaging features) have recently gained popularity in diagnostic and therapeutic strategies. Despite this, radiomics faces many challenges and limitations. This study sheds light on these limitations by reviewing the studies that used radiomics to predict therapeutic response in CRCLM. Despite radiomics’ potential to enhance clinical decision-making, it lacks standardization. According to the results of this study, the instability of radiomics quantification is caused by changes in CT scan parameters used to obtain CT scans, lesion segmentation methods used for contouring liver metastases, feature extraction methods, and dataset size used for experimentation and validation. Accordingly, the study recommends combining radiomics with deep learning to improve prediction accuracy.

Radiomics artificial intelligence modelling for prediction of local control for colorectal liver metastases treated with radiotherapy

•

Computed tomography imaging contains quantifiable information to characterize colorectal liver metastases.

•

Shape, texture, and intensity statistical features quantified the computed tomography liver volume.

•

An artificial intelligence model to predict local progression from radiomic features was developed with high accuracy.

•

Maximum dosage and textural coarseness of liver volume were features with highest predictive value.

Background and Purpose

Prognostic assessment of local therapies for colorectal liver metastases (CLM) is essential for guiding management in radiation oncology. Computed tomography (CT) contains liver texture information which may be predictive of metastatic environments. To investigate the feasibility of analyzing CT texture, we sought to build an automated model to predict progression-free survival using CT radiomics and artificial intelligence (AI).

Materials and Methods

Liver CT scans and outcomes for N = 97 CLM patients treated with radiotherapy were retrospectively obtained. A survival model was built by extracting 108 radiomic features from liver and tumor CT volumes for a random survival forest (RSF) to predict local progression. Accuracies were measured by concordance indices (C-index) and integrated Brier scores (IBS) with 4-fold cross-validation. This was repeated with different liver segmentations and radiotherapy clinical variables as inputs to the RSF. Predictive features were identified by perturbation importances.

Results

The AI radiomics model achieved a C-index of 0.68 (CI: 0.62–0.74) and IBS below 0.25 and the most predictive radiomic feature was gray tone difference matrix strength (importance: 1.90 CI: 0.93–2.86) and most predictive treatment feature was maximum dose (importance: 3.83, CI: 1.05–6.62). The clinical data only model achieved a similar C-index of 0.62 (CI: 0.56–0.69), suggesting that predictive signals exist in radiomics and clinical data.

Conclusions

The AI model achieved good prediction accuracy for progression-free survival of CLM, providing support that radiomics or clinical data combined with machine learning may aid prognostic assessment and management.

Performance of quantitative CT texture analysis in differentiation of gastric tumors

PURPOSE

To examine the computed tomography (CT) images of patients with a diagnosis of gastric tumor by texture analysis and to investigate its place in differential diagnosis.

MATERIALS AND METHODS

Contrast enhanced venous phase CT images of 163 patients with pathological diagnosis of gastric adenocarcinoma (n = 125), gastric lymphoma (n = 12) and gastrointestinal stromal tumors (n = 26) were retrospectively analyzed. Pixel size adjustment, gray-level discretization and gray-level normalization procedures were applied as pre-processing steps. Region of interest (ROI) was determined from the axial slice that represented the largest lesion area and a total of 40 texture features were calculated for each patient. Texture features were compared between the tumor subtypes and between adenocarcinoma grades. Statistically significant texture features were combined into a single parameter by logistic regression analysis. The sensitivity and specificity of these features and the combined parameter were measured to differentiate tumor subtypes by receiver-operating characteristic curve (ROC) analysis.

RESULTS

Classifications between adenocarcinoma versus lymphoma, adenocarcinoma vs. gastrointestinal stromal tumor (GIST) and well-differentiated adenocarcinoma versus poorly differentiated adenocarcinoma using texture features yielded successful results with high sensitivity (98, 91, 96%, respectively) and specificity (75, 77, 80%, respectively).

CONCLUSIONS

CT texture analysis is a non-invasive promising method for classifying gastric tumors and predicting gastric adenocarcinoma differentiation.

Tumor heterogeneity evaluated by computed tomography detects muscle-invasive upper tract urothelial carcinoma that is associated with inflammatory tumor microenvironment

To detect muscle-invasive upper tract urothelial carcinoma, we evaluated the internal texture of the tumor using texture analysis of computed tomography images in 86 cases of upper tract urothelial carcinoma. The internal texture of the tumor was evaluated as the value of computed tomography attenuation number of the unenhanced image, and the median, standard deviation, skewness and kurtosis were calculated. Each parameter was compared with clinicopathological factors, and their associations with postoperative prognosis were investigated. Immunohistochemistry was performed to investigate the histological and molecular mechanisms of the inflammatory tumor microenvironment. The histogram of computed tomography attenuation number in non-muscle invasive tumor was single-peaked, whereas muscle invasive tumor showed a multi-peaked shape. In the parameters obtained by texture analysis, standard deviation was significantly associated with pathological stage (p < 0.0001), tumor grade (p = 0.0053), lymphovascular invasion (p = 0.0078) and concomitant carcinoma in situ (p = 0.0177) along with recurrence-free (p = 0.0191) and overall survival (p = 0.0184). The standard deviation value correlated with the amount of stromal components (p < 0.0001) and number of tumor-infiltrating macrophages (p < 0.0001). In addition, higher expression of high mobility group box 1 was found in heterogeneous tumor. Tumor heterogeneity evaluated by texture analysis was associated with muscle-invasive upper tract urothelial carcinoma and represented an inflammatory tumor microenvironment and useful as the clinical assessment to differentiate muscle invasive tumor.

Advanced Imaging Analysis in Prostate MRI: Building a Radiomic Signature to Predict Tumor Aggressiveness

Radiomics allows the extraction quantitative features from imaging, as imaging biomarkers of disease. The objective of this exploratory study is to implement a reproducible radiomic-pipeline for the extraction of a magnetic resonance imaging (MRI) signature for prostate cancer (PCa) aggressiveness. One hundred and two consecutive patients performing preoperative prostate multiparametric magnetic resonance imaging (mpMRI) and radical prostatectomy were enrolled. Multiparametric images, including T2-weighted (T2w), diffusion-weighted and dynamic contrast-enhanced images, were acquired at 1.5 T. Ninety-three imaging features (Ifs) were extracted from segmentation of index lesion. Ifs were ranked based on a stability rank and redundant Ifs were excluded. Using unsupervised hierarchical clustering, patients were grouped on the basis of similar radiomic patterns, whose association with Gleason Grade Group (GGG), extracapsular extension (ECE), and nodal involvement (pN) was tested. Signatures composed by IFs from T2w-images and Apparent Diffusion Coefficient (ADC) maps were tested for the prediction of GGG, ECE, and pN. T2w radiomic pattern was associated with pN, ECE, and GGG (p = 0.027, 0.05, 0.03) and ADC radiomic pattern was associated with GGG (p = 0.004). The best performance was reached by the signature combing IFs from multiparametric images (0.88, 0.89, and 0.84 accuracy for GGG, pN, and ECE). A reliable multiparametric MRI radiomic signature was extracted, potentially able to predict PCa aggressiveness, to be further validated on an independent sample.

Value of contrast-enhanced CT texture analysis in predicting IDH mutation status of intrahepatic cholangiocarcinoma

To explore the value of contrast-enhanced CT texture analysis in predicting isocitrate dehydrogenase (IDH) mutation status of intrahepatic cholangiocarcinomas (ICCs). Institutional review board approved this study. Contrast-enhanced CT images of 138 ICC patients (21 with IDH mutation and 117 without IDH mutation) were retrospectively reviewed. Texture analysis was performed for each lesion and compared between ICCs with and without IDH mutation. All textural features in each phase and combinations of textural features (p < 0.05) by Mann–Whitney U tests were separately used to train multiple support vector machine (SVM) classifiers. The classification generalizability and performance were evaluated using a tenfold cross-validation scheme. Among plain, arterial phase (AP), portal venous phase (VP), equilibrium phase (EP) and Sig classifiers, VP classifier showed the highest accuracy of 0.863 (sensitivity, 0.727; specificity, 0.885), with a mean area under the receiver operating characteristic curve of 0.813 in predicting IDH mutation in validation cohort. Texture features of CT images in portal venous phase could predict IDH mutation status of ICCs with SVM classifier preoperatively.

Quantitative tumor heterogeneity MRI profiling improves machine learning-based prognostication in patients with metastatic colon cancer

OBJECTIVES

Intra-tumor heterogeneity has been previously shown to be an independent predictor of patient survival. The goal of this study is to assess the role of quantitative MRI-based measures of intra-tumor heterogeneity as predictors of survival in patients with metastatic colorectal cancer.

METHODS

In this IRB-approved retrospective study, we identified 55 patients with stage 4 colon cancer with known hepatic metastasis on MRI. Ninety-four metastatic hepatic lesions were identified on post-contrast images and manually volumetrically segmented. A heterogeneity phenotype vector was extracted from each lesion. Univariate regression analysis was used to assess the contribution of 110 extracted features to survival prediction. A random forest-based machine learning technique was applied to the feature vector and to the standard prognostic clinical and pathologic variables. The dataset was divided into a training and test set at a ratio of 4:1. ROC analysis and confusion matrix analysis were used to assess classification performance.

RESULTS

Mean survival time was 39 ± 3.9 months for the study population. A total of 22 texture features were associated with patient survival (p < 0.05). The trained random forest machine learning model that included standard clinical and pathological prognostic variables resulted in an area under the ROC curve of 0.83. A model that adds imaging-based heterogeneity features to the clinical and pathological variables resulted in improved model performance for survival prediction with an AUC of 0.94.

CONCLUSIONS

MRI-based texture features are associated with patient outcomes and improve the performance of standard clinical and pathological variables for predicting patient survival in metastatic colorectal cancer.

KEY POINTS

• MRI-based tumor heterogeneity texture features are associated with patient survival outcomes. • MRI-based tumor texture features complement standard clinical and pathological variables for prognosis prediction in metastatic colorectal cancer. • Agglomerative hierarchical clustering shows that patient survival outcomes are associated with different MRI tumor profiles.

Contrast-Enhanced CT-based Textural Parameters as Potential Prognostic Factors of Survival for Colorectal Cancer Patients Receiving Targeted Therapy

PURPOSE

This study was designed to estimate the clinical significance of the contrast-enhanced computed tomography (CT) textural features for prediction of survival in colorectal cancer (CRC) patients receiving targeted therapy (bevacizumab and cetuximab).

PROCEDURES

The LifeX software was used to extract the textural parameters of the tumor lesions in the contrast-enhanced CT. We used the least absolute shrinkage and selection operator (LASSO) Cox regression and random forest method to screen the non-redundant radiomic features and constructed the CT imaging score. Univariate and multivariate analyses through the Cox proportional hazards model were performed to assess the prognostic clinical factor. Based on the result of multivariate analysis and CT imaging score, combined nomogram model was constructed to predict the overall survival (OS) of patients. Decision curves analysis was employed to evaluate the performance of the combined model and clinical model.

RESULTS

After comparative analysis of the area under curve of the receiver operating characteristic (ROC) curve, we chose the result of random forest model as CT imaging score. Considering the clinical practice and the result of analysis, age, surgery, and lactate dehydrogenase (LDH) level have been introduced into clinical model. Based on the result of analysis and the CT imaging score, we constructed the nomogram combined model. C-index and calibration curve verified the goodness of fit and discrimination of the combined model. Decision curve analysis (DCA) demonstrated that the combined model showed the better net benefit for a 3-year OS than clinical model.

CONCLUSIONS

In conclusion, the study provides preliminary evidences that several radiomic parameters of tumor lesions derived from CT images were prognostic factors and predictive markers for CRC patients who are candidates for targeted therapy (bevacizumab and cetuximab).

MRI texture analysis in acromegaly and its role in predicting response to somatostatin receptor ligands

PURPOSE

Given the paucity of reliable predictors of tumor recurrence, progression, or response to somatostatin receptor ligand (SRL) therapy in acromegaly, we attempted to determine whether preoperative MR image texture was predictive of these clinical outcomes. We also determined whether image texture could differentiate somatotroph adenomas from non-functioning pituitary adenomas (NFPAs).

METHODS

We performed a retrospective study of patients with acromegaly due to a macroadenoma who underwent transsphenoidal surgery at our institution between 2007 and 2015. Clinical data were extracted from electronic medical records. MRI texture analysis was performed on preoperative non-enhanced T1-weighted images using ImageJ (NIH). Logistic and Cox models were used to determine if image texture parameters predicted outcomes.

RESULTS

Eighty-nine patients had texture parameters measured, which were compared to that of NFPAs, while 64 of these patients had follow-up and were included in the remainder of analyses. Minimum pixel intensity, skewness, and kurtosis were significantly different in somatotroph adenomas versus NFPAs (area under the receiver operating characteristic curve, 0.7771, for kurtosis). Furthermore, those with a maximum pixel intensity above the median had an increased odds of IGF-I normalization on SRL therapy (OR 5.96, 95% CI 1.33-26.66), which persisted after adjusting for several potential predictors of response. Image texture did not predict tumor recurrence or progression.

CONCLUSION

Our data suggest that MRI texture analysis can distinguish NFPAs from somatotroph macroadenomas with good diagnostic accuracy and can predict normalization of IGF-I with SRL therapy.

Usefulness of CT texture analysis in differentiating benign and malignant renal tumours

AIM

To elucidate visually imperceptible differences between benign and malignant renal tumours using computed tomography texture analysis (CTTA) using filtration histogram based parameters.

MATERIALS AND METHODS

A retrospective study was performed by texture analysis of pretreatment contrast-enhanced CT examinations in 354 histopathologically confirmed renal cell carcinomas (RCCs) and 147 benign renal tumours. A region-of-interest was drawn encompassing the largest cross-section of the tumour on venous phase axial CT. CTTA features of entropy, kurtosis, mean positive pixel density, and skewness at different spatial filters were calculated and compared in an attempt to differentiate benign lesions from malignancy.

RESULTS

Entropy with fine spatial filter was significantly higher in RCC than benign renal tumours (p=0.022). Entropy with fine and medium filters was higher in RCC than lipid-poor angiomyolipoma (p=0.050 and 0.052, respectively). Entropy >5.62 had high specificity of 85.7%, but low sensitivity of 31.3%, respectively, for predicting RCC.

CONCLUSIONS

Differences in entropy were helpful in differentiating RCC from lipid-poor angiomyolipoma, and chromophobe RCC from oncocytoma. This technique may be useful to differentiate lesions that appear equivocal on visual assessment or alter management in poor surgical candidates.

The 2017 Assisi Think Tank Meeting on rectal cancer: A positioning paper

BACKGROUND AND PURPOSES

To describe current practice in the management of rectal cancer, to identify uncertainties that usually arise in the multidisciplinary team (MDT)'s discussions ('grey zones') and propose next generation studies which may provide answers to them.

MATERIALS AND METHODS

A questionnaire on the areas of controversy in managing T2, T3 and T4 rectal cancer was drawn up and distributed to the Rectal-Assisi Think Tank Meeting (ATTM) Expert European Board. Less than 70% agreement on a treatment option was indicated as uncertainty and selected as a 'grey zone'. Topics with large disagreement were selected by the task force group for discussion at the Rectal-ATTM.

RESULTS

The controversial clinical issues that had been identified within cT2-cT3-cT4 needed further investigation. The discussions focused on the role of (1) neoadjuvant therapy and organ preservation on cT2-3a low-middle rectal cancer; (2) neoadjuvant therapy in cT3 low rectal cancer without high risk features; (3) total neoadjuvant therapy, radiotherapy boost and the best chemo-radiotherapy schedule in T4 tumors. A description of each area of investigation and trial proposals are reported.

CONCLUSION

The meeting successfully identified 'grey zones' and, in the light of new evidence, proposed clinical trials for treatment of early, intermediate and advanced stage rectal cancer.

Texture analysis versus conventional MRI prognostic factors in predicting tumor response to neoadjuvant chemotherapy in patients with locally advanced cancer of the uterine cervix

INTRODUCTION

To determine the performance of texture analysis and conventional MRI parameters in predicting tumoral response to neoadjuvant chemotherapy and to assess whether a relationship exists between texture tissue heterogeneity and histological type of uterine cervix cancer.

METHOD AND MATERIALS

Twenty-eight patients with local advanced cervical cancer (FIGO IB2-IIIB), underwent MRI before chemotherapy. Texture analysis parameters were quantified on T2-weighted sequences, as well as the maximum diameter expressed in mm. ADC values were obtained on the ADC map. Statistical analysis included unpaired t test and ROC curve.

RESULTS

No statistical correlation was found between conventional parameters and response to NACT. Mean and skewness showed a strong correlation with the histological type: Adenocarcinomas presented higher mean and skewness values (69.8 ± 10.5 and 0.55 ± 0.19) in comparison with squamous cell carcinomas. Using a cutoff value ≥ 29 for mean it was possible to differentiate the two histological types with a sensitivity of 100% and a specificity of 81%. Kurtosis showed a positive correlation with tumor response to NACT resulting higher in responders (v.m. 5.7 ± 1.1) in comparison with non-responders (2.3 ± 0.5). The optimal Kurtosis cutoff value for the identification of non-responders tumors was ≤ 3.7 with a sensitivity of 92% and a specificity of 75%.

CONCLUSION

Texture analysis applied to T2-weighted images of uterine cervical cancer exceeded the role of conventional prognostic factors in predicting tumoral response; moreover, they showed a potential role to differentiate histological tumor types.

Unresectable pancreatic ductal adenocarcinoma: Role of CT quantitative imaging biomarkers for predicting outcomes of patients treated with chemotherapy

OBJECTIVES

The primary aim of this study was to determine if computed tomographic (CT) texture analysis measurements of the tumor are independently associated with progression-free survival (PFS) and overall survival (OS) in patients with unresectable pancreatic ductal adenocarcinoma (PDAC), including both unresectable locally advanced and metastatic PDAC, who were treated with chemotherapy.

METHODS

After an institutional review board waiver was obtained, contrast material-enhanced CT studies in 41 patients with unresectable PDAC who underwent contrast-enhanced CT before chemotherapy between 2014 and 2017 were analyzed in terms of tumor texture, with quantification of mean gray-level intensity (Mean), entropy, mean of positive pixels (MPP), kurtosis, standard deviation (SD), and skewness for fine to coarse textures (spatial scaling factor (SSF) 0-6, respectively). The association between pretreatment and posttreatment texture parameters, as well as Δ value (difference between posttreatment and pretreatment texture parameters), and survival time was assessed by using Cox proportional hazards models and Kaplan-Meier analysis.

RESULTS

Findings from the multivariate Cox model indicated that tumor size, tumor SD (HR, 0.942; 95% CI: 0.898, 0.988) and skewness (HR, 0.407; 95% CI: 0.172, 0.962) measurements with SSF = 3, and tumor SD (HR, 0.958; 95% CI: 0.92, 0.997) measurements with SSF = 4 were significantly and independently associated with PFS, while tumor size and tumor SD (HR, 0.928; 95% CI: 0.882, 0.976) measurements with SSF = 3 were significantly and independently associated with OS. None of the post-therapy texture parameters or Δ value had a significant association with OS or PFS in multivariate Cox regression models. Medium SD (SSF = 3) of more than 38.38 and coarse SD (SSF = 4) of more than 40.67 were associated with longer PFS after chemotherapy (for SSF = 3, median PFS was 10.0 vs 6.0 months [P = 0.024], and for SSF = 4, median PFS was 12.0 vs 6.0 months [P = 0.003]). SD of 38.38 or greater (SSF = 3) as a dichotomized variable was a significant positive prognostic factor for OS (median OS, 20.0 vs 9.0 months [P = 0.04]). Survival models that included a combination of pretreatment SD (SSF = 3) with tumor size, had the potential to perform better than SD alone, while having no statistical significance in this study (area under the ROC curve, 0.756 vs 0.715 [P = 0.066]).

CONCLUSIONS

Pretreatment CT quantitative imaging biomarkers from texture analysis are associated with PFS and OS in patients with unresectable PDAC who were treated with chemotherapy, and the combination of pretreatment texture parameters and tumor size have the potential to perform better in survival models than imaging biomarker alone.

Radiomics robustness assessment and classification evaluation: A two-stage method demonstrated on multivendor FFDM

PURPOSE

Radiomic texture analysis is typically performed on images acquired under specific, homogeneous imaging conditions. These controlled conditions may not be representative of the range of imaging conditions implemented clinically. We aim to develop a two-stage method of radiomic texture analysis that incorporates the reproducibility of individual texture features across imaging conditions to guide the development of texture signatures which are robust across mammography unit vendors.

METHODS

Full-field digital mammograms were retrospectively collected for women who underwent screening mammography on both a Hologic Lorad Selenia and GE Senographe 2000D system. Radiomic features were calculated on manually placed regions of interest in each image. In stage one (robustness assessment), we identified a set of nonredundant features that were reproducible across the two different vendors. This was achieved through hierarchical clustering and application of robustness metrics. In stage two (classification evaluation), we performed stepwise feature selection and leave-one-out quadratic discriminant analysis (QDA) to construct radiomic signatures. We refer to this two-state method as robustness assessment, classification evaluation (RACE). These radiomic signatures were used to classify the risk of breast cancer through receiver operator characteristic (ROC) analysis, using the area under the ROC curve as a figure of merit in the task of distinguishing between women with and without high-risk factors present. Generalizability was investigated by comparing the classification performance of a feature set on the images from which they were selected (intravendor) to the classification performance on images from the vendor on which it was not selected (intervendor). Intervendor and intravendor performances were also compared to the performance obtained by implementing ComBat, a feature-level harmonization method and to the performance by implementing ComBat followed by RACE.

RESULTS

Generalizability, defined as the difference between intervendor and intravendor classification performance, was shown to monotonically decrease as the number of clusters used in stage one increased (Mann-Kendall P < 0.001). Intravendor performance was not shown to be statistically different from ComBat harmonization while intervendor performance was significantly higher than ComBat. No significant difference was observed between either of the single methods and the use of ComBat followed by RACE.

CONCLUSIONS

A two-stage method for robust radiomic signature construction is proposed and demonstrated in the task of breast cancer risk assessment. The proposed method was used to assess generalizability of radiomic texture signatures at varying levels of feature robustness criteria. The results suggest that generalizability of feature sets monotonically decreases as reproducibility of features decreases. This trend suggests that considerations of feature robustness in feature selection methodology could improve classifier generalizability in multifarious full-field digital mammography datasets collected on various vendor units. Additionally, harmonization methods such as ComBat may hold utility in classification schemes and should continue to be investigated.

Predicting locally advanced rectal cancer response to neoadjuvant therapy with 18F-FDG PET and MRI radiomics features

PURPOSE

Pathological complete response (pCR) following neoadjuvant chemoradiotherapy or radiotherapy in locally advanced rectal cancer (LARC) is reached in approximately 15-30% of cases, therefore it would be useful to assess if pretreatment of ¹⁸F-FDG PET/CT and/or MRI texture features can reliably predict response to neoadjuvant therapy in LARC.

METHODS

Fifty-two patients were dichotomized as responder (pR+) or non-responder (pR-) according to their pathological tumor regression grade (TRG) as follows: 22 as pR+ (nine with TRG = 1, 13 with TRG = 2) and 30 as pR- (16 with TRG = 3, 13 with TRG = 4 and 1 with TRG = 5). First-order parameters and 21 second-order texture parameters derived from the Gray-Level Co-Occurrence matrix were extracted from semi-automatically segmented tumors on T2w MRI, ADC maps, and PET/CT acquisitions. The role of each texture feature in predicting pR+ was assessed with monoparametric and multiparametric models.

RESULTS

In the mono-parametric approach, PET homogeneity reached the maximum AUC (0.77; sensitivity = 72.7% and specificity = 76.7%), while PET glycolytic volume and ADC dissimilarity reached the highest sensitivity (both 90.9%). In the multiparametric analysis, a logistic regression model containing six second-order texture features (five from PET and one from T2w MRI) yields the highest predictivity in distinguish between pR+ and pR- patients (AUC = 0.86; sensitivity = 86%, and specificity = 83% at the Youden index).

CONCLUSIONS

If preliminary results of this study are confirmed, pretreatment PET and MRI could be useful to personalize patient treatment, e.g., avoiding toxicity of neoadjuvant therapy in patients predicted pR-.

Variation in algorithm implementation across radiomics software

Given the increased need for consistent quantitative image analysis, variations in radiomics feature calculations due to differences in radiomics software were investigated. Two in-house radiomics packages and two freely available radiomics packages, MaZda and IBEX, were utilized. Forty 256 × 256 - pixel regions of interest (ROIs) from 40 digital mammograms were studied along with 39 manually delineated ROIs from the head and neck (HN) computed tomography (CT) scans of 39 patients. Each package was used to calculate first-order histogram and second-order gray-level co-occurrence matrix (GLCM) features. Friedman tests determined differences in feature values across packages, whereas intraclass-correlation coefficients (ICC) quantified agreement. All first-order features computed from both mammography and HN cases (except skewness in mammography) showed significant differences across all packages due to systematic biases introduced by each package; however, based on ICC values, all but one first-order feature calculated on mammography ROIs and all but two first-order features calculated on HN CT ROIs showed excellent agreement, indicating the observed differences were small relative to the feature values but the bias was systematic. All second-order features computed from the two databases both differed significantly and showed poor agreement among packages, due largely to discrepancies in package-specific default GLCM parameters. Additional differences in radiomics features were traced to variations in image preprocessing, algorithm implementation, and naming conventions. Large variations in features among software packages indicate that increased efforts to standardize radiomics processes must be conducted.

MRI texture analysis as a predictor of tumor recurrence or progression in patients with clinically non-functioning pituitary adenomas

BACKGROUND

There are limited predictors of prognosis in patients with clinically non-functioning pituitary adenomas (NFPAs). We hypothesized that MRI texture analysis may predict tumor recurrence or progression in patients with NFPAs undergoing transsphenoidal pituitary surgery (TSS).

OBJECTIVE

To characterize texture parameters on preoperative MRI examinations in patients with NFPAs in relation to prognosis.

METHODS

Retrospective study of patients with NFPAs who underwent TSS at our institution between 2009 and 2010. Clinical, radiological and histopathological data were extracted from electronic medical records. MRI texture analysis was performed on coronal T1-weighted non-enhanced MR images using ImageJ (NIH). MRI texture parameters were used to predict tumor recurrence or progression. Both logistic regression and Cox proportional hazard analyses were conducted to adjust for potential confounders.

RESULTS

Data on 78 patients were analyzed. On both crude and multivariable-adjusted analyses, mean, median, mode, minimum and maximum pixel intensity were associated with the risk of pituitary tumor recurrence or progression after TSS. Patients whose tumor mean pixel intensity was above the median for the population had a hazard ratio of 0.44 (95% CI: 0.21-0.94, P = 0.034) for recurrence or progression in comparison with tumors below the median.

CONCLUSIONS

Our data suggest that MRI texture analysis can predict the risk of tumor recurrence or progression in patients with NFPAs.

CT texture analysis of pancreatic cancer

OBJECTIVES

We investigated the value of CT texture analysis (CTTA) in predicting prognosis of unresectable pancreatic cancer.

METHODS

Sixty patients with unresectable pancreatic cancers at presentation were enrolled for post-processing with CTTA using commercially available software (TexRAD Ltd, Cambridge, UK). The largest cross-section of the tumour on axial CT was chosen to draw a region-of-interest. CTTA parameters (mean value of positive pixels (MPP), kurtosis, entropy, skewness), arterial and venous invasion, metastatic disease and tumour size were correlated with overall and progression-free survivals.

RESULTS

The median overall and progression-free survivals of cohort were 13.3 and 7.8 months, respectively. On multivariate Cox proportional hazard regression analysis, presence of metastatic disease at presentation had the highest association with overall survival (p = 0.003-0.05) and progression-free survival (p < 0.001 to p = 0.004). MPP at medium spatial filter was significantly associated with poor overall survival (p = 0.04). On Kaplan-Meier survival analysis of CTTA parameters at medium spatial filter, MPP of more than 31.625 and kurtosis of more than 0.565 had significantly worse overall survival (p = 0.036 and 0.028, respectively).

CONCLUSIONS

CTTA features were significantly associated with overall survival in pancreas cancer, particularly in patients with non-metastatic, locally advanced disease.

KEY POINTS

• CT texture analysis is easy to perform on contrast-enhanced CT. • CT texture analysis can determine prognosis in patients with unresectable pancreas cancer. • The best predictors of poor prognosis were high kurtosis and MPP.

Could texture features from preoperative CT image be used for predicting occult peritoneal carcinomatosis in patients with advanced gastric cancer?

Purpose

To retrospectively investigate whether texture features obtained from preoperative CT images of advanced gastric cancer (AGC) patients could be used for the prediction of occult peritoneal carcinomatosis (PC) detected during operation.

Materials and methods

51 AGC patients with occult PC detected during operation from January 2009 to December 2012 were included as occult PC group. For the control group, other 51 AGC patients without evidence of distant metastasis including PC, and whose clinical T and N stage could be matched to those of the patients of the occult PC group, were selected from the period of January 2011 to July 2012. Each group was divided into test (n = 41) and validation cohort (n = 10). Demographic and clinical data of these patients were acquired from the hospital database. Texture features including average, standard deviation, kurtosis, skewness, entropy, correlation, and contrast were obtained from manually drawn region of interest (ROI) over the omentum on the axial CT image showing the omentum at its largest cross sectional area. After using Fisher's exact and Wilcoxon signed-rank test for comparison of the clinical and texture features between the two groups of the test cohort, conditional logistic regression analysis was performed to determine significant independent predictor for occult PC. Using the optimal cut-off value from receiver operating characteristic (ROC) analysis for the significant variables, diagnostic sensitivity and specificity were determined in the test cohort. The cut-off value of the significant variables obtained from the test cohort was then applied to the validation cohort. Bonferroni correction was used to adjust P value for multiple comparisons.

Results

Between the two groups, there was no significant difference in the clinical features. Regarding the texture features, the occult PC group showed significantly higher average, entropy, standard deviation, and significantly lower correlation (P value < 0.004 for all). Conditional logistic regression analysis demonstrated that entropy was significant independent predictor for occult PC. When the cut-off value of entropy (> 7.141) was applied to the validation cohort, sensitivity and specificity for the prediction of occult PC were 80% and 90%, respectively.

Conclusion

For AGC patients whose PC cannot be detected with routine imaging such as CT, texture analysis may be a useful adjunct for the prediction of occult PC.

Influence of CT acquisition and reconstruction parameters on radiomic feature reproducibility

High-dimensional imaging features extracted from diagnostic imaging, called radiomics, are increasingly reported for diagnosis, prognosis, and response to therapy. Establishing the sensitivity of radiomic features to variation in scan protocols is necessary because acquisition and reconstruction parameters can vary widely across and within institutions. Our objective was to assess the reproducibility of radiomic features derived from computed tomography (CT) images by varying tube current (mA), noise index, and reconstruction [adaptive statistical iterative reconstruction (ASiR)], parameters increasingly varied by institutions seeking to reduce radiation dose in their patients. We extracted radiomic features from CT images of a uniform water phantom, anthropomorphic phantom, and a human scan. Scans were acquired from the phantoms with six tube currents (50, 100, 200, 300, 400, and 500 mA) and five noise index levels (12, 14, 16, 18, and 20), respectively. Scans of the phantoms and patient were reconstructed from 0% ASiR (i.e., filtered back projection) to 100% ASiR in increments of 10%. Two hundred and forty-eight well-known radiomic features were extracted from all scans. The concordance correlation coefficient was used to assess agreement of features. Our analysis suggests that image acquisition parameters (tube current, noise index) as well as the reconstruction technique strongly influence radiomic feature reproducibility and demonstrate a subset of reproducible features potentially usable in clinical practice.

CT Texture Analysis: Definitions, Applications, Biologic Correlates, and Challenges

This review discusses potential oncologic and nononcologic applications of CT texture analysis ( CTTA CT texture analysis ), an emerging area of "radiomics" that extracts, analyzes, and interprets quantitative imaging features. CTTA CT texture analysis allows objective assessment of lesion and organ heterogeneity beyond what is possible with subjective visual interpretation and may reflect information about the tissue microenvironment. CTTA CT texture analysis has shown promise in lesion characterization, such as differentiating benign from malignant or more biologically aggressive lesions. Pretreatment CT texture features are associated with histopathologic correlates such as tumor grade, tumor cellular processes such as hypoxia or angiogenesis, and genetic features such as KRAS or epidermal growth factor receptor (EGFR) mutation status. In addition, and likely as a result, these CT texture features have been linked to prognosis and clinical outcomes in some tumor types. CTTA CT texture analysis has also been used to assess response to therapy, with decreases in tumor heterogeneity generally associated with pathologic response and improved outcomes. A variety of nononcologic applications of CTTA CT texture analysis are emerging, particularly quantifying fibrosis in the liver and lung. Although CTTA CT texture analysis seems to be a promising imaging biomarker, there is marked variability in methods, parameters reported, and strength of associations with biologic correlates. Before CTTA CT texture analysis can be considered for widespread clinical implementation, standardization of tumor segmentation and measurement techniques, image filtration and postprocessing techniques, and methods for mathematically handling multiple tumors and time points is needed, in addition to identification of key texture parameters among hundreds of potential candidates, continued investigation and external validation of histopathologic correlates, and structured reporting of findings. ^©RSNA, 2017.

Computed tomography texture analysis to facilitate therapeutic decision making in hepatocellular carcinoma

This study explored the potential of computed tomography (CT) textural feature analysis for the stratification of single large hepatocellular carcinomas (HCCs) > 5 cm, and the subsequent determination of patient suitability for liver resection (LR) or transcatheter arterial chemoembolization (TACE). Wavelet decomposition was performed on portal-phase CT images with three bandwidth responses (filter 0, 1.0, and 1.5). Nine textural features of each filter were extracted from regions of interest. Wavelet-2-H (filter 1.0) in LR and wavelet-2-V (filter 0 and 1.0) in TACE were related to survival. Subsequently, LR and TACE patients were divided based on the wavelet-2-H and wavelet-2-V median at filter 1.0 into two subgroups (+ or −). LR+ patients showed the best survival, followed by LR-, TACE+, and TACE-. We estimated that LR+ patients treated using TACE would exhibit a survival similar to TACE- patients and worse than TACE+ patients, with a severe compromise in overall survival. LR was recommended for TACE- patients, whereas TACE was preferred for LR- and TACE+ patients. Independent of tumor size, CT textural features showed positive and negative correlations with survival after LR and TACE, respectively. Although further validation is needed, texture analysis demonstrated the feasibility of using HCC patient stratification for determining the suitability of LR vs. TACE.

Prognostic value of histogram analysis in advanced non-small cell lung cancer: a radiomic study

Introduction

Quantitative assessment of heterogeneity by histogram analysis (HA) of tumor images can potentially provide a non-invasive prognostic biomarker. We assessed the prognostic value of HA and evaluated a correlation with molecular signature.

Results

CT scans performed between July 2009 and January 2015 from 692 patients were reviewed. HA was performed on scans from 313 patients in the training dataset and 108 in the validation dataset. Median follow-up were 33.7 months [range: 1.7 - 65.5] and 29 months [range: 1.1 - 35.6] with a median overall survival (OS) of 11.7 months [95%CI: 10.7 - 13.1] and 9.5 months [95%CI: 7.9 - 12.7] respectively. Primary mass entropy in coarse texture with spatial filter 3.3 was prognostic for OS in a multivariate Cox analysis (HR: 1.3 [95%CI: 1.1 - 1.5], p=0.001). Results were not reproduced in our validation set and no correlation with molecular signature was identified.

Materials and Methods

HA using filtration-histogram method was applied to the region of interest on the primary tumor in enhanced-CT acquired as diagnostic/staging routine, from a cohort of patients with advanced non-small cell lung cancer (NSCLC) treated with platinum-based chemotherapy. The resultants parameters were prospectively applied to a validation dataset. CT scans, clinical and molecular data were retrospectively collected. Cox proportional hazard models were used for survival analysis and Wilcoxon test for correlations.

Conclusion

Primary mass entropy was significantly associated with survival in the training set but was not validated in the validation cohort, raising doubt over the reliability of published data from small cohorts.

Evaluation of Texture Analysis Parameter for Response Prediction in Patients with Hepatocellular Carcinoma Undergoing Drug-eluting Bead Transarterial Chemoembolization (DEB-TACE) Using Biphasic Contrast-enhanced CT Image Data: Correlation with Liver Perfusion CT

RATIONALE AND OBJECTIVES

This study aimed to evaluate the potential role of computed tomography texture analysis (CTTA) of arterial and portal-venous enhancement phase image data for prediction and accurate assessment of response of hepatocellular carcinoma undergoing drug-eluting bead transarterial chemoembolization (TACE) by comparison to liver perfusion CT (PCT).

MATERIALS AND METHODS

Twenty-eight patients (27 male; mean age 67.2 ± 10.4) with 56 hepatocellular carcinoma-typical liver lesions were included. Arterial and portal-venous phase CT data obtained before and after TACE with a mean time of 39.93 ± 62.21 days between examinations were analyzed. TACE was performed within 48 hours after first contrast-enhanced CT. CTTA software was a prototype. CTTA analysis was performed blinded (for results) by two observers separately. Combined results of modified Response Evaluation Criteria In Solid Tumors (mRECIST) and PCT of the liver were used as the standard of reference. Time to progression was additionally assessed for all patients. CTTA parameters included heterogeneity, intensity, average, deviation, skewness, and entropy of co-occurrence. Each parameter was compared to those of PCT (blood flow [BF], blood volume, arterial liver perfusion [ALP], portal-venous perfusion, and hepatic perfusion index) measured before and after TACE.

RESULTS

mRECIST + PCT yielded 28.6% complete response (CR), 42.8% partial response, and 28.6% stable disease. Significant correlations were registered in the arterial phase in CR between changes in mean heterogeneity and BF (P = .004, r = -0.815), blood volume (P = .002, r = -0.851), and ALP (P = .002, r = -0.851), respectively. In the partial response group, changes in mean heterogeneity correlated with changes in ALP (P = .003) and to a lesser degree with hepatic perfusion index (P = .027) in the arterial phase. In the stable disease group, BF correlated with entropy of nonuniformity (P = .010). In the portal-venous phase, no statistically significant correlations were registered in all groups. Receiver operating characteristic analysis of CTTA parameters yielded predictive cutoff values for CR in the arterial contrast-enhanced CT phase for uniformity of skewness (sensitivity: 90.0%; specificity: 45.8%), and in the portal-venous phase for uniformity of heterogeneity (sensitivity: 92.3%; specificity: 81.8%).

CONCLUSIONS

Significant correlations exist between CTTA parameters and those derived from PCT both in the pre- and the post-TACE settings, and some of them have predictive value for TACE midterm outcome.

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.

Harmonizing the pixel size in retrospective computed tomography radiomics studies

Consistent pixel sizes are of fundamental importance for assessing texture features that relate intensity and spatial information in radiomics studies. To correct for the effects of variable pixel sizes, we combined image resampling with Butterworth filtering in the frequency domain and tested the correction on computed tomography (CT) scans of lung cancer patients reconstructed 5 times with pixel sizes varying from 0.59 to 0.98 mm. One hundred fifty radiomics features were calculated for each preprocessing and field-of-view combination. Intra-patient agreement and inter-patient agreement were compared using the overall concordance correlation coefficient (OCCC). To further evaluate the corrections, hierarchical clustering was used to identify patient scans before and after correction. To assess the general applicability of the corrections, they were applied to 17 CT scans of a radiomics phantom. The reduction in the inter-scanner variability relative to non–small cell lung cancer patient scans was quantified. The variation in pixel sizes caused the intra-patient variability to be large (OCCC <95%) relative to the inter-patient variability in 79% of the features. However, with the resampling and filtering corrections, the intra-patient variability was relatively large in only 10% of the features. With the filtering correction, 8 of 8 patients were correctly clustered, in contrast to only 2 of 8 without the correction. In the phantom study, resampling and filtering the images of a rubber particle cartridge substantially reduced variability in 61% of the radiomics features and substantially increased variability in only 6% of the features. Surprisingly, resampling without filtering tended to increase the variability. In conclusion, applying a correction based on resampling and Butterworth low-pass filtering in the frequency domain effectively reduced variability in CT radiomics features caused by variations in pixel size. This correction may also reduce the variability introduced by other CT scan acquisition parameters.

A peek into the future of radiology using big data applications

Big data is extremely large amount of data which is available in the radiology department. Big data is identified by four Vs – Volume, Velocity, Variety, and Veracity. By applying different algorithmic tools and converting raw data to transformed data in such large datasets, there is a possibility of understanding and using radiology data for gaining new knowledge and insights. Big data analytics consists of 6Cs – Connection, Cloud, Cyber, Content, Community, and Customization. The global technological prowess and per-capita capacity to save digital information has roughly doubled every 40 months since the 1980's. By using big data, the planning and implementation of radiological procedures in radiology departments can be given a great boost. Potential applications of big data in the future are scheduling of scans, creating patient-specific personalized scanning protocols, radiologist decision support, emergency reporting, virtual quality assurance for the radiologist, etc. Targeted use of big data applications can be done for images by supporting the analytic process. Screening software tools designed on big data can be used to highlight a region of interest, such as subtle changes in parenchymal density, solitary pulmonary nodule, or focal hepatic lesions, by plotting its multidimensional anatomy. Following this, we can run more complex applications such as three-dimensional multi planar reconstructions (MPR), volumetric rendering (VR), and curved planar reconstruction, which consume higher system resources on targeted data subsets rather than querying the complete cross-sectional imaging dataset. This pre-emptive selection of dataset can substantially reduce the system requirements such as system memory, server load and provide prompt results. However, a word of caution, “big data should not become “dump data” due to inadequate and poor analysis and non-structured improperly stored data. In the near future, big data can ring in the era of personalized and individualized healthcare.

Whole liver CT texture analysis to predict the development of colorectal liver metastases-A multicentre study

OBJECTIVES

CT texture analysis has shown promise to differentiate colorectal cancer patients with/without hepatic metastases.

AIM

To investigate whether whole-liver CT texture analysis can also predict the development of colorectal liver metastases.

MATERIAL AND METHODS

Retrospective multicentre study (n=165). Three subgroups were assessed: patients [A] without metastases (n=57), [B] with synchronous metastases (n=54) and [C] who developed metastases within ≤24 months (n=54). Whole-liver texture analysis was performed on primary staging CT. Mean grey-level intensity, entropy and uniformity were derived with different filters (σ0.5-2.5). Univariable logistic regression (group A vs. B) identified potentially predictive parameters, which were tested in multivariable analyses to predict development of metastases (group A vs. C), including subgroup analyses for early (≤6 months), intermediate (7-12 months) and late (13-24 months) metastases.

RESULTS

Univariable analysis identified uniformity (σ0.5), sex, tumour site, nodal stage and carcinoembryonic antigen as potential predictors. Uniformity remained a significant predictor in multivariable analysis to predict early metastases (OR 0.56). None of the parameters could predict intermediate/late metastases.

CONCLUSIONS

Whole-liver CT-texture analysis has potential to predict patients at risk of developing early liver metastases ≤6 months, but is not robust enough to identify patients at risk of developing metastases at later stage.

Quantitative CT texture analysis for evaluating histologic grade of urothelial carcinoma

PURPOSE

To investigate the feasibility of using CT texture analysis (CTTA) to differentiate between low- versus high-grade urothelial carcinoma.

METHODS

A total of 105 patients with high-grade urothelial carcinoma (HGUC, n = 106) and low-grade urothelial carcinoma (LGUC, n = 18) were included in this retrospective study. Both unenhanced and enhanced CT images representing the largest cross-sectional area of the tumor were chosen for CTTA performed using TexRAD software. Comparison of texture parameters, mean gray-level intensity (Mean), standard deviation, entropy, mean of positive pixels (MPP), skewness, and kurtosis were made for the objective. Receiver operating characteristic (ROC) analysis was performed and the area under the ROC curve was calculated for texture parameters that were significantly different (P < 0.05) for the purpose. Sensitivity (Se), specificity (Sp), positive predictive value, negative predictive value, and accuracy were calculated using the cut-off value of texture parameter with the highest AUC.

RESULTS

Compared to HGUC, LGUC had significantly lower Mean (P = 0.001), Entropy (P = 0.002), and MPP (P < 0.001) on unenhanced and enhanced images and lower SD (P = 0.048) on enhanced images. There was no significant difference in skewness or kurtosis at any texture scale on unenhanced and enhanced images. A MPP <24.13 at fine texture scale on unenhanced images identified LGUC from HGUC with the highest AUC of 0.779 ± 0.065 (Se = 72.2%, Sp = 84.9%, PPV = 44.8%, NPV = 94.7%, and accuracy = 83.1%).

CONCLUSIONS

CTTA proved to be a feasible tool for differentiating LGUC from HGUC. MPP quantified from fine texture scale on unenhanced images was the optimal diagnostic parameter for estimating histologic grade of urothelial carcinoma.

Texture Analysis of Abnormal Cell Images for Predicting the Continuum of Colorectal Cancer

Abnormal cell (ABC) is a markedly heterogeneous tissue area and can be categorized into three main types: benign hyperplasia (BH), carcinoma (Ca), and intraepithelial neoplasia (IN) or precursor cancerous lesion. In this study, the goal is to determine and characterize the continuum of colorectal cancer by using a 3D-texture approach. ABC was segmented in preprocessing step using an active contour segmentation technique. Cell types were analyzed based on textural features extracted from the gray level cooccurrence matrices (GLCMs). Significant texture features were selected using an analysis of variance (ANOVA) of ABC with a p value cutoff of p < 0.01. Features selected were reduced with a principal component analysis (PCA), which accounted for 97% of the cumulative variance from significant features. The simulation results identified 158 significant features based on ANOVA from a total of 624 texture features extracted from GLCMs. Performance metrics of ABC discrimination based on significant texture features showed 92.59% classification accuracy, 100% sensitivity, and 94.44% specificity. These findings suggest that texture features extracted from GLCMs are sensitive enough to discriminate between the ABC types and offer the opportunity to predict cell characteristics of colorectal cancer.

Quantitative Computed Tomography Features for Predicting Tumor Recurrence in Patients with Surgically Resected Adenocarcinoma of the Lung

Purpose

The purpose of this study was to determine if preoperative quantitative computed tomography (CT) features including texture and histogram analysis measurements are associated with tumor recurrence in patients with surgically resected adenocarcinoma of the lung.

Methods

The study included 194 patients with surgically resected lung adenocarcinoma who underwent preoperative CT between January 2013 and December 2013. Quantitative CT feature analysis of the lung adenocarcinomas were performed using in-house software based on plug-in package for ImageJ. Ten quantitative features demonstrating the tumor size, attenuation, shape and texture were extracted. The CT parameters obtained from 1-mm and 5-mm data were compared using intraclass correlation coefficients. Univariate and multivariable logistic regression methods were used to investigate the association between tumor recurrence and preoperative CT findings.

Results

The 1-mm and 5-mm data were highly correlated in terms of diameter, perimeter, area, mean attenuation and entropy. Circularity and aspect ratio were moderately correlated. However, skewness and kurtosis were poorly correlated. Multivariable logistic regression analysis revealed that area (odds ratio [OR], 1.002 for each 1-mm² increase; P = 0.003) and mean attenuation (OR, 1.005 for each 1.0-Hounsfield unit increase; P = 0.022) were independently associated with recurrence. The receiver operating curves using these two independent predictive factors showed high diagnostic performance in predicting recurrence (C-index = 0.81, respectively).

Conclusion

Tumor area and mean attenuation are independently associated with recurrence in patients with surgically resected adenocarcinoma of the lung.

T2-weighted MRI-derived textural features reflect prostate cancer aggressiveness: preliminary results

PURPOSE

To evaluate the diagnostic relevance of T2-weighted (T2W) MRI-derived textural features relative to quantitative physiological parameters derived from diffusion-weighted (DW) and dynamic contrast-enhanced (DCE) MRI in Gleason score (GS) 3+4 and 4+3 prostate cancers.

MATERIALS AND METHODS

3T multiparametric-MRI was performed on 23 prostate cancer patients prior to prostatectomy. Textural features [angular second moment (ASM), contrast, correlation, entropy], apparent diffusion coefficient (ADC), and DCE pharmacokinetic parameters (K^trans and V_e) were calculated from index tumours delineated on the T2W, DW, and DCE images, respectively. The association between the textural features and prostatectomy GS and the MRI-derived parameters, and the utility of the parameters in differentiating between GS 3+4 and 4+3 prostate cancers were assessed statistically.

RESULTS

ASM and entropy correlated significantly (p < 0.05) with both GS and median ADC. Contrast correlated moderately with median ADC. The textural features correlated insignificantly with K^trans and V_e. GS 4+3 cancers had significantly lower ASM and higher entropy than 3+4 cancers, but insignificant differences in median ADC, K^trans, and V_e. The combined texture-MRI parameters yielded higher classification accuracy (91%) than the individual parameter sets.

CONCLUSION

T2W MRI-derived textural features could serve as potential diagnostic markers, sensitive to the pathological differences in prostate cancers.

KEY POINTS

• T2W MRI-derived textural features correlate significantly with Gleason score and ADC. • T2W MRI-derived textural features differentiate Gleason score 3+4 from 4+3 cancers. • T2W image textural features could augment tumour characterization.

Characterization of Portal Vein Thrombosis (Neoplastic Versus Bland) on CT Images Using Software-Based Texture Analysis and Thrombus Density (Hounsfield Units)

OBJECTIVE

The purpose of this study was to investigate the role of CT texture analysis and thrombus density (measured in Hounsfield units) in distinguishing between neoplastic and bland portal vein thrombosis (PVT) on portal venous phase CT.

MATERIALS AND METHODS

In this retrospective study, 117 contrast-enhanced CT studies of 109 patients were included for characterization of PVT. Assessment of PVT was performed by estimation of CT textural features using CT texture analysis software and measurement of attenuation values. For CT texture analysis, filtered and unfiltered images were assessed to quantify heterogeneity using a set of predefined histogram-based texture parameters. The Mann-Whitney U test and binary logistic regression were applied for statistical significance. ROC curves were used to identify accuracy and optimal cutoff values.

RESULTS

Of the 117 CT studies, 63 neoplastic thrombi and 54 bland thrombi were identified on the images. The two most discriminative CT texture analysis parameters to differentiate neoplastic from bland thrombus were mean value of positive pixels (without filtration, p < 0.001) and entropy (with fine filtration, p < 0.001). Mean thrombus density values could also reliably distinguish neoplastic (81.39 HU) and bland (32.88 HU) thrombi (p < 0.001). The AUCs were 0.97 for mean value of positive pixels (p < 0.001), 0.93 for entropy (p < 0.001), 0.99 for the model combining mean value of positive pixels and entropy (p < 0.001), 0.91 for thrombus density (p < 0.001), and 0.61 for the radiologist's subjective evaluation (p = 0.037). The optimal cutoffs values were 56.9 for mean value of positive pixels, 4.50 for entropy, and 54.0 HU for thrombus density.

CONCLUSION

CT texture analysis and CT attenuation values allow reliable differentiation between neoplastic and bland thrombi on a single portal venous phase CT examination.

Prognostic value of computed tomography texture features in non-small cell lung cancers treated with definitive concomitant chemoradiotherapy

OBJECTIVES

The aim of this study was to investigate whether the computed tomography (CT) texture features of primary tumors are associated with the overall survival (OS) of non-small cell lung cancer (NSCLC) patients undergoing definitive concomitant chemoradiotherapy (CCRT).

MATERIALS AND METHODS

In this retrospective study, 98 patients (83 men and 15 women; mean age, 61.9 ± 8.0 years) with unresectable NSCLCs (stage IIIA, 45; stage IIIB, 53) underwent definitive CCRT at our institution from January 2006 to December 2011. Patients were followed up for 3 years or until death. The CT texture parameters of primary tumors were extracted from contrast-enhanced CT images taken before CCRT using an in-house software program. Each texture parameter was dichotomized based on their optimal cutoff values obtained from receiver operating characteristics curve analysis. Three-year OS was compared between the dichotomized subgroups using Kaplan-Meier analysis and the log-rank test. Multivariate Cox regression analysis was performed to determine significant prognostic factors.

RESULTS

The 3-year cumulative survival rate was 0.51. The mean 3-year OS was 24.0 months (95% confidence interval, 21.5-26.6 months). There were no significant differences in 3-year OS according to tumor stage or histologic subtypes. However, entropy (P = 0.030), skewness (P = 0.021), and mean attenuation (P = 0.030) were shown to be significantly associated with 3-year OS. Multivariate Cox regression analysis revealed that higher entropy (adjusted hazard ratio [HR],2.31; P = 0.040), higher skewness (adjusted HR,1.92; P = 0.046), and higher mean attenuation (adjusted HR,1.93; P = 0.028) were independent predictors of decreased 3-year OS.

CONCLUSIONS

Computed tomography texture features have the potential to be used as prognostic biomarkers in unresectable NSCLC patients undergoing definitive CCRT.

Radiomics: a new application from established techniques

The increasing use of biomarkers in cancer have led to the concept of personalized medicine for patients. Personalized medicine provides better diagnosis and treatment options available to clinicians. Radiological imaging techniques provide an opportunity to deliver unique data on different types of tissue. However, obtaining useful information from all radiological data is challenging in the era of "big data". Recent advances in computational power and the use of genomics have generated a new area of research termed Radiomics. Radiomics is defined as the high throughput extraction of quantitative imaging features or texture (radiomics) from imaging to decode tissue pathology and creating a high dimensional data set for feature extraction. Radiomic features provide information about the gray-scale patterns, inter-pixel relationships. In addition, shape and spectral properties can be extracted within the same regions of interest on radiological images. Moreover, these features can be further used to develop computational models using advanced machine learning algorithms that may serve as a tool for personalized diagnosis and treatment guidance.

Role of quantitative computed tomography texture analysis in the differentiation of primary lung cancer and granulomatous nodules

BACKGROUND

Texture analysis is a computer tool that enables quantification of gray-level patterns, pixel interrelationships, and spectral properties of an image. It can enhance visual methods of image analysis. Primary lung cancer and granulomatous nodules have identical CT imaging features. The purpose of this study was to assess the sensitivity and specificity of CT texture analysis in differentiating lung cancer and granulomas.

METHODS

This retrospective study evaluated 55 patients with primary lung cancer and granulomatous nodules who had contrast-enhanced (CE) and/or non-contrast-enhanced (NCE) CT within 3 months of biopsy. Textural features were extracted from 61 nodules. Mann-Whitney U tests were used to compare values for nodules. Receiver operating characteristic (ROC) curves were constructed and the area under the curve (AUC) calculated with histopathology as outcome. Combinations of features were entered as predictors in logistic regression models and optimal threshold criteria were used to estimate sensitivity and specificity.

RESULTS

The model generated by sum of squares, sum difference, and sum entropy features for NCE CT yielded 88% sensitivity and 92% specificity (AUC =0.90±0.06, P<0.0001). For nodules with fluorodeoxyglucose positron emission tomography (FDG-PET)/CT, sensitivity for detection of lung cancer was 79.2% (CI: 57.8-92.9%), specificity was 38.5% (CI: 13.9-68.4%) and accuracy was 64.8%.

CONCLUSIONS

Quantitative CT texture analysis has the potential to differentiate primary lung cancer and granulomatous lesions.

Texture analysis as imaging biomarker of tumoral response to neoadjuvant chemoradiotherapy in rectal cancer patients studied with 3-T magnetic resonance

OBJECTIVE

The aim of this study was to determine whether texture features of rectal cancer on T2-weighted (T2w) magnetic resonance images can predict tumoral response in patients treated with neoadjuvant chemoradiotherapy (CRT).

MATERIALS AND METHODS

We prospectively enrolled 15 consecutive patients (6 women, 63.2 ± 13.4 years) with rectal cancer, who underwent pretreatment and midtreatment 3-T magnetic resonance imaging. Treatment protocol consisted of neoadjuvant CRT with oxaliplatin and 5-fluorouracile. Texture analysis using a filtration-histogram technique was performed using a commercial research software algorithm (TexRAD Ltd, Somerset, England, United Kingdom) on unenhanced axial T2w images by manually delineating a region of interest around the tumor outline for the largest cross-sectional area. The technique selectively filters and extracts textures at different anatomic scales followed by quantification of the histogram using kurtosis, entropy, skewness, and mean value of positive pixels. After CRT, all patients underwent complete surgical resection and the surgical specimen served as the gold standard.

RESULTS

Six patients showed pathological complete response (pCR), and 4 patients, partial response (PR). Five patients were classified as nonresponders (NRs). Pretreatment medium texture-scale quantified as kurtosis was significantly lower in the pCR subgroup in comparison with the PR + NR subgroup (P = 0.01). Midtreatment kurtosis without filtration was significantly higher in pCR in comparison with PR + NR (P = 0.045). The change in kurtosis between midtreatment and pretreatment images was significantly lower in the PR + NR subgroup compared with the pCR subgroup (P = 0.038). Pretreatment area under the receiver operating characteristic curves, to discriminate between pCR and PR + NR, was significantly higher for kurtosis (0.907, P < 0.001) compared with all other parameters. The optimal cutoff value for pretreatment kurtosis was 0.19 or less. Using this value, the sensitivity and specificity for pCR prediction were 100% and 77.8%, respectively.

CONCLUSION

Texture parameters derived from T2w images of rectal cancer have the potential to act as imaging biomarkers of tumoral response to neoadjuvant CRT.

Three-dimensional texture analysis of contrast enhanced CT images for treatment response assessment in Hodgkin lymphoma: comparison with F-18-FDG PET

PURPOSE

To determine the diagnostic performance of three-dimensional (3D) texture analysis (TA) of contrast-enhanced computed tomography (CE-CT) images for treatment response assessment in patients with Hodgkin lymphoma (HL), compared with F-18-fludeoxyglucose (FDG) positron emission tomography/CT.

METHODS

3D TA of 48 lymph nodes in 29 patients was performed on venous-phase CE-CT images before and after chemotherapy. All lymph nodes showed pathologically elevated FDG uptake at baseline. A stepwise logistic regression with forward selection was performed to identify classic CT parameters and texture features (TF) that enable the separation of complete response (CR) and persistent disease.

RESULTS

The TF fraction of image in runs, calculated for the 45° direction, was able to correctly identify CR with an accuracy of 75%, a sensitivity of 79.3%, and a specificity of 68.4%. Classical CT features achieved an accuracy of 75%, a sensitivity of 86.2%, and a specificity of 57.9%, whereas the combination of TF and CT imaging achieved an accuracy of 83.3%, a sensitivity of 86.2%, and a specificity of 78.9%.

CONCLUSIONS

3D TA of CE-CT images is potentially useful to identify nodal residual disease in HL, with a performance comparable to that of classical CT parameters. Best results are achieved when TA and classical CT features are combined.

Preliminary data using computed tomography texture analysis for the classification of hypervascular liver lesions: generation of a predictive model on the basis of quantitative spatial frequency measurements--a work in progress

OBJECTIVE

Computed tomography texture analysis (CTTA) is a method of quantifying lesion heterogeneity based on distribution of pixel intensities within a region of interest. This study investigates the ability of CTTA to distinguish different hypervascular liver lesions and compares CTTA parameters by creating a proof-of-concept model to distinguish between different lesions.

METHODS

Following institutional review board approval, CTTA software (TexRAD Ltd) was used to retrospectively analyze 17 cases of focal nodular hyperplasia, 19 hepatic adenomas, 25 hepatocellular carcinomas, and 19 cases of normal liver parenchyma using arterial phase scans. Two radiologists read the same image series used by the CTTA software and reported their best guess diagnosis. Computed tomography texture analysis parameters were computed from regions of interest using spatial band-pass filters to quantify heterogeneity. Random-forest method was used to construct a predictive model from these parameters, and a separate regression model was created using a subset of parameters.

RESULTS

The random-forest model successfully distinguished the 3 lesion types and normal liver with predicted classification performance accuracy for 91.2% for adenoma, 94.4% for focal nodular hyperplasia, and 98.6% for hepatocellular carcinoma. This error prediction was generated using a subset of data points not used in generation of the model, but not on discrete prospective cases. In contrast, the 2 human readers using the same image series data analyzed by the CTTA software had lower accuracies, of 72.2% and 65.6%, respectively. The explicit regression model with a subset of image parameters had intermediate overall accuracy of 84.9%.

CONCLUSIONS

Computed tomography texture analysis may prove valuable in lesion characterization. Differentiation between common hypervascular lesion types could be aided by the judicious incorporation of texture parameters into clinical analysis.

Quantitative Imaging in Radiation Oncology: An Emerging Science and Clinical Service

Radiation oncology has long required quantitative imaging approaches for the safe and effective delivery of radiation therapy. The past 10 years has seen a remarkable expansion in the variety of novel imaging signals and analyses that are starting to contribute to the prescription and design of the radiation treatment plan. These include a rapid increase in the use of magnetic resonance imaging, development of contrast-enhanced imaging techniques, integration of fluorinated deoxyglucose-positron emission tomography, evaluation of hypoxia imaging techniques, and numerous others. These are reviewed with an effort to highlight challenges related to quantification and reproducibility. In addition, several of the emerging applications of these imaging approaches are also highlighted. Finally, the growing community of support for establishing quantitative imaging approaches as we move toward clinical evaluation is summarized and the need for a clinical service in support of the clinical science and delivery of care is proposed.

Texture analysis of preoperative CT images for prediction of postoperative hepatic insufficiency: a preliminary study

BACKGROUND

Texture analysis is a promising method of analyzing imaging data to potentially enhance diagnostic capability. This approach involves automated measurement of pixel intensity variation that may offer further insight into disease progression than do standard imaging techniques alone. We postulated that postoperative liver insufficiency, a major source of morbidity and mortality, correlates with preoperative heterogeneous parenchymal enhancement that can be quantified with texture analysis of cross-sectional imaging.

STUDY DESIGN

A retrospective case-matched study (waiver of informed consent and HIPAA authorization, approved by the Institutional Review Board) was performed comparing patients who underwent major hepatic resection and developed liver insufficiency (n = 12) with a matched group of patients with no postoperative liver insufficiency (n = 24) by procedure, remnant volume, and year of procedure. Texture analysis (with gray-level co-occurrence matrices) was used to quantify the heterogeneity of liver parenchyma on preoperative CT scans. Statistical significance was evaluated using Wilcoxon's signed rank and Pearson's chi-square tests.

RESULTS

No statistically significant differences were found between study groups for preoperative patient demographics and clinical characteristics, with the exception of sex (p < 0.05). Two texture features differed significantly between the groups: correlation (linear dependency of gray levels on neighboring pixels) and entropy (randomness of brightness variation) (p < 0.05).

CONCLUSIONS

In this preliminary study, the texture of liver parenchyma on preoperative CT was significantly more varied, less symmetric, and less homogeneous in patients with postoperative liver insufficiency. Therefore, texture analysis has the potential to provide an additional means of preoperative risk stratification.

CT texture analysis of renal masses: pilot study using random forest classification for prediction of pathology

RATIONALE AND OBJECTIVES

Computed tomography texture analysis (CTTA) allows quantification of heterogeneity within a region of interest. This study investigates the possibility of distinguishing between several common renal masses using CTTA-derived parameters by developing and validating a predictive model.

MATERIALS AND METHODS

CTTA software was used to analyze 20 clear cell renal cell carcinomas (RCCs), 20 papillary RCCs, 20 oncocytomas, and 20 renal cysts. Regions of interest were drawn around each mass on multiple slices in the arterial, venous, and delayed phases on renal mass protocol CT scans. Unfiltered images and spatial band-pass filtered images were analyzed to quantify heterogeneity. Random forest method was used to construct a predictive model to classify lesions using quantitative parameters. The model was externally validated on a separate set of 19 unknown cases.

RESULTS

The random forest model correctly categorized oncocytomas in 89% of cases (sensitivity = 89%, specificity = 99%), clear cell RCCs in 91% of cases (sensitivity = 91%, specificity = 97%), cysts in 100% of cases (sensitivity = 100%, specificity = 100%), and papillary RCCs in 100% of cases (sensitivity = 100%, specificity = 98%).

CONCLUSIONS

CTTA, in conjunction with random forest modeling, demonstrates promise as a tool to characterize lesions. Various renal masses were accurately classified using quantitative information derived from routine scans.

High quality machine-robust image features: identification in nonsmall cell lung cancer computed tomography images

PURPOSE

For nonsmall cell lung cancer (NSCLC) patients, quantitative image features extracted from computed tomography (CT) images can be used to improve tumor diagnosis, staging, and response assessment. For these findings to be clinically applied, image features need to have high intra and intermachine reproducibility. The objective of this study is to identify CT image features that are reproducible, nonredundant, and informative across multiple machines.

METHODS

Noncontrast-enhanced, test-retest CT image pairs were obtained from 56 NSCLC patients imaged on three CT machines from two institutions. Two machines ("M1" and "M2") used cine 4D-CT and one machine ("M3") used breath-hold helical 3D-CT. Gross tumor volumes (GTVs) were semiautonomously segmented then pruned by removing voxels with CT numbers less than a prescribed Hounsfield unit (HU) cutoff. Three hundred and twenty eight quantitative image features were extracted from each pruned GTV based on its geometry, intensity histogram, absolute gradient image, co-occurrence matrix, and run-length matrix. For each machine, features with concordance correlation coefficient values greater than 0.90 were considered reproducible. The Dice similarity coefficient (DSC) and the Jaccard index (JI) were used to quantify reproducible feature set agreement between machines. Multimachine reproducible feature sets were created by taking the intersection of individual machine reproducible feature sets. Redundant features were removed through hierarchical clustering based on the average correlation between features across multiple machines.

RESULTS

For all image types, GTV pruning was found to negatively affect reproducibility (reported results use no HU cutoff). The reproducible feature percentage was highest for average images (M1 = 90.5%, M2 = 94.5%, M1∩M2 = 86.3%), intermediate for end-exhale images (M1 = 75.0%, M2 = 71.0%, M1∩M2 = 52.1%), and lowest for breath-hold images (M3 = 61.0%). Between M1 and M2, the reproducible feature sets generated from end-exhale images were relatively machine-sensitive (DSC = 0.71, JI = 0.55), and the reproducible feature sets generated from average images were relatively machine-insensitive (DSC = 0.90, JI = 0.87). Histograms of feature pair correlation distances indicated that feature redundancy was machine-sensitive and image type sensitive. After hierarchical clustering, 38 features, 28 features, and 33 features were found to be reproducible and nonredundant for M1∩M2 (average images), M1∩M2 (end-exhale images), and M3, respectively. When blinded to the presence of test-retest images, hierarchical clustering showed that the selected features were informative by correctly pairing 55 out of 56 test-retest images using only their reproducible, nonredundant feature set values.

CONCLUSIONS

Image feature reproducibility and redundancy depended on both the CT machine and the CT image type. For each image type, the authors found a set of cross-machine reproducible, nonredundant, and informative image features that would be useful for future image-based models. Compared to end-exhale 4D-CT and breath-hold 3D-CT, average 4D-CT derived image features showed superior multimachine reproducibility and are the best candidates for clinical correlation.

New techniques for assessing response after hypofractionated radiotherapy for lung cancer

Hypofractionated radiotherapy (HFRT) is an effective and increasingly-used treatment for early stage non-small cell lung cancer (NSCLC). Stereotactic ablative radiotherapy (SABR) is a form of HFRT and delivers biologically effective doses (BEDs) in excess of 100 Gy10 in 3-8 fractions. Excellent long-term outcomes have been reported; however, response assessment following SABR is complicated as radiation induced lung injury can appear similar to a recurring tumor on CT. Current approaches to scoring treatment responses include Response Evaluation Criteria in Solid Tumors (RECIST) and positron emission tomography (PET), both of which appear to have a limited role in detecting recurrences following SABR. Novel approaches to assess response are required, but new techniques should be easily standardized across centers, cost effective, with sensitivity and specificity that improves on current CT and PET approaches. This review examines potential novel approaches, focusing on the emerging field of quantitative image feature analysis, to distinguish recurrence from fibrosis after SABR.