Texture Analysis of Imaging: What Radiologists Need to Know. AJR Am J Roentgenol. 2019;212:520-528. [PMID: 30645163 DOI: 10.2214/ajr.18.20624]

Volumetric CT Texture Analysis of Intrahepatic Mass-Forming Cholangiocarcinoma for the Prediction of Postoperative Outcomes: Fully Automatic Tumor Segmentation Versus Semi-Automatic Segmentation

Objective

To determine whether volumetric CT texture analysis (CTTA) using fully automatic tumor segmentation can help predict recurrence-free survival (RFS) in patients with intrahepatic mass-forming cholangiocarcinomas (IMCCs) after surgical resection.

Materials and Methods

This retrospective study analyzed the preoperative CT scans of 89 patients with IMCCs (64 male; 25 female; mean age, 62.1 years; range, 38–78 years) who underwent surgical resection between January 2005 and December 2016. Volumetric CTTA of IMCCs was performed in late arterial phase images using both fully automatic and semi-automatic liver tumor segmentation techniques. The time spent on segmentation and texture analysis was compared, and the first-order and second-order texture parameters and shape features were extracted. The reliability of CTTA parameters between the techniques was evaluated using intraclass correlation coefficients (ICCs). Intra- and interobserver reproducibility of volumetric CTTAs were also obtained using ICCs. Cox proportional hazard regression were used to predict RFS using CTTA parameters and clinicopathological parameters.

Results

The time spent on fully automatic tumor segmentation and CTTA was significantly shorter than that for semi-automatic segmentation: mean ± standard deviation of 1 minutes 37 seconds ± 50 seconds vs. 10 minutes 48 seconds ± 13 minutes 44 seconds (p < 0.001). ICCs of the texture features between the two techniques ranged from 0.215 to 0.980. ICCs for the intraobserver and interobserver reproducibility using fully automatic segmentation were 0.601–0.997 and 0.177–0.984, respectively. Multivariable analysis identified lower first-order mean (hazard ratio [HR], 0.982; p = 0.010), larger pathologic tumor size (HR, 1.171; p < 0.001), and positive lymph node involvement (HR, 2.193; p = 0.014) as significant parameters for shorter RFS using fully automatic segmentation.

Conclusion

Volumetric CTTA parameters obtained using fully automatic segmentation could be utilized as prognostic markers in patients with IMCC, with comparable reproducibility in significantly less time compared with semi-automatic segmentation.

Role of artificial intelligence in multidisciplinary imaging diagnosis of gastrointestinal diseases

The use of artificial intelligence-based tools is regarded as a promising approach to increase clinical efficiency in diagnostic imaging, improve the interpretability of results, and support decision-making for the detection and prevention of diseases. Radiology, endoscopy and pathology images are suitable for deep-learning analysis, potentially changing the way care is delivered in gastroenterology. The aim of this review is to examine the key aspects of different neural network architectures used for the evaluation of gastrointestinal conditions, by discussing how different models behave in critical tasks, such as lesion detection or characterization (i.e. the distinction between benign and malignant lesions of the esophagus, the stomach and the colon). To this end, we provide an overview on recent achievements and future prospects in deep learning methods applied to the analysis of radiology, endoscopy and histologic whole-slide images of the gastrointestinal tract.

A Novel Radiomics Nomogram for the Prediction of Secondary Loss of Response to Infliximab in Crohn's Disease

Purpose

The prediction of the loss of response (LOR) to infliximab (IFX) is crucial for optimizing treatment strategies and shifting biologics. However, a secondary LOR is difficult to predict by endoscopy due to the intestinal stricture, perforation, and fistulas. This study aimed to develop and validate a radiomic nomogram for the prediction of secondary LOR to IFX in patients with Crohn’s disease (CD).

Patients and Methods

A total of 186 biologic-naive patients diagnosed with CD between September 2016 and June 2019 were enrolled. Secondary LOR was determined during week 54. Computed tomography enterography (CTE) texture analysis (TA) features were extracted from lesions and analyzed using LIFEx software. Feature selection was performed by least absolute shrinkage and selection operator (LASSO) and ten-fold cross validation. A nomogram was constructed using multivariable logistic regression, and the internal validation was approached by ten-fold cross validation.

Results

Predictors contained in the radiomics nomogram included three first-order and five second-order signatures. The prediction model presented significant discrimination (AUC, 0.880; 95% CI, 0.816–0.944) and high calibration (mean absolute error of = 0.028). Decision curve analysis (DCA) indicated that the nomogram provided clinical net benefit. Ten-fold cross validation assessed the stability of the nomogram with an AUC of 0.817 and an accuracy of 0.819.

Conclusion

This novel radiomics nomogram provides a predictive tool to assess secondary LOR to IFX in patients with Crohn’s disease. This tool will help physicians decide when to switch therapy.

Volumetric ADC histogram analysis for preoperative evaluation of LVSI status in stage I endometrioid adenocarcinoma

OBJECTIVES

To investigate the value of volumetric ADC histogram metrics for evaluating lymphovascular space invasion (LVSI) status in stage I endometrioid adenocarcinoma (EAC).

METHODS

Preoperative MRI of 227 patients with stage I EAC were retrospectively analyzed. ADC histogram data were derived from the whole tumor with ROIs drawn on all slices of DWI scans (b = 0, 1000 s/mm²). The Student t-test was performed to compare ADC histogram metrics (minADC, maxADC, and meanADC; 10th, 25th, 50th, 75th, and 90th percentiles of ADC; skewness; and kurtosis) between the LVSI-positive and LVSI-negative groups, as well as between stage Ia and Ib EACs. ROC curve analysis was carried out to evaluate the diagnostic performance of ADC histogram metrics in predicting LVSI status in EAC.

RESULTS

The minADC and meanADC and 10th, 25th, 50th, 75th, and 90th percentiles of ADC were significantly lower in LVSI-positive EACs compared with those in the LVSI-negative groups for stage I, Ia, and Ib EACs (all p < 0.05). MeanADC ≤ 0.857 × 10^-3 mm²/s, meanADC ≤ 0.854 × 10^-3 mm²/s, and the 90th percentile of ADC ≤ 1.06 × 10^-3 mm²/s yielded the largest AUC of 0.844, 0.844, and 0.849 for evaluating LVSI positivity in stage I, Ia, and Ib tumors, respectively, with sensitivity of 75.4%, 75.0%, and 76.2%; specificity of 80.0%, 83.1%, and 82.1%; and accuracy of 79.3%, 81.5%, and 79.6%, respectively.

CONCLUSION

Volumetric ADC histogram metrics might be helpful for the preoperative evaluation of LVSI status and personalized clinical management in patients with stage I EAC.

KEY POINTS

• Volumetric ADC histogram analysis helps evaluate LVSI status preoperatively. • LVSI-positive EAC is associated with a reduction in multiple volumetric ADC histogram metrics. • MeanADC and the 90th percentile of ADC were shown to be best in evaluating LVSI- positivity in stage Ia and Ib EACs, respectively.

CT and MRI radiomics of bone and soft-tissue sarcomas: a systematic review of reproducibility and validation strategies

Background

Feature reproducibility and model validation are two main challenges of radiomics. This study aims to systematically review radiomic feature reproducibility and predictive model validation strategies in studies dealing with CT and MRI radiomics of bone and soft-tissue sarcomas. The ultimate goal is to promote achieving a consensus on these aspects in radiomic workflows and facilitate clinical transferability.

Results

Out of 278 identified papers, forty-nine papers published between 2008 and 2020 were included. They dealt with radiomics of bone (n = 12) or soft-tissue (n = 37) tumors. Eighteen (37%) studies included a feature reproducibility analysis. Inter-/intra-reader segmentation variability was the theme of reproducibility analysis in 16 (33%) investigations, outnumbering the analyses focused on image acquisition or post-processing (n = 2, 4%). The intraclass correlation coefficient was the most commonly used statistical method to assess reproducibility, which ranged from 0.6 and 0.9. At least one machine learning validation technique was used for model development in 25 (51%) papers, and K-fold cross-validation was the most commonly employed. A clinical validation of the model was reported in 19 (39%) papers. It was performed using a separate dataset from the primary institution (i.e., internal validation) in 14 (29%) studies and an independent dataset related to different scanners or from another institution (i.e., independent validation) in 5 (10%) studies.

Conclusions

The issues of radiomic feature reproducibility and model validation varied largely among the studies dealing with musculoskeletal sarcomas and should be addressed in future investigations to bring the field of radiomics from a preclinical research area to the clinical stage.

Radiomics Nomograms Based on Non-enhanced MRI and Clinical Risk Factors for the Differentiation of Chondrosarcoma from Enchondroma

BACKGROUND

Differentiating chondrosarcoma from enchondroma using conventional MRI remains challenging. An effective method for accurate preoperative diagnosis could affect the management and prognosis of patients.

PURPOSE

To validate and evaluate radiomics nomograms based on non-enhanced MRI and clinical risk factors for the differentiation of chondrosarcoma from enchondroma.

STUDY TYPE

Retrospective.

POPULATION

A total of 103 patients with pathologically confirmed chondrosarcoma (n = 53) and enchondroma (n = 50) were randomly divided into training (n = 68) and validation (n = 35) groups.

FIELD STRENGTH/SEQUENCE

Axial non-contrast-enhanced T1-weighted images (T1WI) and fat-suppressed T2-weighted images (T2WI-FS) were acquired at 3.0 T.

ASSESSMENT

Clinical risk factors (sex, age, and tumor location) and diagnosis assessment based on morphologic MRI by three radiologists were recorded. Three radiomics signatures were established based on the T1WI, T2WI-FS, and T1WI + T2WI-FS sequences. Three clinical radiomics nomograms were developed based on the clinical risk factors and three radiomics signatures.

STATISTICAL TESTS

The area under the receiver operating characteristic curve (AUC) was used to evaluate the performance of radiomics signatures and clinical radiomics nomograms.

RESULTS

Tumor location was an important clinical risk factor (P < 0.05). The radiomics signature based on T1WI and T1WI + T2WI-FS features performed better than that based on T2WI-FS in the validation group (AUC in the validation group: 0.961, 0.938, and 0.833, respectively; P < 0.05). In the validation group, the three clinical radiomics nomograms (T1WI, T2WI-FS, and T1WI + T2WI-FS) achieved AUCs of 0.938, 0.935, and 0.954, respectively. In all patients, the clinical radiomics nomogram based on T2WI-FS (AUC = 0.967) performed better than that based on T2WI-FS (AUC = 0.901, P < 0.05).

DATA CONCLUSION

The proposed clinical radiomics nomogram showed promising performance in differentiating chondrosarcoma from enchondroma.

LEVEL OF EVIDENCE

4 TECHNICAL EFFICACY: Stage 2.

Machine learning-based radiomic evaluation of treatment response prediction in glioblastoma

AIM

To investigate machine learning based models combining clinical, radiomic, and molecular information to distinguish between early true progression (tPD) and pseudoprogression (psPD) in patients with glioblastoma.

MATERIALS AND METHODS

A retrospective analysis was undertaken of 76 patients (46 tPD, 30 psPD) with early enhancing disease following chemoradiotherapy for glioblastoma. Outcome was determined on follow-up until 6 months post-chemoradiotherapy. Models comprised clinical characteristics, O⁶-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, and 307 quantitative imaging features extracted from enhancing disease and perilesional oedema masks on early post-chemoradiotherapy contrast-enhanced T1-weighted imaging, T2-weighted imaging (T2WI), and apparent diffusion coefficient (ADC) maps. Feature selection was performed within bootstrapped cross-validated recursive feature elimination with a random forest algorithm. Naive Bayes five-fold cross-validation was used to validate the final model.

RESULTS

Top selected features included age, MGMT promoter methylation status, two shape-based features from the enhancing disease mask, three radiomic features from the enhancing disease mask on ADC, and one radiomic feature from the perilesional oedema mask on T2WI. The final model had an area under the receiver operating characteristics curve (AUC) of 0.80, sensitivity 78.2%, specificity 66.7%, and accuracy of 73.7%.

CONCLUSION

Incorporating a machine learning-based approach using quantitative radiomic features from standard-of-care magnetic resonance imaging (MRI), in combination with clinical characteristics and MGMT promoter methylation status has a complementary effect and improves model performance for early prediction of glioblastoma treatment response.

Whole-tumor 3D volumetric MRI-based radiomics approach for distinguishing between benign and malignant soft tissue tumors

OBJECTIVES

Our purpose was to differentiate between malignant from benign soft tissue neoplasms using a combination of MRI-based radiomics metrics and machine learning.

METHODS

Our retrospective study identified 128 histologically diagnosed benign (n = 36) and malignant (n = 92) soft tissue lesions. 3D ROIs were manually drawn on 1 sequence of interest and co-registered to other sequences obtained during the same study. One thousand seven hundred eight radiomics features were extracted from each ROI. Univariate analyses with supportive ROC analyses were conducted to evaluate the discriminative power of predictive models constructed using Real Adaptive Boosting (Adaboost) and Random Forest (RF) machine learning approaches.

RESULTS

Univariate analyses demonstrated that 36.89% of individual radiomics varied significantly between benign and malignant lesions at the p ≤ 0.05 level. Adaboost and RF performed similarly well, with AUCs of 0.77 (95% CI 0.68-0.85) and 0.72 (95% CI 0.63-0.81), respectively, after 10-fold cross-validation. Restricting the machine learning models to only sequences extracted from T2FS and STIR sequences maintained comparable performance, with AUCs of 0.73 (95% CI 0.64-0.82) and 0.75 (95% CI 0.65-0.84), respectively.

CONCLUSION

Machine learning decision classifiers constructed from MRI-based radiomics features show promising ability to preoperatively discriminate between benign and malignant soft tissue masses. Our approach maintains applicability even when the dataset is restricted to T2FS and STIR fluid-sensitive sequences, which may bolster practicality in clinical application scenarios by eliminating the need for complex co-registrations for multisequence analysis.

KEY POINTS

• Predictive models constructed from MRI-based radiomics data and machine learning-augmented approaches yielded good discriminative power to correctly classify benign and malignant lesions on preoperative scans, with AUCs of 0.77 (95% CI 0.68-0.85) and 0.72 (95% CI 0.63-0.81) for Real Adaptive Boosting (Adaboost) and Random Forest (RF), respectively. • Restricting the models to only use metrics extracted from T2 fat-saturated (T2FS) and Short-Tau Inversion Recovery (STIR) sequences yielded similar performance, with AUCs of 0.73 (95% CI 0.64-0.82) and 0.75 (95% CI 0.65-0.84) for Adaboost and RF, respectively. • Radiomics-based machine learning decision classifiers constructed from multicentric data more closely mimic the real-world practice environment and warrant additional validation ahead of prospective implementation into clinical workflows.

Noninvasive prediction of residual disease for advanced high-grade serous ovarian carcinoma by MRI-based radiomic-clinical nomogram

OBJECTIVES

To develop a preoperative MRI-based radiomic-clinical nomogram for prediction of residual disease (RD) in patients with advanced high-grade serous ovarian carcinoma (HGSOC).

METHODS

In total, 217 patients with advanced HGSOC were enrolled from January 2014 to June 2019 and randomly divided into a training set (n = 160) and a validation set (n = 57). Finally, 841 radiomic features were extracted from each tumor on T2-weighted imaging (T2WI) and contrast-enhanced T1-weighted imaging (CE-T1WI) sequence, respectively. We used two fusion methods, the maximal volume of interest (MV) and the maximal feature value (MF), to fuse the radiomic features of bilateral tumors, so that patients with bilateral tumors have the same kind of radiomic features as patients with unilateral tumors. The radiomic signatures were constructed by using mRMR method and LASSO classifier. Multivariable logistic regression analysis was used to develop a radiomic-clinical nomogram incorporating radiomic signature and conventional clinico-radiological features. The performance of the nomogram was evaluated on the validation set.

RESULTS

In total, 342 tumors from 217 patients were analyzed in this study. The MF-based radiomic signature showed significantly better prediction performance than the MV-based radiomic signature (AUC = 0.744 vs. 0.650, p = 0.047). By incorporating clinico-radiological features and MF-based radiomic signature, radiomic-clinical nomogram showed favorable prediction ability with an AUC of 0.803 in the validation set, which was significantly higher than that of clinico-radiological signature and MF-based radiomic signature (AUC = 0.623, 0.744, respectively).

CONCLUSIONS

The proposed MRI-based radiomic-clinical nomogram provides a promising way to noninvasively predict the RD status.

KEY POINTS

• MRI-based radiomic-clinical nomogram is feasible to noninvasively predict residual disease in patients with advanced HGSOC. • The radiomic signature based on MF showed significantly better prediction performance than that based on MV. • The radiomic-clinical nomogram showed a favorable prediction ability with an AUC of 0.803.

Assessment of Renal Cell Carcinoma by Texture Analysis in Clinical Practice: A Six-Site, Six-Platform Analysis of Reliability

Background: Multiple commercial and open-source software applications are available for texture analysis. Nonstandard techniques can cause undesirable variability that impedes result reproducibility and limits clinical utility. Objective: The purpose of this study is to measure agreement of texture metrics extracted by 6 software packages. Methods: This retrospective study included 40 renal cell carcinomas with contrast-enhanced CT from The Cancer Genome Atlas and Imaging Archive. Images were analyzed by 7 readers at 6 sites. Each reader used 1 of 6 software packages to extract commonly studied texture features. Inter and intra-reader agreement for segmentation was assessed with intra-class correlation coefficients. First-order (available in 6 packages) and second-order (available in 3 packages) texture features were compared between software pairs using Pearson correlation. Results: Inter- and intra-reader agreement was excellent (ICC 0.93-1). First-order feature correlations were strong (r>0.8, p<0.001) between 75% (21/28) of software pairs for mean and standard deviation, 48% (10/21) for entropy, 29% (8/28) for skewness, and 25% (7/28) for kurtosis. Of 15 second-order features, only co-occurrence matrix correlation, grey-level non-uniformity, and run-length non-uniformity showed strong correlation between software packages (0.90-1, p<0.001). Conclusion: Variability in first and second order texture features was common across software configurations and produced inconsistent results. Standardized algorithms and reporting methods are needed before texture data can be reliably used for clinical applications. Clinical Impact: It is important to be aware of variability related to texture software processing and configuration when reporting and comparing outputs.

Combining machine learning and texture analysis to differentiate mediastinal lymph nodes in lung cancer patients

Evaluate whether texture analysis associated with machine learning approaches could differentiate between malignant and benign lymph nodes. A total 18 patients with lung cancer were selected, with 39 lymph nodes, being 15 malignant and 24 benign. Retrospective computed tomography scans were utilized both with and without contrast medium. The great differential of this work was the use of 15 textures from mediastinal lymph nodes, with five different physicians as operators. First and second order statistical textures such as gray level run length and co-occurrence matrix were extracted and applied to three different machine learning classifiers. The best machine learning classifier demonstrated a variability of less than 5% among operators. The support vector machine (SVM) classifier presented 95% of the area under the ROC curve (AUC) and 89% of sensitivity for sequences without contrast medium. SVM classifier presented 93% of AUC and 86% of sensitivity for sequences with contrast medium. Texture analysis and machine learning may be helpful in the differentiation between malign and benign lymph nodes. This study can aid the physician in diagnosis and staging of lymph nodes and potentially reduce the number of invasive analysis to histopathological confirmation.

A CT radiomics analysis of COVID-19-related ground-glass opacities and consolidation: Is it valuable in a differential diagnosis with other atypical pneumonias?

PURPOSE

To evaluate the discrimination of parenchymal lesions between COVID-19 and other atypical pneumonia (AP) by using only radiomics features.

METHODS

In this retrospective study, 301 pneumonic lesions (150 ground-glass opacity [GGO], 52 crazy paving [CP], 99 consolidation) obtained from nonenhanced thorax CT scans of 74 AP (46 male and 28 female; 48.25±13.67 years) and 60 COVID-19 (39 male and 21 female; 48.01±20.38 years) patients were segmented manually by two independent radiologists, and Location, Size, Shape, and First- and Second-order radiomics features were calculated.

RESULTS

Multiple parameters showed significant differences between AP and COVID-19-related GGOs and consolidations, although only the Range parameter was significantly different for CPs. Models developed by using the Bayesian information criterion (BIC) for the whole group of GGO and consolidation lesions predicted COVID-19 consolidation and AP GGO lesions with low accuracy (46.1% and 60.8%, respectively). Thus, instead of subjective classification, lesions were reclassified according to their skewness into positive skewness group (PSG, 78 AP and 71 COVID-19 lesions) and negative skewness group (NSG, 56 AP and 44 COVID-19 lesions), and group-specific models were created. The best AUC, accuracy, sensitivity, and specificity were respectively 0.774, 75.8%, 74.6%, and 76.9% among the PSG models and 0.907, 83%, 79.5%, and 85.7% for the NSG models. The best PSG model was also better at predicting NSG lesions smaller than 3 mL. Using an algorithm, 80% of COVID-19 and 81.1% of AP patients were correctly predicted.

CONCLUSION

During periods of increasing AP, radiomics parameters may provide valuable data for the differential diagnosis of COVID-19.

Quality control and whole-gland, zonal and lesion annotations for the PROSTATEx challenge public dataset

PURPOSE

Radiomic features are promising quantitative parameters that can be extracted from medical images and employed to build machine learning predictive models. However, generalizability is a key concern, encouraging the use of public image datasets. We performed a quality assessment of the PROSTATEx training dataset and provide publicly available lesion, whole-gland, and zonal anatomy segmentation masks.

METHOD

Two radiology residents and two experienced board-certified radiologists reviewed the 204 prostate MRI scans (330 lesions) included in the training dataset. The quality of provided lesion coordinate was scored using the following scale: 0 = perfectly centered, 1 = within lesion, 2 = within the prostate without lesion, 3 = outside the prostate. All clearly detectable lesions were segmented individually slice-by-slice on T2-weighted and apparent diffusion coefficient images. With the same methodology, volumes of interest including the whole gland, transition, and peripheral zones were annotated.

RESULTS

Of the 330 available lesion identifiers, 3 were duplicates (1%). From the remaining, 218 received score = 0, 74 score = 1, 31 score = 2 and 4 score = 3. Overall, 299 lesions were verified and segmented. Independently of lesion coordinate score and other issues (e.g., lesion coordinates falling outside DICOM images, artifacts etc.), the whole prostate gland and zonal anatomy were also manually annotated for all cases.

CONCLUSION

While several issues were encountered evaluating the original PROSTATEx dataset, the improved quality and availability of lesion, whole-gland and zonal segmentations will increase its potential utility as a common benchmark in prostate MRI radiomics.

FET-PET radiomics in recurrent glioblastoma: prognostic value for outcome after re-irradiation?

Purpose

The value of O-(2-[18F]fluoroethyl)-L-tyrosine (FET)-positron emission tomography (PET)-radiomics in the outcome assessment of patients with recurrent glioblastoma (rGBM) has not been evaluated until now. The aim of this study was to evaluate whether a prognostic model based on FET-PET radiomics features (RF) is feasible and can identify rGBM patients that would most benefit from re-irradiation.

Methods

We prospectively recruited rGBM patients who underwent FET-PET before re-irradiation (GLIAA-Pilot trial, DRKS00000633). Tumor volume was delineated using a semi-automatic method with a threshold of 1.8 times the standardized-uptake-value of the background. 135 FET-RF (histogram parameters, shape and texture features) were extracted. The analysis involved the characterization of tumor and non-tumor tissue with FET-RF and the evaluation of the prognostic value of FET-RF for time-to-progression (TTP), overall survival (OS) and recurrence location (RL).

Results

Thirty-two rGBM patients constituted our cohort. FET-RF discriminated significantly between tumor and non-tumor. The texture feature Small-Zone-Low-Gray-Level-Emphasis (SZLGE) showed the best performance for the prediction of TTP (p = 0.001, satisfying Bonferroni-multiple-test significance level). Additionally, two radiomics signatures could predict TTP (TTP-radiomics-signature, p = 0.001) and OS (OS-radiomics-signature, p = 0.038). SZLGE and the TTP-radiomics-signature additionally predicted RL. Specifically, high values for TTP-radiomics-signature and for SZLGE indicated not only earlier progression, but also a RL within the initial FET-PET active volume.

Conclusion

Our findings suggest that FET-PET radiomics could contribute to the prognostic assessment and selection of rGBM-patients benefiting from re-irradiation.

Trial registration DRKS00000633. Registered on 8th of December in 2010.

https://www.drks.de/drks_web/navigate.do?navigationId=trial.HTML&TRIAL_ID=DRKS00000633.

The role of CT texture analysis in predicting the clinical outcomes of acute ischemic stroke patients undergoing mechanical thrombectomy

OBJECTIVES

To evaluate the performance of CT-based texture analysis (TA) for predicting clinical outcomes of mechanical thrombectomy (MT) in acute ischemic stroke (AIS).

METHODS

This single-center, retrospective study contained 64 consecutive patients with AIS who underwent MT for large anterior circulation occlusion between December 2016 and January 2020. Patients were divided into 2 groups according to the modified Rankin scale (mRS) scores at 3 months as good outcome (mRS ≤ 2) and bad outcome (mRS > 2). Two observers examined the early ischemic changes for TA on baseline non-contrast CT images independently. Demographic, clinical, periprocedural, and texture variables were compared between the groups and ROC curves were made. Logistic regression analysis was used and a model was created to determine the independent predictors of a bad outcome.

RESULTS

Sixty-four patients (32 female, 32 male; mean age 63.03 ± 14.42) were included in the study. Fourteen texture parameters were significantly different between patients with good and bad outcomes. The long-run high gray-level emphasis (LRHGE), which is a gray-level run-length matrix (GLRLM) feature, showed the highest sensitivity (80%) and specificity (70%) rates to predict disability. The GLRLM_LRHGE value of > 4885.0 and the time from onset to puncture of > 237.5 mi were found as independent predictors of the bad outcome. The diagnostic rate was 80.0% when using the combination of the GLRLM_LRHGE and the time from onset to puncture cutoff values.

CONCLUSION

CT-based TA might be a promising modality to predict clinical outcome after MT in patients with AIS.

KEY POINTS

• The gray-level run-length matrix parameters displayed higher diagnostic performance among the texture features. • The long-run high gray-level emphasis showed the highest sensitivity and specificity rates for predicting a bad outcome in stroke patients undergoing mechanical thrombectomy. • The gray-level run-length matrix_long-run high gray-level emphasis value of > 4885.0 (OR = 11.06; 95% CI = 2.51 - 48.77; p = 0.001) and the time from onset to puncture of > 237.5 min (OR = 8.55; 95% CI = 1.96 - 37.21; p = 0.004) were found as independent predictors of the bad outcome.

Utility of texture analysis on T2-weighted MR for differentiating tumor deposits from mesorectal nodes in rectal cancer patients, in a retrospective cohort

OBJECTIVE

The purpose of the study was to evaluate the utility of MR texture analysis for differentiating tumor deposits from mesorectal nodes in rectal cancer.

MATERIALS AND METHODS

Pretreatment MRI of 40 patients performed between 2006 and 2018 with pathologically proven tumor deposits and/or malignant nodes in the setting of rectal cancer were retrospectively reviewed. In total, 25 tumor deposits (TDs) and 71 positive lymph nodes (LNs) were analyzed for morphological and first-order texture analysis features on T2-weighted axial images. MR morphological features (lesion shape, size, signal heterogeneity, contrast enhancement) were analyzed and agreed in consensus by two experienced radiologists followed by assessment with Fisher's exact test. Texture analysis of the lesions was performed using TexRAD, a proprietary software algorithm. First-order texture analysis features (mean, standard deviation, skewness, entropy, kurtosis, MPP) were obtained after applying spatial scaling filters (SSF; 0, 2, 3, 4, 5, 6). Univariate analysis was performed with non-parametric Mann-Whitney U test. The results of univariate analysis were reassessed with generalized estimating equations followed by multivariate analysis. Using histopathology as a gold standard, diagnostic accuracy was assessed by obtaining area under the receiver operating curve.

RESULTS

MR morphological parameter, lesion shape was a strong discriminator between TDs and LNs with a p value of 0.02 (AUC: 0.76, 95% CI of 0.66 to 0.84, SE: 0.06) and sensitivity, specificity of 90% and 68%, respectively. Skewness extracted at fine filter (SSF-2) was the only significant texture analysis parameter for distinguishing TDs from LNs with p value of 0.03 (AUC: 0.70, 95% CI of 0.59 to 0.79, SE: 0.06) and sensitivity, specificity of 70% and 72%, respectively. When lesion shape and skewness-2 were combined into a single model, the diagnostic accuracy was improved with AUC of 0.82 (SE: 0.05, 95% CI of 0.72 to 0.88 with p value of < 0.01). This model also showed a high sensitivity of 91% with specificity of 68%.

CONCLUSION

Lesion shape on MR can be a useful predictor for distinguishing TDs from positive LNs in rectal cancer patients. When interpreted along with MR texture parameter of skewness, accuracy is further improved.

Identification of robust and reproducible CT-texture metrics using a customized 3D-printed texture phantom

OBJECTIVE

The objective of this study was to evaluate the robustness and reproducibility of computed tomography-based texture analysis (CTTA) metrics extracted from CT images of a customized texture phantom built for assessing the association of texture metrics to three-dimensional (3D) printed progressively increasing textural heterogeneity.

MATERIALS AND METHODS

A custom-built 3D-printed texture phantom comprising of six texture patterns was used to evaluate the robustness and reproducibility of a radiomics panel under a variety of routine abdominal imaging protocols. The phantom was scanned on four CT scanners (Philips, Canon, GE, and Siemens) to assess reproducibility. The robustness assessment was conducted by imaging the texture phantom across different CT imaging parameters such as slice thickness, field of view (FOV), tube voltage, and tube current for each scanner. The texture panel comprised of 387 features belonging to 15 subgroups of texture extraction methods (e.g., Gray-level Co-occurrence Matrix: GLCM). Twelve unique image settings were tested on all the four scanners (e.g., FOV125). Interclass correlation two-way mixed with absolute agreement (ICC3) was used to assess the robustness and reproducibility of radiomic features. Linear regression was used to test the association between change in radiomic features and increased texture heterogeneity. Results were summarized in heat maps.

RESULTS

A total of 5612 (23.2%) of 24 090 features showed excellent robustness and reproducibility (ICC ≥ 0.9). Intensity, GLCM 3D, and gray-level run length matrix (GLRLM) 3D features showed best performance. Among imaging variables, changes in slice thickness affected all metrics more intensely compared to other imaging variables in reducing the ICC3. From the analysis of linear trend effect of the CTTA metrics, the top three metrics with high linear correlations across all scanners and scanning settings were from the GLRLM 2D/3D and discrete cosine transform (DCT) texture family.

CONCLUSION

The choice of scanner and imaging protocols affect texture metrics. Furthermore, not all CTTA metrics have a linear association with linearly varying texture patterns.

Texture Analysis in the Evaluation of Covid-19 Pneumonia in Chest X-Ray Images: a Proof of Concept Study

BACKGROUND

One of the most challenging aspects related to Covid-19 is to establish the presence of infection in early phase of the disease. Texture analysis might be an additional tool for the evaluation of Chest X-ray in patients with clinical suspicion of Covid-19 related pneumonia.

OBJECTIVE

To evaluate the diagnostic performance of texture analysis and machine learning models for the diagnosis of Covid-19 interstitial pneumonia in Chest X-ray images.

METHODS

Chest X-ray images were accessed from a publicly available repository (https://www.kaggle.com/tawsifurrahman/covid19-radiography-database). Lung areas were manually segmented using a polygonal regions of interest covering both lung areas, using MaZda, a freely available software for texture analysis. A total of 308 features per ROI was extracted. One hundred-ten Covid-19 Chest X-ray images were selected for the final analysis.

RESULTS

Six models, namely NB, GLM, DL, GBT, ANN and PLS-DA were selected and ensembled. According to Youden's index, the Covid-19 Ensemble Machine Learning Score showing the highest Area Under the Curve (0.971±0.015) was 132.57. Assuming this cut-off the Ensemble model performance was estimated evaluating both true and false positive/negative, resulting in 91.8% accuracy with 93% sensitivity and 90% specificity. Moving the cut-off value to -100, although the accuracy resulted lower (90.6%), the Ensemble Machine Learning showed 100% sensitivity, with 80% specificity.

CONCLUSION

Texture analysis of Chest X-ray images and machine learning algorithms may help in differentiating patients with Covid-19 pneumonia. Despite several limitations, this study can lay ground for future researches in this field and help developing more rapid and accurate screening tools for these patients.

MRI Texture Analysis for the Prediction of Stereotactic Radiosurgery Outcomes in Brain Metastases from Lung Cancer

This study aims to evaluate the utility of texture analysis in predicting the outcome of stereotactic radiosurgery (SRS) for brain metastases from lung cancer. From 83 patients with lung cancer who underwent SRS for brain metastasis, a total of 118 metastatic lesions were included. Two neuroradiologists independently performed magnetic resonance imaging (MRI)-based texture analysis using the Imaging Biomarker Explorer software. Inter-reader reliability as well as univariable and multivariable analyses were performed for texture features and clinical parameters to determine independent predictors for local progression-free survival (PFS) and overall survival (OS). Furthermore, Harrell’s concordance index (C-index) was used to assess the performance of the independent texture features. The primary tumor histology of small cell lung cancer (SCLC) was the only clinical parameter significantly associated with local PFS in multivariable analysis. Run-length non-uniformity (RLN) and short-run emphasis were the independent texture features associated with local PFS. In the non-SCLC (NSCLC) subgroup analysis, RLN and local range mean were associated with local PFS. The C-index of independent texture features was 0.79 for the all-patients group and 0.73 for the NSCLC subgroup. In conclusion, texture analysis on pre-treatment MRI of lung cancer patients with brain metastases may have a role in predicting SRS response.

Texture Analysis Using CT and Chemical Shift Encoding-Based Water-Fat MRI Can Improve Differentiation Between Patients With and Without Osteoporotic Vertebral Fractures

PURPOSE

Osteoporosis is a highly prevalent skeletal disease that frequently entails vertebral fractures. Areal bone mineral density (BMD) derived from dual-energy X-ray absorptiometry (DXA) is the reference standard, but has well-known limitations. Texture analysis can provide surrogate markers of tissue microstructure based on computed tomography (CT) or magnetic resonance imaging (MRI) data of the spine, thus potentially improving fracture risk estimation beyond areal BMD. However, it is largely unknown whether MRI-derived texture analysis can predict volumetric BMD (vBMD), or whether a model incorporating texture analysis based on CT and MRI may be capable of differentiating between patients with and without osteoporotic vertebral fractures.

MATERIALS AND METHODS

Twenty-six patients (15 females, median age: 73 years, 11 patients showing at least one osteoporotic vertebral fracture) who had CT and 3-Tesla chemical shift encoding-based water-fat MRI (CSE-MRI) available were analyzed. In total, 171 vertebral bodies of the thoracolumbar spine were segmented using an automatic convolutional neural network (CNN)-based framework, followed by extraction of integral and trabecular vBMD using CT data. For CSE-MRI, manual segmentation of vertebral bodies and consecutive extraction of the mean proton density fat fraction (PDFF) and T2* was performed. First-order, second-order, and higher-order texture features were derived from texture analysis using CT and CSE-MRI data. Stepwise multivariate linear regression models were computed using integral vBMD and fracture status as dependent variables.

RESULTS

Patients with osteoporotic vertebral fractures showed significantly lower integral and trabecular vBMD when compared to patients without fractures (p<0.001). For the model with integral vBMD as the dependent variable, T2* combined with three PDFF-based texture features explained 40% of the variance (adjusted R²[Ra2] = 0.40; p<0.001). Furthermore, regarding the differentiation between patients with and without osteoporotic vertebral fractures, a model including texture features from CT and CSE-MRI data showed better performance than a model based on integral vBMD and PDFF only ( Ra2 = 0.47 vs. Ra2  = 0.81; included texture features in the final model: integral vBMD, CT_Short-run_emphasis, CT_Varianceglobal, and PDFF_Variance).

CONCLUSION

Using texture analysis for spine CT and CSE-MRI can facilitate the differentiation between patients with and without osteoporotic vertebral fractures, implicating that future fracture prediction in osteoporosis may be improved.

Radiomics of hepatocellular carcinoma

The diagnosis of hepatocellular carcinoma relies largely on non-invasive imaging, and is well suited for radiomics analysis. Radiomics is an emerging method for quantification of tumor heterogeneity by mathematically analyzing the spatial distribution and relationships of gray levels in medical images. The published studies on radiomics analysis of HCC provide encouraging data demonstrating potential utility for prediction of tumor biology, molecular profiles, post-therapy response, and outcome. The combination of radiomics data and clinical/laboratory information provides added value in many studies. Radiomics is a multi-step process that requires optimization and standardization, the development of semi-automated or automated segmentation methods, robust data quality control, and refinement of algorithms and modeling approaches for high-throughput data analysis. While radiomics remains largely in the research setting, the strong associations of predictive models and nomograms with certain pathologic, molecular, and immune markers with tumor aggressiveness and patient outcomes, provide great potential for clinical applications to inform optimized treatment strategies and patient prognosis.

Stratification of cystic renal masses into benign and potentially malignant: applying machine learning to the bosniak classification

PURPOSE

To create a CT texture-based machine learning algorithm that distinguishes benign from potentially malignant cystic renal masses as defined by the Bosniak Classification version 2019.

METHODS

In this IRB-approved, HIPAA-compliant study, 4,454 adult patients underwent renal mass protocol CT or CT urography from January 2011 to June 2018. Of these, 257 cystic renal masses were included in the final study cohort. Each mass was independently classified using Bosniak version 2019 by three radiologists, resulting in 185 benign (Bosniak I or II) and 72 potentially malignant (Bosniak IIF, III or IV) masses. Six texture features: mean, standard deviation, mean of positive pixels, entropy, skewness, kurtosis were extracted using commercial software TexRAD (Feedback PLC, Cambridge, UK). Random forest (RF), logistic regression (LR), and support vector machine (SVM) machine learning algorithms were implemented to classify cystic renal masses into the two groups and tested with tenfold cross validations.

RESULTS

Higher mean, standard deviation, mean of positive pixels, entropy, skewness were statistically associated with the potentially malignant group (P ≤ 0.0015 each). Sensitivity, specificity, positive predictive value, negative predictive value, and area under curve of RF model was 0.67, 0.91, 0.75, 0.88, 0.88; of LR model was 0.63, 0.93, 0.78, 0.86, 0.90, and of SVM model was 0.56, 0.91, 0.71, 0.84, 0.89, respectively.

CONCLUSION

Three CT texture-based machine learning algorithms demonstrated high discriminatory capability in distinguishing benign from potentially malignant cystic renal masses as defined by the Bosniak Classification version 2019. If validated, CT texture-based machine learning algorithms may help reduce interreader variability when applying the Bosniak classification.

Texture Analysis Using Preoperative Positron Emission Tomography Images May Predict the Prognosis of Patients With Resectable Oral Squamous Cell Carcinoma

PURPOSE

Texture analysis is a computer-assisted technique used to measure intratumoral heterogeneity, which is known to have important roles in cancer research. This study aimed to assess the potential prognostic values of textural features extracted from preoperative ¹⁸F-fluorodeoxyglucose positron emission tomography images in patients with resectable oral squamous cell carcinoma.

PATIENTS AND METHODS

This retrospective cohort study included patients with oral squamous cell carcinoma who underwent resection surgery. We extracted 31 textural indices from preoperative positron emission tomography images. Overall survival (OS) and disease-free survival (DFS) were chosen as the primary outcome variables, and the primary predictor variables were age, sex, primary tumor location, pathological T and N classification, histologic differentiation, resected margin, perineural and lymphovascular invasion, maximum standardized uptake value, and the 14 textural indices selected in the factor analysis. We analyzed OS and DFS using Kaplan-Meier curves, and the differences between survival curves were determined using a log-rank test. The independent prognostic factors were assessed using the Cox-proportional hazards model.

RESULTS

We enrolled 81 patients (median age, 67.3 years; range, 32 to 88 years). The median follow-up duration was 50.1 months (range, 6.3 to 133.7 months). The univariable and multivariable analyses revealed that higher entropy values (≥1.91) were associated with worse OS (hazard ratio, 21.49; 95% confidence interval, 1.36 to 340.71; P = .03) and DFS (hazard ratio, 50.69; 95% confidence interval, 5.23 to 491.18; P = .001).

CONCLUSIONS

This study showed that entropy is a statistically significant prognostic factor of both OS and DFS. Texture analysis using preoperative positron emission tomography images may contribute to risk stratification.

Feasibility of T1 mapping with histogram analysis for the diagnosis and staging of liver fibrosis: Preclinical results

PURPOSE

To compare the diagnostic accuracy of parameters derived from the histogram analysis of precontrast, 10-min hepatobiliary phase (HBP) and 20-min HBP T1 maps for staging liver fibrosis (LF).

METHODS

LF was induced in New Zealand white rabbits by subcutaneous injections of carbon tetrachloride for 4-16 weeks (n = 120), and 20 rabbits injected with saline served as controls. Precontrast, 10-min and 20-min HBP modified Look-Locker inversion recovery (MOLLI) T1 mapping was performed. Histogram analysis of T1 maps was performed, and the mean, median, skewness, kurtosis, entropy, inhomogeneity and 10th/25th/75th/90th percentiles of T1_native, T1_10min and T1_20min were derived. Quantitative histogram parameters were compared. For significant parameters, further receiver operating characteristic (ROC) analyses were performed to evaluate the potential diagnostic performance in differentiating LF stages.

RESULTS

Finally, 17, 20, 21, 21 and 20 rabbits were included for the F0, F1, F2, F3, and F4 pathological grades of fibrosis, respectively. The mean/75th of T1_native, entropy of T1_10min and entropy/mean/median/10th of T1_20min demonstrated a significant good correlation with the LF stage (|r| = 0.543-0.866, all P < 0.05). The 75th of T1_native, entropy_10min, and entropy_20min were the three most reliable imaging markers in reflecting the stage of LF. The area under the ROC curve of entropy_20min was larger than that of entropy_10min (P < 0.05 for LF ≥ F2, ≥F3, and ≥ F4) and the 75th of T1_native (P < 0.05 for LF ≥ F2 and ≥ F3) for staging LF.

CONCLUSION

Magnetic resonance histogram analysis of T1 maps, particularly the entropy derived from 20-min HBP T1 mapping, is promising for predicting the LF stage.

Radiomics and "radi-…omics" in cancer immunotherapy: a guide for clinicians

In recent years the concept of precision medicine has become a popular topic particularly in medical oncology. Besides the identification of new molecular prognostic and predictive biomarkers and the development of new targeted and immunotherapeutic drugs, imaging has started to play a central role in this new era. Terms such as "radiomics", "radiogenomics" or "radi…-omics" are becoming increasingly common in the literature and soon they will represent an integral part of clinical practice. The use of artificial intelligence, imaging and "-omics" data can be used to develop models able to predict, for example, the features of the tumor immune microenvironment through imaging, and to monitor the therapeutic response beyond the standard radiological criteria. The aims of this narrative review are to provide a simplified guide for clinicians to these concepts, and to summarize the existing evidence on radiomics and "radi…-omics" in cancer immunotherapy.

Texture Analysis as a Radiomic Marker for Differentiating Benign From Malignant Adrenal Tumors

OBJECTIVE

The aim of this study was to evaluate the use of texture analysis for differentiation between benign from malignant adrenal lesions on contrast-enhanced abdominal computed tomography (CT).

METHODS

After institutional review board approval, a retrospective analysis was performed, including an electronic search of pathology records for all biopsied adrenal lesions. Patients were included if they also had a contrast-enhanced abdominal CT in the portal venous phase. Computed tomographic images were manually segmented, and texture analysis of the segmented tumors was performed. Texture analysis results of benign and malignant tumors were compared, and areas under the curve (AUCs) were calculated.

RESULTS

One hundred twenty-five patients were included in the analysis. Excellent discriminators of benign from malignant lesions were identified, including entropy and standard deviation. These texture features demonstrated lower values for benign lesions compared with malignant lesions. Entropy values of benign lesions averaged 3.95 using a spatial scaling factor of 4 compared with an average of 5.08 for malignant lesions (P < .0001). Standard deviation values of benign lesions averaged 19.94 on the unfiltered image compared with an average of 34.32 for malignant lesions (P < .0001). Entropy demonstrated AUCs ranging from 0.95 to 0.97 for discriminating tumors, with sensitivities and specificities ranging from 81% to 95% and 88% to 100%, respectively. Standard deviation demonstrated AUCs ranging from 0.91 to 0.94 for discriminating tumors, with sensitivities and specificities ranging from 73% to 93% and 86% to 95%, respectively.

CONCLUSION

Texture analysis offers a noninvasive tool for differentiating benign from malignant adrenal tumors on contrast-enhanced CT images. These results support the further development of texture analysis as a quantitative biomarker for characterizing adrenal tumors.

Differentiation of Endometriomas from Ovarian Hemorrhagic Cysts at Magnetic Resonance: The Role of Texture Analysis

Background and Objectives: To assess ovarian cysts with texture analysis (TA) in magnetic resonance (MRI) images for establishing a differentiation criterion for endometriomas and functional hemorrhagic cysts (HCs) that could potentially outperform their classic MRI diagnostic features. Materials and Methods: Forty-three patients with known ovarian cysts who underwent MRI were retrospectively included (endometriomas, n = 29; HCs, n = 14). TA was performed using dedicated software based on T2-weighted images, by incorporating the whole lesions in a three-dimensional region of interest. The most discriminative texture features were highlighted by three selection methods (Fisher, probability of classification error and average correlation coefficients, and mutual information). The absolute values of these parameters were compared through univariate, multivariate, and receiver operating characteristic analyses. The ability of the two classic diagnostic signs ("T2 shading" and "T2 dark spots") to diagnose endometriomas was assessed by quantifying their sensitivity (Se) and specificity (Sp), following their conventional assessment on T1-and T2-weighted images by two radiologists. Results: The diagnostic power of the one texture parameter that was an independent predictor of endometriomas (entropy, 75% Se and 100% Sp) and of the predictive model composed of all parameters that showed statistically significant results at the univariate analysis (100% Se, 100% Sp) outperformed the ones shown by the classic MRI endometrioma features ("T2 shading", 75.86% Se and 35.71% Sp; "T2 dark spots", 55.17% Se and 64.29% Sp). Conclusion: Whole-lesion MRI TA has the potential to offer a superior discrimination criterion between endometriomas and HCs compared to the classic evaluation of the two lesions' MRI signal behaviors.

Radiomics and Bladder Cancer: Current Status

PURPOSE: To systematically review the current literature and discuss the applications and limitations of radiomics and machine-learning augmented radiomics in the management of bladder cancer. METHODS: Pubmed ®, Scopus ®, and Web of Science ® databases were searched systematically for all full-text English-language articles assessing the impact of Artificial Intelligence OR Radiomics OR Machine Learning AND Bladder Cancer AND (staging OR grading OR prognosis) published up to January 2020. RESULTS: Of the 686 articles that were identified, 13 studies met the criteria for quantitative analysis. Staging, Grading and Tumor Classification, Prognosis, and Therapy Response were discussed in 7, 3, 2 and 7 studies, respectively. Data on cost of implementation were not reported. CT and MRI were the most common imaging approaches. CONCLUSION: Radiomics shows potential in bladder cancer detection, staging, grading, and response to therapy, thereby supporting the physician in personalizing patient management. Extension and validation of this promising technology in large multisite prospective trials is warranted to pave the way for its clinical translation.

Three dimensional texture analysis of noncontrast chest CT in differentiating solitary solid lung squamous cell carcinoma from adenocarcinoma and correlation to immunohistochemical markers

Background

The aim of the study was to investigate 3D texture analysis (3D‐TA) in noncontrast enhanced computed tomography (CT) (NCECT) to differentiate squamous cell carcinoma (SCC) from adenocarcinoma (AC), and the correlation with immunohistochemical markers.

Methods

A total of 70 patients confirmed with SCC (n = 29) and AC (n = 41) were enrolled in this retrospective study. 3D‐TA was utilized to calculate TA parameters of all the tumor lesions based on NCECT images, and all the patients were divided into the training and the test groups. The TA parameters were selected by dimensionality reduction, and the model was established to differentiate SCC from AC according to the training group. The ROC curve was used to evaluate the diagnostic efficiency of the model in both the training and the test groups. Spearman correlation were used to assess the correlation between the selected feature parameters and immunohistochemical markers (P63, P40, and TTF‐1).

Results

Five TA parameters, including volume count, relative deviation, Haralick correlation, gray‐level nonuniformity and run length nonuniformity, were obtained to differentiate SCC from AC by multistep dimensionality reduction. The new model combined with all five TA parameters yielded a high diagnostic performance to differentiate SCC from AC (AUC 0.803) in test group, with a specificity of 89% and a sensitivity of 77%. There was weak correlation between the five texture feature parameters and P63 as well as P40 in all patients (P < 0.05), respectively.

Conclusions

The model including five TA parameters on NECT has a good diagnostic performance in differentiating SCC from AC.

Key points

• Significant findings of the study

The model created by five selected textural feature parameters can differentiate solid SCC from AC without contrast media. The selected five texture feature parameters are correlated to the immunohistochemical markers P63 and P40.

• What this study adds

The textural feature parameters' model can identify SCC from AC without contrast media.

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

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

Texture analysis of routine T2 weighted fat-saturated images can identify head and neck paragangliomas - A pilot study

PURPOSE

To evaluate the role of the first and second-order texture parameters obtained from T2-weighted fat-saturated DIXON images in differentiating paragangliomas from other neck masses, and to develop a statistical model to classify them.

METHOD

We retrospectively evaluated 38 paragangliomas, 18 nerve-sheath tumours and 14 other miscellaneous neck lesions obtained from an IRB approved study conducted between January 2016 and June 2019; using a composite gold standard of histopathology, cytology and DOTANOC PET CT (A total of 70 lesions in 63 patients). Fat-suppressed T2weighted-DIXON axial images were used. First and second-order texture-parameters were calculated from the original and filtered images. Feature selection using F-statistics and collinearity analysis provided 14 texture parameters for further analysis. Mann-Whitney-U test was used to compare between the groups and p-values were adjusted for multiple comparisons. ROC curve analysis was used to obtain optimal cut-offs.

RESULTS

A total of ten texture features were found to be significantly different between paragangliomas and non-paraganglioma lesions. Minimum from the histogram of grey levels was lower in paragangliomas with a cut off of ≤113.462 obtaining 62.9 % sensitivity and 77.27 % specificity in differentiating paragangliomas from non-paragangliomas. Logistic regression model was trained (n-49) using forward feature selection, which when evaluated on the validation set(n-21)- obtained an AUC of 0.855(95 %CI, 0.633 to 0.968) with a positive likelihood ratio of 4.545 (95 %CI, 1.298-15.923) in differentiating paragangliomas from non-paragangliomas.

CONCLUSION

Texture analysis of a routine imaging sequence can identify paragangliomas with high accuracy. Further development of texture analysis would enable better imaging workflow, resource utilisation and imaging cost reductions.

Decoding and Systematization of Medical Imaging Features of Multiple Human Malignancies

Purpose

To summarize the data of previously reported medical imaging features on human malignancies to provide a scientific basis for more credible imaging feature selection for future studies.

Materials and Methods

A search was performed in PubMed from database inception through March 23, 2018, for studies clearly stating the decoding of medical imaging features for malignancy-related objectives and/or hypotheses. The Newcastle-Ottawa scale was used for quality assessment of the included studies. Unsupervised hierarchical clustering was performed on the manually extracted features from each included study to identify the application rules of medical imaging features across human malignancies. CT images of 1000 retrospective patients with non–small cell lung cancer were used to reveal a pattern for the value distribution of complex texture features.

Results

A total of 5026 imaging features of malignancies affecting 20 parts of the human body from 930 original articles were collated and assessed in this study. A meta-feature construct was proposed to facilitate the investigation of details of any high-dimensional complex imaging features of malignancy. A correlation atlas was constructed to clarify the general rules of applying medical imaging features to the analysis of human malignancy. Assessment of this data revealed a pattern of value distributions of the most commonly reported texture features across human malignancies. Furthermore, the significant expression of the gene mutational signature 1B across human cancer was highly consistent with the presence of the run length imaging feature across different human malignancy types.

Conclusion

The results of this study may facilitate more credible imaging feature selection in all oncology tasks across a wide spectrum of human malignancies and help to reduce bias and redundancies in future medical imaging studies.

Keywords: Computer Aided Diagnosis (CAD), Computer Applications-General (Informatics), Evidence Based Medicine, Informatics, Research Design, Statistics, Technology Assessment

Supplemental material is available for this article.

Published under a CC BY 4.0 license.

Prediction of Glioma Grades Using Deep Learning with Wavelet Radiomic Features

Gliomas are the most common primary brain tumors. They are classified into 4 grades (Grade I–II-III–IV) according to the guidelines of the World Health Organization (WHO). The accurate grading of gliomas has clinical significance for planning prognostic treatments, pre-diagnosis, monitoring and administration of chemotherapy. The purpose of this study is to develop a deep learning-based classification method using radiomic features of brain tumor glioma grades with deep neural network (DNN). The classifier was combined with the discrete wavelet transform (DWT) the powerful feature extraction tool. This study primarily focuses on the four main aspects of the radiomic workflow, namely tumor segmentation, feature extraction, analysis, and classification. We evaluated data from 121 patients with brain tumors (Grade II, n = 77; Grade III, n = 44) from The Cancer Imaging Archive, and 744 radiomic features were obtained by applying low sub-band and high sub-band 3D wavelet transform filters to the 3D tumor images. Quantitative values were statistically analyzed with MannWhitney U tests and 126 radiomic features with significant statistical properties were selected in eight different wavelet filters. Classification performances of 3D wavelet transform filter groups were measured using accuracy, sensitivity, F1 score, and specificity values using the deep learning classifier model. The proposed model was highly effective in grading gliomas with 96.15% accuracy, 94.12% precision, 100% recall, 96.97% F1 score, and 98.75% Area under the ROC curve. As a result, deep learning and feature selection techniques with wavelet transform filters can be accurately applied using the proposed method in glioma grade classification.

New boundaries of liver imaging: from morphology to function

From an invisible organ to one of the most explored non-invasively, the liver is, today, one of the cornerstones for current cross-sectional imaging techniques and minimally invasive procedures. After the achievements of US, CT and, most recently, MRI in providing highly accurate morphological and structural information about the organ, a significant scientific development has gained momentum for the last decades, coupling morphology to liver function and contributing far most to what we know today as precision medicine. In fact, dedicated tailor-made investigations are now possible in order to detect and, most of all, quantify physiopathological processes with unprecedented certitude. It is the intention of this review to provide a better insight to the reader of several functional imaging techniques applied to liver imaging. Contrast enhanced imaging, diffusion weighted imaging, elastography, spectral computed tomography and fat and iron assessment techniques are commonly performed clinically. Diffusion kurtosis imaging, magnetic resonance spectroscopy, T1 relaxometry and radiomics remain largely limited to advanced clinical research. Each of them has its own value and place on the diagnostic armamentarium and provide unique qualitative and quantitative information regarding the pathophysiology of diseases, contributing at a large scale to model therapeutic decisions and patient follow-up. Therefore, state-of-the-art liver imaging acts today as a non-invasive surrogate biomarker of many focal and diffuse liver diseases.

New advances in radiomics of gastrointestinal stromal tumors

Gastrointestinal stromal tumors (GISTs) are uncommon neoplasms of the gastrointestinal tract with peculiar clinical, genetic, and imaging characteristics. Preoperative knowledge of risk stratification and mutational status is crucial to guide the appropriate patients’ treatment. Predicting the clinical behavior and biological aggressiveness of GISTs based on conventional computed tomography (CT) and magnetic resonance imaging (MRI) evaluation is challenging, unless the lesions have already metastasized at the time of diagnosis. Radiomics is emerging as a promising tool for the quantification of lesion heterogeneity on radiological images, extracting additional data that cannot be assessed by visual analysis. Radiomics applications have been explored for the differential diagnosis of GISTs from other gastrointestinal neoplasms, risk stratification and prediction of prognosis after surgical resection, and evaluation of mutational status in GISTs. The published researches on GISTs radiomics have obtained excellent performance of derived radiomics models on CT and MRI. However, lack of standardization and differences in study methodology challenge the application of radiomics in clinical practice. The purpose of this review is to describe the new advances of radiomics applied to CT and MRI for the evaluation of gastrointestinal stromal tumors, discuss the potential clinical applications that may impact patients’ management, report limitations of current radiomics studies, and future directions.

Mid-term response assessment in multiple myeloma using a texture analysis approach on dual energy-CT-derived bone marrow images - A proof of principle study

PURPOSE

To identify textural features on dual-energy CT (DECT)-generated virtual non calcium (VNC) bone marrow images in a small group of patients with multiple myeloma undergoing systemic treatment which could potentially help for mid-term response assessment.

METHODS

44 patients (59.1 ± 11.2 yr.) with multiple myeloma who underwent unenhanced whole-body reduced-dose DECT before and after systemic therapy were evaluated. All patients had current hematologic laboratory tests including serum levels of immunoglobulins, albumin, and total proteins. Using DECT post-processing, bone marrow images of the axial skeleton were reconstructed. The vertebral bodies T10-L5 were segmented for quantification of 1st order (n = 18) and 2nd order Gray Level Co-occurrence Matrix (GLCM) textural features (n = 23) based on an open-source radiomics library (Pyradiomics), which were then compared with the hematologic response category to treatment. Five patients underwent only active surveillance at intervals after previous successful therapy.

RESULTS

According to hematologic diagnosis, 29 patients were classified as complete response (CR), 10 as partial response (PR) and 5 as stable disease (SD). We observed a significant drop of the 1st order textural features "10th percentile" (p = 0.009), "median" (p = 0.01), and "minimum" (p < 0.0001) after treatment, whereas the 1st order feature "range" (p = 0.0004) and the 2nd order GLCM feature "difference variance" (p = 0.007) significantly increased in patients experiencing CR. A similar trend, however, without statistical significance, could be observed in patients achieving PR after treatment. 2nd order GLCM feature "difference variance" proved to be a significant discriminator (p = 0.01) between patients with CR and PR (sensitivity 0.93, specificity 0.70) for a cut-off value of -0.28. In patients classified CR, both the mean serum protein and the beta-2 microglobulin decreased after treatment, whereas the serum albumin increased (p < 0.01). The same trend without significance could be observed in patients classified PR.

CONCLUSIONS

Changes in textural features applied on VNC bone marrow images in the pre- and posttreatment settings correlate well with myeloma-specific hematologic parameters and provide complementary information for the assessment of the late effects of treatment on the bone marrow.

Textural Characteristics of Biopsy-proven Metastatic Axillary Nodes on Preoperative Breast MRI in Breast Cancer Patients: A Feasibility Study

OBJECTIVE

To determine the diagnostic accuracy of MRI textural analysis (TA) to differentiate malignant from benign axillary lymph nodes in patients with breast cancer.

METHODS

This was an institutional review board-approved retrospective study of axillary lymph nodes in women with breast cancer that underwent ultrasound-guided biopsy and contrast-enhanced (CE) breast MRI from January 2015 to December 2018. TA of axillary lymph nodes was performed on 3D dynamic CE T1-weighted fat-suppressed, 3D delayed CE T1-weighted fat-suppressed, and T2-weighted fat-suppressed MRI sequences. Quantitative parameters used to measure TA were compared with pathologic diagnoses. Areas under the curve (AUC) were calculated using receiver operating characteristic curve analysis to distinguish between malignant and benign lymph nodes.

RESULTS

Twenty-three biopsy-proven malignant lymph nodes and 24 benign lymph nodes were analyzed. The delayed CE T1-weighted fat-suppressed sequence had the greatest ability to differentiate malignant from benign outcome at all spatial scaling factors, with the highest AUC (0.84-0.93), sensitivity (0.78 [18/23] to 0.87 [20/23]), and specificity (0.76 [18/24] to 0.88 [21/24]). Kurtosis on the 3D delayed CE T1-weighted fat-suppressed sequence was the most prominent TA parameter differentiating malignant from benign lymph nodes (P < 0.0001).

CONCLUSION

This study suggests that MRI TA could be helpful in distinguishing malignant from benign axillary lymph nodes. Kurtosis has the greatest potential on 3D delayed CE T1-weighted fat-suppressed sequences to distinguish malignant and benign lymph nodes.

Radiomics for the Diagnosis and Differentiation of Pancreatic Cystic Lesions

Radiomics, also known as quantitative imaging or texture analysis, involves extracting a large number of features traditionally unmeasured in conventional radiological cross-sectional images and converting them into mathematical models. This review describes this approach and its use in the evaluation of pancreatic cystic lesions (PCLs). This discipline has the potential of more accurately assessing, classifying, risk stratifying, and guiding the management of PCLs. Existing studies have provided important insight into the role of radiomics in managing PCLs. Although these studies are limited by the use of retrospective design, single center data, and small sample sizes, radiomic features in combination with clinical data appear to be superior to the current standard of care in differentiating cyst type and in identifying mucinous PCLs with high-grade dysplasia. Combining radiomic features with other novel endoscopic diagnostics, including cyst fluid molecular analysis and confocal endomicroscopy, can potentially optimize the predictive accuracy of these models. There is a need for multicenter prospective studies to elucidate the role of radiomics in the management of PCLs.

Feasibility of computed tomography texture analysis of hepatic fibrosis using dual-energy spectral detector computed tomography

PURPOSE

To evaluate feasibility of computer tomography texture analysis (CTTA) at different energy level using dual-energy spectral detector CT for liver fibrosis.

MATERIALS AND METHODS

Eighty-seven patients who underwent a spectral CT examination and had a reference standard of liver fibrosis (histopathologic findings, n = 61, or clinical findings for normal, n = 26) were included. Mean gray-level intensity, mean number of positive pixels (MPP), entropy, skewness, and kurtosis using commercially available software (TexRAD) were compared at different energy levels. Optimal CTTA parameter cutoffs to diagnose liver fibrosis were evaluated. CTTA parameters at different energy levels correlated with liver fibrosis. The association of CTTA parameters with energy level was evaluated.

RESULTS

Mean gray-level intensity, skewness, kurtosis, and entropy showed significant differences between patients with and without clinically significant hepatic fibrosis (P < 0.05). Mean gray-level intensity at 50 keV was significantly positively correlated with liver fibrosis (ρ = 0.502, P < 0.001). To diagnose stages F2-F4, entropy and mean gray-level intensity at low keV level showed the largest area under the curve (AUC; 0.79 and 0.79). Estimated marginal means (EMMs) of mean gray-level intensity showed prominent differences at low energy levels.

CONCLUSION

CTTA parameters from different keV levels demonstrated meaningful accuracy for diagnosis of liver fibrosis or clinically significant hepatic fibrosis.

Texture analysis in susceptibility-weighted imaging may be useful to differentiate acute from chronic multiple sclerosis lesions

OBJECTIVES

To evaluate the diagnostic performance of texture analysis (TA) applied on non-contrast-enhanced susceptibility-weighted imaging (SWI) to differentiate acute (enhancing) from chronic (non-enhancing) multiple sclerosis (MS) lesions.

METHODS

We analyzed 175 lesions from 58 patients with relapsing-remitting MS imaged on a 3.0 T MRI scanner and applied TA on T2-w and SWI images to extract texture features. We evaluated the presence or absence of lesion enhancement on T1-w post-contrast images and performed a computational statistical analysis to assess if there was any significant correlation between the texture features and the presence of lesion activity. ROC curves and leave-one-out cross-validation were used to evaluate the performance of individual features and multiparametric models in the identification of active lesions.

RESULTS

Multiple TA features obtained from SWI images showed a significantly different distribution in acute and chronic lesions (AUC, 0.617-0.720). Multiparametric predictive models based on logistic ridge regression and partial least squares regression yielded an AUC of 0.778 and 0.808, respectively. Results from T2-w images did not show any significant predictive ability of neither individual features nor multiparametric models.

CONCLUSIONS

Texture analysis on SWI sequences may be useful to differentiate acute from chronic MS lesions. The good diagnostic performance could help to reduce the need of intravenous contrast agent administration in follow-up MRI studies.

KEY POINTS

• Texture analysis applied on SWI sequences may be useful to differentiate acute from chronic multiple sclerosis lesions • The good diagnostic performance could help to minimize the need of intravenous contrast agent administration in follow-up MRI studies.

MR Imaging Texture Analysis in the Abdomen and Pelvis

Add "which is a" before "distribution"? Texture analysis (TA) is a form of radiomics that refers to quantitative measurements of the histogram, distribution and/or relationship of pixel intensities or gray scales within a region of interest on an image. TA can be applied to MR images of the abdomen and pelvis, with the main strength quantitative analysis of pixel intensities and heterogeneity rather than subjective/qualitative analysis. There are multiple limitations of MRTA. Despite these limitations, there is a growing body of literature supporting MRTA. This review discusses application of MRTA to the abdomen and pelvis.

MRI Texture Analysis for Differentiating Nonfunctional Pancreatic Neuroendocrine Neoplasms From Solid Pseudopapillary Neoplasms of the Pancreas

RATIONALE AND OBJECTIVES

To evaluate the value of texture analysis on preoperative magnetic resonance imaging (MRI) for identifying nonfunctional pancreatic neuroendocrine neoplasms (NF-PNENs) and solid pseudopapillary neoplasms (SPNs).

MATERIALS AND METHODS

This retrospective study included 119 patients who underwent MRI, including T2-weighted imaging with fat-suppression, diffusion-weighted imaging (DWI), apparent diffusion coefficient, precontrast T1-weighted imaging with fat-suppression (T₁WI+fs), and dynamic contrast-enhanced (DCE)-T₁WI+fs. Raw data analysis, principal component analysis, linear discriminant analysis, and nonlinear discriminant analysis (NDA) were used to classify NF-PNENs and SPNs. The results are reported as misclassification rates. The images were simultaneously evaluated by an experienced senior radiologist without knowledge of the pathological results. The misclassification rate of the radiologist was compared to the MaZda (texture analysis software) results. Neural network classifier testing was used for validation. In addition, 30 textures for each MRI sequence were investigated.

RESULTS

The misclassification rate of NDA was lower than that of other analyses. In NDA, DWI obtained the lowest value of 7.92%, but there was no significant difference among the sequences. The misclassification rate of the radiologist (34.65%) was significantly higher than that of NDA for all sequences. The validation results were good in the arterial phase and delayed phase. In the training set, entropy and sum entropy were optimal texture features on DWI and precontrast T₁WI+fs, while the mean and percentile seemed to be the more discriminative features on DCE-T₁WI+fs.

CONCLUSION

Texture analysis can sensitively distinguish between NF-PNENs and SPNs on MRI, and percentile and mean of DCE-T₁WI+fs images were informative for differentiation of neoplasms.

Texture Analysis of Apparent Diffusion Coefficient Maps in Cervical Carcinoma: Correlation with Histopathologic Findings and Prognosis

Purpose

To determine the feasibility of texture analysis of apparent diffusion coefficient (ADC) maps and to assess the performance of texture analysis and ADC to predict histologic grade, parametrial invasion, lymph node metastasis, International Federation of Gynecology and Obstetrics (FIGO) stage, recurrence, and recurrence-free survival (RFS) in patients with cervical carcinoma.

Materials and Methods

This retrospective study included 58 patients with cervical carcinoma who were examined with a 1.5-T MRI system and diffusion-weighted imaging with b values of 0 and 1000 sec/mm². Software with volumes of interest on ADC maps was used to extract 45 texture features, including higher-order texture features. Receiver operating characteristic (ROC) analysis was performed to compare the diagnostic performance of ADC map random forest models and of ADC values. Dunnett test, Spearman rank correlation coefficient, Kaplan-Meier analyses, log-rank test, and Cox proportional hazards regression analyses were also used for statistical analyses.

Results

The ADC map random forest models showed a significantly larger area under the ROC curve (AUC) than the AUC of ADC values for predicting high-grade cervical carcinoma (P = .0036), but not for parametrial invasion, lymph node metastasis, stages III-IV, and recurrence (P = .0602, .3176, .0924, and .5633, respectively). The random forest models predicted that the mean RFS rates were significantly shorter for high-grade cervical carcinomas, parametrial invasion, lymph node metastasis, stages III-IV, and recurrence (P = .0405, < .0001, .0344, .0001, and .0015, respectively); the random forest models for parametrial invasion and stages III-IV were more useful than ADC values (P = .0018) for predicting RFS.

Conclusion

The ADC map random forest models were more useful for noninvasively evaluating histologic grade, parametrial invasion, lymph node metastasis, FIGO stage, and recurrence and for predicting RFS in patients with cervical carcinoma than were ADC values.Keywords: Comparative Studies, Genital/Reproductive, MR-Diffusion Weighted Imaging, MR-Imaging, Neoplasms-Primary, Pathology, Pelvis, Tissue Characterization, UterusSupplemental material is available for this article.© RSNA, 2020See also the commentary by Reinhold and Nougaret in this issue.

MRI-based texture analysis for differentiating pediatric craniofacial rhabdomyosarcoma from infantile hemangioma

OBJECTIVES

To evaluate the diagnostic performance of MRI texture analysis (TA) for differentiation of pediatric craniofacial rhabdomyosarcoma (RMS) from infantile hemangioma (IH).

METHODS

This study included 15 patients with RMS and 42 patients with IH who underwent MRI before an invasive procedure. All patients had a solitary lesion. T2-weighted and fat-suppressed contrast-enhanced T1-weighted axial images were used for TA. Two readers delineated the tumor borders for TA independently and evaluated the qualitative MRI characteristics in consensus. The differences of the texture features' values between the groups were assessed and ROC curves were calculated. Logistic regression analysis was used to analyze the value of TA with and without the combination of the qualitative MRI characteristics. A p value < 0.05 was considered statistically significant.

RESULTS

Thirty-eight texture features were calculated for each tumor. Eighteen features on T2-weighted images and 25 features on contrast-enhanced T1-weighted images were significantly different between the RMSs and IHs. On contrast-enhanced T1-weighted images, the short-zone emphasis (SZE), which was a gray-level zone length matrix (GLZLM) parameter, had the largest area under the curve: 0.899 (sensitivity 93%, specificity 87%). The independent predictor for the RMS among the qualitative MRI characteristics was heterogeneous contrast enhancement (p < 0.001). Using only a GLZLM_SZE value of lower than 0.72 was found to be the best diagnostic parameter in predicting RMS (p < 0.001; 95% CI, 8.770-992.4).

CONCLUSION

MRI-based TA may contribute to differentiate RMS from IH without invasive procedures.

KEY POINTS

• Texture analysis may help to distinguish between rhabdomyosarcoma and infantile hemangioma without invasive procedures. • The gray-level zone length matrix parameters, especially the short-zone emphasis, may be a potential predictor for rhabdomyosarcoma. • Using contrast-enhanced T1-weighted images may be superior to T2-weighted images to differentiate rhabdomyosarcoma from infantile hemangioma in texture analysis.

MRI image analysis methods and applications: an algorithmic perspective using brain tumors as an exemplar

The use of magnetic resonance imaging (MRI) in healthcare and the emergence of radiology as a practice are both relatively new compared with the classical specialties in medicine. Having its naissance in the 1970s and later adoption in the 1980s, the use of MRI has grown exponentially, consequently engendering exciting new areas of research. One such development is the use of computational techniques to analyze MRI images much like the way a radiologist would. With the advent of affordable, powerful computing hardware and parallel developments in computer vision, MRI image analysis has also witnessed unprecedented growth. Due to the interdisciplinary and complex nature of this subfield, it is important to survey the current landscape and examine the current approaches for analysis and trend trends moving forward.

MRI features and texture analysis for the early prediction of therapeutic response to neoadjuvant chemoradiotherapy and tumor recurrence of locally advanced rectal cancer

OBJECTIVES

This study aimed to evaluate the efficiency of imaging features and texture analysis (TA) based on baseline rectal MRI for the early prediction of therapeutic response to neoadjuvant chemoradiotherapy (nCRT) and tumor recurrence in patients with locally advanced rectal cancer (LARC).

METHODS

Consecutive patients with LARC who underwent rectal MRI between January 2014 and December 2015 and surgical resection after completing nCRT were retrospectively enrolled. Imaging features were analyzed, and TA parameters were extracted from the tumor volume of interest (VOI) from baseline rectal MRI. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the optimal TA parameter cutoff values to stratify the patients. Logistic and Cox regression analyses were performed to assess the efficacy of each imaging feature and texture parameter in predicting tumor response and disease-free survival.

RESULTS

In total, 78 consecutive patients were enrolled. In the logistic regression, good treatment response was associated with lower tumor location (OR = 13.284, p = 0.012), low Conv_Min (OR = 0.300, p = 0.013) and high Conv_Std (OR = 3.174, p = 0.016), Shape_Sphericity (OR = 3.170, p = 0.015), and Shape_Compacity (OR = 2.779, p = 0.032). In the Cox regression, a greater risk of tumor recurrence was related to higher cT stage (HR = 5.374, p = 0.044), pelvic side wall lymph node positivity (HR = 2.721, p = 0.013), and gray-level run length matrix_long-run low gray-level emphasis (HR = 2.268, p = 0.046).

CONCLUSIONS

Imaging features and TA based on baseline rectal MRI could be valuable for predicting the treatment response to nCRT for rectal cancer and tumor recurrence.

KEY POINTS

• Imaging features and texture parameters of T2-weighted MR images of rectal cancer can help to predict treatment response and the risk for tumor recurrence. • Tumor location as well as conventional and shape indices of texture features can help to predict treatment response for rectal cancer. • Clinical T stage, positive pelvic side wall lymph nodes, and the high-order texture parameter, GLRLM_LRLGE, can help to predict tumor recurrence for rectal cancer.

Radiomics with artificial intelligence: a practical guide for beginners

Radiomics is a relatively new word for the field of radiology, meaning the extraction of a high number of quantitative features from medical images. Artificial intelligence (AI) is broadly a set of advanced computational algorithms that basically learn the patterns in the data provided to make predictions on unseen data sets. Radiomics can be coupled with AI because of its better capability of handling a massive amount of data compared with the traditional statistical methods. Together, the primary purpose of these fields is to extract and analyze as much and meaningful hidden quantitative data as possible to be used in decision support. Nowadays, both radiomics and AI have been getting attention for their remarkable success in various radiological tasks, which has been met with anxiety by most of the radiologists due to the fear of replacement by intelligent machines. Considering ever-developing advances in computational power and availability of large data sets, the marriage of humans and machines in future clinical practice seems inevitable. Therefore, regardless of their feelings, the radiologists should be familiar with these concepts. Our goal in this paper was three-fold: first, to familiarize radiologists with the radiomics and AI; second, to encourage the radiologists to get involved in these ever-developing fields; and, third, to provide a set of recommendations for good practice in design and assessment of future works.