Multiparametric radiomics improve prediction of lymph node metastasis of rectal cancer compared with conventional radiomics

Diagnostic performance of Node Reporting and Data System (Node-RADS) for assessing mesorectal lymph node in rectal cancer by CT

BACKGROUND

To compare the diagnostic performance of the Node-RADS scoring system and lymph node (LN) size in preoperative LN assessment for rectal cancer (RC), and to investigate whether the selection of size as the primary criterion whereas morphology as the secondary criterion for LNs can be considered the preferred method for clinical assessment.

METHODS

Preoperative CT data of 146 RC patients treated with radical resection surgery were retrospectively analyzed. The Node-RADS score and short-axis diameter of size-prioritized LNs and the morphology-prioritized LNs were obtained. The correlations of Node-RADS score to the pN stage, LNM number and lymph node ratio (LNR) were investigated. The performances on assessing pathological lymph node metastasis were compared between Node-RADS score and short-axis diameter. A nomogram combined the Node-RADS score and clinical features was also evaluated.

RESULTS

Node-RADS score showed significant correlation with pN stage, LNM number and LNR (Node-RADS of size-prioritized LN: r = 0.600, 0.592, and 0.606; Node-RADS of morphology-prioritized LN: r = 0.547, 0.538, and 0.527; Node-RADSmax: r = 0.612, 0.604, and 0.610; all p < 0.001). For size-prioritized LN, Node-RADS achieved an AUC of 0.826, significantly superior to short-axis diameter (0.826 vs. 0.743, p = 0.009). For morphology-prioritized LN, Node-RADS exhibited an AUC of 0.758, slightly better than short-axis diameter (0.758 vs. 0.718, p = 0.098). The Node-RADS score of size-prioritized LN was significantly better than that of morphology-prioritized LN (0.826 vs. 0.758, p = 0.038). The nomogram achieved the best diagnostic performance (AUC = 0.861) than all the other assessment methods (p < 0.05).

CONCLUSIONS

The Node-RADS scoring system outperforms the short-axis diameter in predicting lymph node metastasis in RC. Size-prioritized LN demonstrates superior predictive efficacy compared to morphology-prioritized LN. The nomogram combined the Node-RADS score of size-prioritized LN with clinical features exhibits the best diagnostic performance. Moreover, a clear relationship was demonstrated between the Node-RADS score and the quantity-dependent pathological characteristics of LNM.

Clinical development of MRI-based multi-sequence multi-regional radiomics model to predict lymph node metastasis in rectal cancer

OBJECTIVE

We aim to construct a magnetic resonance imaging (MRI)-based multi-sequence multi-regional radiomics model that will improve the preoperative prediction ability of lymph node metastasis (LNM) in T3 rectal cancer.

METHODS

Multi-sequence MRI data from 190 patients with T3 rectal cancer were retrospectively analyzed, with 94 patients in the LNM group and 96 patients in the non-LNM group. The clinical factors, subjective imaging features, and the radiomic features of tumor and peritumoral mesorectum region of patients were extracted from T2WI and ADC images. Spearman's rank correlation coefficient, Mann-Whitney's U test, and the least absolute shrinkage and selection operator were used for feature selection and dimensionality reduction. Logistic regression was used to construct six models. The predictive performance of each model was evaluated by the receiver operating characteristic curve (ROC). The differences of each model were characterized by area under the curve (AUC) via the DeLong test.

RESULTS

The AUCs of T2WI, ADC single-sequence radiomics model and multi-sequence radiomics model were 0.73, 0.75, and 0.78, respectively. The multi-sequence multi-regional radiomics model with improved performance was created by combining the radiomics characteristics of the peritumoral mesorectum region with the multi-sequence radiomics model (AUC, 0.87; p < 0.01). The AUC of the clinical model was 0.68, and the MRI-clinical composite evaluation model was obtained by incorporating the clinical data with the multi-sequence multi-regional radiomics features, with an AUC of 0.89.

CONCLUSION

The MRI-based multi-sequence multi-regional radiomics model significantly improved the prediction ability of LNM for T3 rectal cancer and could be applied to guide surgical decision-making in patients with T3 rectal cancer.

T2WI-based texture analysis predicts preoperative lymph node metastasis of rectal cancer

BACKGROUND

To prospectively develop and validate the T2WI texture analysis model based on a node-by-node comparison for improving the diagnostic accuracy of lymph node metastasis (LNM) in rectal cancer.

METHODS

A total of 381 histopathologically confirmed lymph nodes (LNs) were collected. LNs texture features were extracted from MRI-T2WI. Spearman's rank correlation coefficient and the least absolute shrinkage and selection operator were used for feature selection to construct the LN rad-score. Then the clinical risk factors and LN texture features were combined to establish combined predictive model. Model performance was assessed by the area under the receiver operating characteristic (ROC) curve (AUC). Decision curve analysis (DCA) and nomogram were used to evaluate the clinical application of the model.

RESULTS

A total of 107 texture features were extracted from LN-MRI images. After selection and dimensionality reduction, the radiomics prediction model consisting of 8 texture features showed well-predictive performance in the training and validation cohorts (AUC, 0.676; 95% CI 0.582-0.771) (AUC, 0.774; 95% CI 0.648-0.899). A clinical-radiomics prediction model with the best performance was created by combining clinical and radiomics features, 0.818 (95% CI 0.742-0.893) for the training and 0.922 (95% CI 0.863-0.980) for the validation cohort. The LN Rad-score in clinical-radiomics nomogram obtained the highest classification contribution and was well calibrated. DCA demonstrated the superiority of the clinical-radiomics model.

CONCLUSION

The lymph node T2WI-based texture features can help to improve the preoperative prediction of LNM.

A radiomics model based on T2WI and clinical indexes for prediction of lateral lymph node metastasis in rectal cancer

OBJECTIVE

The aim of this study was to explore the clinical value of a radiomics prediction model based on T2-weighted imaging (T2WI) and clinical indexes in predicting lateral lymph node (LLN) metastasis in rectal cancer patients.

METHODS

This was a retrospective analysis of 106 rectal cancer patients who had undergone LLN dissection. The clinical risk factors for LLN metastasis were selected by multivariable logistic regression analysis of the clinical indicators of the patients. The LLN radiomics features were extracted from the pelvic T2WI of the patients. The least absolute shrinkage and selection operator algorithm and backward stepwise regression method were adopted for feature selection. Three LLN metastasis prediction models were established through logistic regression analysis based on the clinical risk factors and radiomics features. Model performance was assessed in terms of discriminability and decision curve analysis in the training, verification and test sets.

RESULTS

The model based on the combined T2WI radiomics features and clinical risk factors demonstrated the highest accuracy, surpassing the models based solely on either T2WI radiomics features or clinical risk factors. Specifically, the model achieved an AUC value of 0.836 in the test set. Decision curve analysis revealed that this model had the greatest clinical utility for the vast majority of the threshold probability range from 0.4 to 1.0.

CONCLUSION

Combining T2WI radiomics features with clinical risk factors holds promise for the noninvasive assessment of the biological characteristics of the LLNs in rectal cancer, potentially aiding in therapeutic decision-making and optimizing patient outcomes.

Development and validation of a combined nomogram for predicting perineural invasion status in rectal cancer via computed tomography-based radiomics

AIM

This study aimed to create and validate a clinic-radiomics nomogram based on computed tomography (CT) imaging for predicting preoperative perineural invasion (PNI) of rectal cancer (RC).

MATERIAL AND METHODS

This study enrolled 303 patients with RC who were divided into training (n = 242) and test datasets (n = 61) in an 8:2 ratio with all their clinical outcomes. A total of 3,296 radiomic features were extracted from CT images. Five machine learning (ML) models (logistic regression (LR)/K-nearest neighbor (KNN)/multilayer perceptron (MLP)/support vector machine (SVM)/light gradient boosting machine (LightGBM)) were developed using radiomic features derived from the arterial and venous phase images, and the model with the best diagnostic performance was selected. By combining the radiomics and clinical signatures, a fused nomogram model was constructed.

RESULTS

After using the Mann-Whitney U-test and least absolute shrinkage and selection operator (LASSO) to remove redundant features, the MLP model proved to be the most efficient among the five ML models. The fusion nomogram based on MLP prediction probability further improves the ability to predict the PNI status. The area under the curve (AUC) of the training and test sets was 0.883 and 0.889, respectively, which were higher than those of the clinical (training set, AUC = 0.710; test set, AUC = 0.762) and radiomic models (training set, AUC = 0.840; test set, AUC = 0.834).

CONCLUSIONS

The clinical-radiomics combined nomogram model based on enhanced CT images efficiently predicted the PNI status of patients with RC.

Diagnostic value of a radiomics model based on CT and MRI for prediction of lateral lymph node metastasis of rectal cancer

This study aimed to develop a radiomics model for predicting lateral lymph node (LLN) metastasis in rectal cancer patients using MR-T2WI and CT images, and assess its clinical value. This prospective study included rectal cancer patients with complete MR-T2WI and portal enhanced CT images who underwent LLN dissection at Tianjin Union Medical Center between June 2017 and November 2022. Primary lesions and LLN were segmented using 3D slicer. Radiomics features were extracted from the region of interest using pyradiomics in Python. Least absolute shrinkage and selection operator algorithm and backward stepwise regression were employed for feature selection. Three LLN metastasis radiomics prediction models were established via multivariable logistic regression analysis. The performance of the model was evaluated using receiver operating characteristic curve analysis, and the area under the curve (AUC), sensitivity, specificity were calculated for the training, validation, and test sets. A nomogram was constructed for visualization, and decision curve analysis (DCA) was performed to evaluate clinical value. We included 94 eligible patients in the analysis. For each patient, we extracted a total of 1344 radiomics features. The CT combined with MR-T2WI model had the highest AUC for all sets compared to CT and MR-T2WI models. AUC values for the CT combined with MR-T2WI model in the training, validation, and test sets were 0.957, 0.901, and 0.936, respectively. DCA revealed high prediction value for the combined MR-T2WI and CT model. A radiomics model based on CT and MR-T2WI data effectively predicted LLN metastasis in rectal cancer patients preoperatively.

Progress in the diagnosis of lymph node metastasis in rectal cancer: a review

Historically, the chief focus of lymph node metastasis research has been molecular and clinical studies of a few essential pathways and genes. Recent years have seen a rapid accumulation of massive omics and imaging data catalyzed by the rapid development of advanced technologies. This rapid increase in data has driven improvements in the accuracy of diagnosis of lymph node metastasis, and its analysis further demands new methods and the opportunity to provide novel insights for basic research. In fact, the combination of omics data, imaging data, clinical medicine, and diagnostic methods has led to notable advances in our basic understanding and transformation of lymph node metastases in rectal cancer. Higher levels of integration will require a concerted effort among data scientists and clinicians. Herein, we review the current state and future challenges to advance the diagnosis of lymph node metastases in rectal cancer.

Artificial intelligence - based ultrasound elastography for disease evaluation - a narrative review

Ultrasound elastography (USE) provides complementary information of tissue stiffness and elasticity to conventional ultrasound imaging. It is noninvasive and free of radiation, and has become a valuable tool to improve diagnostic performance with conventional ultrasound imaging. However, the diagnostic accuracy will be reduced due to high operator-dependence and intra- and inter-observer variability in visual observations of radiologists. Artificial intelligence (AI) has great potential to perform automatic medical image analysis tasks to provide a more objective, accurate and intelligent diagnosis. More recently, the enhanced diagnostic performance of AI applied to USE have been demonstrated for various disease evaluations. This review provides an overview of the basic concepts of USE and AI techniques for clinical radiologists and then introduces the applications of AI in USE imaging that focus on the following anatomical sites: liver, breast, thyroid and other organs for lesion detection and segmentation, machine learning (ML) - assisted classification and prognosis prediction. In addition, the existing challenges and future trends of AI in USE are also discussed.

MRI-based multiregional radiomics for predicting lymph nodes status and prognosis in patients with resectable rectal cancer

Purpose

To establish and evaluate multiregional T2-weighted imaging (T2WI)-based clinical-radiomics model for predicting lymph node metastasis (LNM) and prognosis in patients with resectable rectal cancer.

Methods

A total of 346 patients with pathologically confirmed rectal cancer from two hospitals between January 2019 and December 2021 were prospectively enrolled. Intra- and peritumoral features were extracted separately, and least absolute shrinkage and selection operator regression was applied for feature selection. Radiomics signatures were built using the selected features from different regions. The clinical-radiomic nomogram was developed by combining the intratumoral and peritumoral radiomics signatures score (radscore) and the most predictive clinical parameters. The diagnostic performances of the nomogram and clinical model were evaluated using the area under the receiver operating characteristic curve (AUC). The prognostic model for 3-year recurrence-free survival (RFS) was constructed using univariate and multivariate Cox analysis.

Results

The intratumoral radscore (radscore 1) included four features, the peritumoral radscore (radscore 2) included five features, and the combined intratumoral and peritumoural radscore (radscore 3) included ten features. The AUCs for radscore 3 were higher than that of radscore 1 in training cohort (0.77 vs. 0.71, P=0.182) and internal validation cohort (0.76 vs. 0.64, P=0.041). The AUCs for radscore 3 were higher than that of radscore 2 in training cohort (0.77 vs. 0.74, P=0.215) and internal validation cohort (0.76 vs. 0.68, P=0.083). A clinical-radiomic nomogram showed a higher AUC compared with the clinical model in training cohort (0.84 vs. 0.67, P<0.001) and internal validation cohort (0.78 vs. 0.64, P=0.038) but not in external validation (0.72 vs. 0.76, P=0.164). Multivariate Cox analysis showed MRI-reported extramural vascular invasion (EMVI) (HR=1.099, 95%CI: 0.462-2.616; P=0.031) and clinical-radiomic nomogram-based LNM (HR=2.232, 95%CI:1.238-7.439; P=0.017) were independent risk factors for assessing 3-year RFS. Combined clinical-radiomic nomogram based LNM and MRI-reported EMVI showed good performance in training cohort (AUC=0.748), internal validation cohort (AUC=0.706) and external validation (AUC=0.688) for predicting 3-year RFS.

Conclusion

A clinical-radiomics nomogram exhibits good performance for predicting preoperative LNM. Combined clinical-radiomic nomogram based LNM and MRI-reported EMVI showed clinical potential for assessing 3-year RFS.

Endorectal ultrasound radiomics in locally advanced rectal cancer patients: despeckling and radiotherapy response prediction using machine learning

PURPOSE

The current study aimed to evaluate the association of endorectal ultrasound (EUS) radiomics features at different denoising filters based on machine learning algorithms and to predict radiotherapy response in locally advanced rectal cancer (LARC) patients.

METHODS

The EUS images of forty-three LARC patients, as a predictive biomarker for predicting the treatment response of neoadjuvant chemoradiotherapy (NCRT), were investigated. For despeckling, the EUS images were preprocessed by traditional filters (bilateral, wiener, lee, frost, median, and wavelet filters). The rectal tumors were delineated by two readers separately, and radiomics features were extracted. The least absolute shrinkage and selection operator were used for feature selection. Classifiers including logistic regression (LR), K-nearest neighbor (KNN), support vector machine (SVM), random forest, naive Bayes, and decision tree were trained using stratified fivefold cross-validation for model development. The area under the curve (AUC) of the receiver operating characteristic curve followed by accuracy, precision, sensitivity, and specificity were obtained for model performance assessment.

RESULTS

The wavelet filter had the best results with means of AUC: 0.83, accuracy: 77.41%, precision: 82.15%, and sensitivity: 79.41%. LR and SVM by having AUC: 0.71 and 0.76; accuracy: 70.0% and 71.5%; precision: 75.0% and 73.0%; sensitivity: 69.8% and 80.2%; and specificity: 70.0% and 60.9% had the highest model's performance, respectively.

CONCLUSION

This study demonstrated that the EUS-based radiomics model could serve as pretreatment biomarkers in predicting pathologic features of rectal cancer. The wavelet filter and machine learning methods (LR and SVM) had good results on the EUS images of rectal cancer.

Prospective clinical research of radiomics and deep learning in oncology: A translational review

Radiomics and deep learning (DL) hold transformative promise and substantial and significant advances in oncology; however, most methods have been tested in retrospective or simulated settings. There is considerable interest in the biomarker validation, clinical utility, and methodological robustness of these studies and their deployment in real-world settings. This review summarizes the characteristics of studies, the level of prospective validation, and the overview of research on different clinical endpoints. The discussion of methodological robustness shows the potential for independent external replication of prospectively reported results. These in-depth analyses further describe the barriers limiting the translation of radiomics and DL into primary care options and provide specific recommendations regarding clinical deployment. Finally, we propose solutions for integrating novel approaches into the treatment environment to unravel the critical process of translating AI models into the clinical routine and explore strategies to improve personalized medicine.

Radiomics for differentiating tumor deposits from lymph node metastasis in rectal cancer

BACKGROUND

Tumor deposits (TDs) are not equivalent to lymph node (LN) metastasis (LNM) but have become independent adverse prognostic factors in patients with rectal cancer (RC). Although preoperatively differentiating TDs and LNMs is helpful in designing individualized treatment strategies and achieving improved prognoses, it is a challenging task.

AIM

To establish a computed tomography (CT)-based radiomics model for preoperatively differentiating TDs from LNM in patients with RC.

METHODS

This study retrospectively enrolled 219 patients with RC [TDs⁺LNM^- (n = 89); LNM⁺ TDs^- (n = 115); TDs⁺LNM⁺ (n = 15)] from a single center between September 2016 and September 2021. Single-positive patients (i.e., TDs⁺LNM^- and LNM⁺TDs^-) were classified into the training (n = 163) and validation (n = 41) sets. We extracted numerous features from the enhanced CT (region 1: The main tumor; region 2: The largest peritumoral nodule). After deleting redundant features, three feature selection methods and three machine learning methods were used to select the best-performing classifier as the radiomics model (Rad-score). After validating Rad-score, its performance was further evaluated in the field of diagnosing double-positive patients (i.e., TDs⁺LNM⁺) by outlining all peritumoral nodules with diameter (short-axis) > 3 mm.

RESULTS

Rad-score 1 (radiomics signature of the main tumor) had an area under the curve (AUC) of 0.768 on the training dataset and 0.700 on the validation dataset. Rad-score 2 (radiomics signature of the largest peritumoral nodule) had a higher AUC (training set: 0.940; validation set: 0.918) than Rad-score 1. Clinical factors, including age, gender, location of RC, tumor markers, and radiological features of the largest peritumoral nodule, were excluded by logistic regression. Thus, the combined model was comprised of Rad-scores of 1 and 2. Considering that the combined model had similar AUCs with Rad-score 2 (P = 0.134 in the training set and 0.594 in the validation set), Rad-score 2 was used as the final model. For the diagnosis of double-positive patients in the mixed group [TDs⁺LNM⁺ (n = 15); single-positive (n = 15)], Rad-score 2 demonstrated moderate performance (sensitivity, 73.3%; specificity, 66.6%; and accuracy, 70.0%).

CONCLUSION

Radiomics analysis based on the largest peritumoral nodule can be helpful in preoperatively differentiating between TDs and LNM.

A Dual-Energy CT Radiomics of the Regional Largest Short-Axis Lymph Node Can Improve the Prediction of Lymph Node Metastasis in Patients With Rectal Cancer

Objective

To explore the value of dual-energy computed tomography (DECT) radiomics of the regional largest short-axis lymph nodes for evaluating lymph node metastasis in patients with rectal cancer.

Materials and Methods

One hundred forty-one patients with rectal cancer (58 in LNM+ group, 83 in LNM- group) who underwent preoperative total abdominal DECT were divided into a training group and testing group (7:3 ratio). After post-processing DECT venous phase images, 120kVp-like images and iodine (water) images were obtained. The highest-risk lymph nodes were identified, and their long-axis and short-axis diameter and DECT quantitative parameters were measured manually by two experienced radiologists who were blind to the postoperative pathological results. Four DECT parameters were analyzed: arterial phase (AP) normalized iodine concentration, AP normalized effective atomic number, the venous phase (VP) normalized iodine concentration, and the venous phase normalized effective atomic number. The carcinoembryonic antigen (CEA) levels were recorded one week before surgery. Radiomics features of the largest lymph nodes were extracted, standardized, and reduced before modeling. Radomics signatures of 120kVp-like images (Rad-signature120kVp) and iodine map (Rad-signatureImap) were built based on Logistic Regression via Least Absolute Shrinkage and Selection Operator (LASSO).

Results

Eight hundred thirty-three features were extracted from 120kVp-like and iodine images, respectively. In testing group, the radiomics features based on 120kVp-like images showed the best diagnostic performance (AUC=0.922) compared to other predictors [CT morphological indicators (short-axis diameter (AUC=0.779, IDI=0.262) and long-axis diameter alone (AUC=0.714, IDI=0.329)), CEA alone (AUC=0.540, IDI=0.414), and normalized DECT parameters alone (AUC=0.504-0.718, IDI=0.290-0.476)](P&lt;0.05 in Delong test). Contrary, DECT iodine map-based radiomic signatures showed similar performance in predicting lymph node metastasis (AUC=0.866). The decision curve showed that the 120kVp-like-based radiomics signature has the highest net income.

Conclusion

Predictive model based on DECT and the largest short-axis diameter lymph nodes has the highest diagnostic value in predicting lymph node metastasis in patients with rectal cancer.

Lymph Nodes Evaluation in Rectal Cancer: Where Do We Stand and Future Perspective

The assessment of nodal involvement in patients with rectal cancer (RC) is fundamental in disease management. Magnetic Resonance Imaging (MRI) is routinely used for local and nodal staging of RC by using morphological criteria. The actual dimensional and morphological criteria for nodal assessment present several limitations in terms of sensitivity and specificity. For these reasons, several different techniques, such as Diffusion Weighted Imaging (DWI), Intravoxel Incoherent Motion (IVIM), Diffusion Kurtosis Imaging (DKI), and Dynamic Contrast Enhancement (DCE) in MRI have been introduced but still not fully validated. Positron Emission Tomography (PET)/CT plays a pivotal role in the assessment of LNs; more recently PET/MRI has been introduced. The advantages and limitations of these imaging modalities will be provided in this narrative review. The second part of the review includes experimental techniques, such as iron-oxide particles (SPIO), and dual-energy CT (DECT). Radiomics analysis is an active field of research, and the evidence about LNs in RC will be discussed. The review also discusses the different recommendations between the European and North American guidelines for the evaluation of LNs in RC, from anatomical considerations to structured reporting.

Establishment and validation of a nomogram for predicting potential lateral pelvic lymph node metastasis in low rectal cancer

BACKGROUND

Identifying lateral pelvic lymph node (LPN) metastasis in low rectal cancer is crucial before treatment. Several risk factors and prediction models for LPN metastasis have been reported. However, there is no useful tool to accurately predict LPN metastasis. Therefore, we aimed to construct a nomogram for predicting LPN metastasis in rectal cancer.

METHODS

We analyzed the risk factors for potential LPN metastasis by logistic regression analysis in 705 patients who underwent primary resection of low rectal cancer. We included patients at 49 institutes of the Japan Society of Laparoscopic Colorectal Surgery between June 2010 and February 2012. Clinicopathological factors and magnetic resonance imaging findings were evaluated. The nomogram performance was assessed using the c-index and calibration plots, and the nomogram was validated using an external cohort.

RESULTS

In the univariable logistic regression analysis, age, sex, carcinoembryonic antigen, tumor location, clinical T stage, tumor size, circumferential resection margin (CRM), extramural vascular invasion (EMVI), and the short and long axes of LPN and perirectal lymph node (PRLN) were nominated as risk factors for potential LPN metastasis. We identified a combination of the short axis of LPN, tumor location, EMVI, and short axis of PRLN as optimal for predicting potential LPN metastasis and developed a nomogram using these factors. This model had a c-index of 0.74 and was moderately calibrated and well-validated.

CONCLUSIONS

This is the first study to construct a well-validated nomogram for predicting potential LPN metastasis in rectal cancer, and its performance was high.

A Nomogram of Combining IVIM-DWI and MRI Radiomics From the Primary Lesion of Rectal Adenocarcinoma to Assess Nonenlarged Lymph Node Metastasis Preoperatively

BACKGROUND

Lymph node (LN) staging plays an important role in treatment decision-making. Current problem is that preoperative detection of LN involvement is always highly challenging for radiologists.

PURPOSE

To explore the value of the nomogram model combining intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) and radiomics features from the primary lesion of rectal adenocarcinoma in assessing the non-enlarged lymph node metastasis (N-LNM) preoperatively.

STUDY TYPE

Retrospective.

POPULATION

A total of 126 patients (43% female) comprising a training group (n = 87) and a validation group (n = 39) with pathologically confirmed rectal adenocarcinoma.

FIELD STRENGTH/SEQUENCE

A 3.0 Tesla (T); T₂ -weighted imaging (T₂ WI) with fast spin-echo (FSE) sequence; IVIM-DWI spin-echo echo-planar imaging sequence.

ASSESSMENT

Based on pathological analysis of the surgical specimen, patients were classified into negative LN (LN-) and positive LN (LN+) groups. Apparent diffusion coefficient (ADC), diffusion coefficient (D), pseudo diffusion coefficient (D*) and microvascular volume fraction (f) values of primary lesion of rectal adenocarcinoma were measured. Three-dimensional (3D) radiomics features were measured on T₂ WI and IVIM-DWI. A nomogram model including IVIM-DWI and radiomics features was developed.

STATISTICAL TESTS

General_univariate_analysis and multivariate logistic regression were used for radiomics features selection. The performance of the nomogram was assessed by the receiver operating characteristic (ROC) curve, calibration, and decision curve analysis (DCA).

RESULTS

The LN+ group had a significantly lower D* value ([13.20 ± 13.66 vs. 23.25 ± 18.71] × 10^-3  mm² /sec) and a higher f value (0.43 ± 0.12 vs. 0.34 ± 0.10) than the LN- group in the training cohort. The nomogram model combined D*, f, and radiomics features had a better evaluated performance (AUC = 0.864) than any other model in the training cohort.

DATE CONCLUSION

The nomogram model including IVIM-DWI and MRI radiomics features in the primary lesion of rectal adenocarcinoma was associated with the N-LNM.

EVIDENCE LEVEL

4 TECHNICAL EFFICACY: Stage 2.

Stiffness Could be a Predictor of AJCC Prognostic Stage Groups in Preoperative Invasive Ductal Carcinoma

OBJECTIVES

This study aimed to evaluate the stiffness of 2-dimensional (2D) shear wave elastography (SWE) in preoperatively predicting the prognostic stage groups of invasive ductal carcinoma (IDC).

METHODS

Eighty-six newly diagnosed lesions on 83 patients with IDCs were analyzed. All parameters from conventional ultrasound and stiffness to virtual touch tissue imaging and quantification were collected, and mean shear wave velocity (SWVmean) was calculated. Data on maximum diameter, estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), histologic grading system and Tumor Node Metastasis (TNM) stages were collected. The levels of maximum shear wave velocity (SWVmax), minimum shear wave velocity (SWVmin) and SWVmean were compared. In receiver operating characteristic (ROC) curves analysis, the diagnostic efficacy was found in area under the curve (AUC). Parallel mode was used to improve the predictive value of sensitivity.

RESULTS

The median stiffness of SWVmax and SWVmean for IDCs were 9.38 and 6.32 m/s for late stage (stages II, III, IV) and 6.39 m/s and 4.72 m/s for early stage (stage I) of the prognostic stage groups, respectively. The median stiffness values in the late stage were significantly higher than those in the early stage (P = .003, P = .005). The optimal cutoff stiffness of SWVmax and SWVmean were 8.62 and 6.13 m/s, respectively. In ROC curves analysis, the AUC for SWVmax was 0.742, and it showed a better diagnostic value than SWVmean (0.725). In predictive diagnosis, the sensitivity for SWVmax and SWVmean were both 62.50%. The parallel mode improved the prediction power of sensitivity to 68.75%.

CONCLUSIONS

Preoperative SWV level may serve as a promising prognostic imaging indicator for breast IDCs.

Development of a Nomogram Combining Clinical Risk Factors and Dual-Energy Spectral CT Parameters for the Preoperative Prediction of Lymph Node Metastasis in Patients With Colorectal Cancer

Objective

This study aimed to develop a dual-energy spectral computed tomography (DESCT) nomogram that incorporated both clinical factors and DESCT parameters for individual preoperative prediction of lymph node metastasis (LNM) in patients with colorectal cancer (CRC).

Material and Methods

We retrospectively reviewed 167 pathologically confirmed patients with CRC who underwent enhanced DESCT preoperatively, and these patients were categorized into training (n = 117) and validation cohorts (n = 50). The monochromatic CT value, iodine concentration value (IC), and effective atomic number (Eff-Z) of the primary tumors were measured independently in the arterial phase (AP) and venous phase (VP) by two radiologists. DESCT parameters together with clinical factors were input into the prediction model for predicting LNM in patients with CRC. Logistic regression analyses were performed to screen for significant predictors of LNM, and these predictors were presented as an easy-to-use nomogram. The receiver operating characteristic curve and decision curve analysis (DCA) were used to evaluate the clinical usefulness of the nomogram.

Results

The logistic regression analysis showed that carcinoembryonic antigen, carbohydrate antigen 199, pericolorectal fat invasion, ICAP, ICVP, and Eff-ZVP were independent predictors in the predictive model. Based on these predictors, a quantitative nomogram was developed to predict individual LNM probability. The area under the curve (AUC) values of the nomogram were 0.876 in the training cohort and 0.852 in the validation cohort, respectively. DCA showed that our nomogram has outstanding clinical utility.

Conclusions

This study presents a clinical nomogram that incorporates clinical factors and DESCT parameters and can potentially be used as a clinical tool for individual preoperative prediction of LNM in patients with CRC.

Artificial intelligence for pre-operative lymph node staging in colorectal cancer: a systematic review and meta-analysis

Background

Artificial intelligence (AI) is increasingly being used in medical imaging analysis. We aimed to evaluate the diagnostic accuracy of AI models used for detection of lymph node metastasis on pre-operative staging imaging for colorectal cancer.

Methods

A systematic review was conducted according to PRISMA guidelines using a literature search of PubMed (MEDLINE), EMBASE, IEEE Xplore and the Cochrane Library for studies published from January 2010 to October 2020. Studies reporting on the accuracy of radiomics models and/or deep learning for the detection of lymph node metastasis in colorectal cancer by CT/MRI were included. Conference abstracts and studies reporting accuracy of image segmentation rather than nodal classification were excluded. The quality of the studies was assessed using a modified questionnaire of the QUADAS-2 criteria. Characteristics and diagnostic measures from each study were extracted. Pooling of area under the receiver operating characteristic curve (AUROC) was calculated in a meta-analysis.

Results

Seventeen eligible studies were identified for inclusion in the systematic review, of which 12 used radiomics models and five used deep learning models. High risk of bias was found in two studies and there was significant heterogeneity among radiomics papers (73.0%). In rectal cancer, there was a per-patient AUROC of 0.808 (0.739–0.876) and 0.917 (0.882–0.952) for radiomics and deep learning models, respectively. Both models performed better than the radiologists who had an AUROC of 0.688 (0.603 to 0.772). Similarly in colorectal cancer, radiomics models with a per-patient AUROC of 0.727 (0.633–0.821) outperformed the radiologist who had an AUROC of 0.676 (0.627–0.725).

Conclusion

AI models have the potential to predict lymph node metastasis more accurately in rectal and colorectal cancer, however, radiomics studies are heterogeneous and deep learning studies are scarce.

Trial registration

PROSPERO CRD42020218004.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-021-08773-w.

Prediction of lymph node metastasis in rectal cancer: comparison between shear-wave elastography based ultrasomics and MRI

PURPOSE

We aimed to explore the diagnostic efficiency of shear-wave elastography (SWE) ultrasomics in the preoperative prediction of lymph node (LN) metastasis in rectal cancer.

METHODS

This study included 87 patients with pathologically confirmed rectal cancer, with data gathered from August 2017 to August 2018. A total of 1044 ultrasomics features of rectal tumor were collected with AK software from the SWE examinations. The least absolute shrinkage and selection operator (LASSO) regression model was used for feature selection and building a SWE ultrasomics signature. The diagnostic performance was evaluated with an area under the receiver operating characteristic curve (AUC) analysis. Then, the diagnostic performance of the SWE ultrasomics signature was compared with magnetic resonance imaging (MRI).

RESULTS

Of the 87 patients, 40 (46.0%) had LN metastasis. Thirteen ultrasomics features of rectal tumor were selected as the most significant features. The SWE ultrasomics signature correlated with LN metastasis (p < 0.001). Patients with LN metastasis had higher signature than patients without LN metastasis. In terms of diagnostic performance, SWE ultrasomics signature was significantly superior to MRI (AUC, 0.883 vs. 0.760, p = 0.034). The diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of SWE ultrasomics signature were 82.8%, 87.5%, 78.8%, 77.8%, and 88.1%, respectively, while those of MRI were 75.9%, 77.5%, 74.5%, 72.1%, and 79.6%, respectively.

CONCLUSION

SWE ultrasomics is a more accurate predictive method for identifying LN metastasis preoperatively than MRI. Thus, SWE ultrasomics might be used to better guide preoperative individual therapies for patients with rectal cancer.

Preoperative prediction of perineural invasion with multi-modality radiomics in rectal cancer

Perineural invasion (PNI) as a grossly underreported independent risk predictor in rectal cancer is hard to identify preoperatively. We aim to predict PNI status in rectal cancer using multi-modality radiomics. In total, 396 radiomics features were extracted from T2-weighted images (T2WIs), diffusion-weighted images (DWIs), and portal venous phase of contrast-enhanced CT (CE-CT) respectively of 94 consecutive patients with histologically confirmed rectal cancer. T2WI score, DWI score, and CT score were calculated via the radiomics features selection and optimization. Discrimination, calibration, and clinical benefit ability were used to evaluate the performance of the radiomics scores in both training and testing datasets. CT score and T2WI score were independent risk predictors [CT score, OR (95% CI) = 4.218 (1.070–16.620); T2WI score, OR (95% CI) = 105.721 (3.091–3615.790)]. The concise score which combined CT score and T2WI score, showed the best performance [training dataset, AUC (95% CI) = 0.906 (0.833–0.979); testing dataset, AUC (95% CI) = 0.884 (0.761–1.000)] and good calibration (P > 0.05 in the Hosmer–Lemeshow test for the training and testing datasets). Decision curve analysis showed that the multi-modality radiomics nomogram had a higher clinical net benefit. The multi-modality radiomics score could be used to preoperatively assess PNI status in rectal cancer.

Preoperative volumetric synthetic magnetic resonance imaging of the primary tumor for a more accurate prediction of lymph node metastasis in rectal cancer

Background

An accurate assessment of lymph node (LN) status in patients with rectal cancer is important for treatment planning and an essential factor for predicting local recurrence and overall survival. In this study, we explored the potential value of histogram parameters of synthetic magnetic resonance imaging (SyMRI) in predicting LN metastasis in rectal cancer and compared their predictive performance with traditional morphological characteristics and chemical shift effect (CSE).

Methods

A total of 70 patients with pathologically proven rectal adenocarcinoma who received direct surgical resection were enrolled in this prospective study. Preoperative rectal MRI, including SyMRI, were performed, and morphological characteristics and CSE of LN were assessed. Histogram parameters were extracted on a T1 map, T2 map, and proton density (PD) map, including mean, variance, maximum, minimum, 10th percentile, median, 90th percentile, energy, kurtosis, entropy, and skewness. Receiver operating characteristic (ROC) curves were used to explore their predictive performance for assessing LN status.

Results

Significant differences in the energy of the T1, T2, and PD maps were observed between LN-negative and LN-positive groups [all P<0.001; the area under the ROC curve (AUC) was 0.838, 0.858, and 0.823, respectively]. The maximum and kurtosis of the T2 map, maximum, and variance of PD map could also predict LN metastasis with moderate diagnostic power (P=0.032, 0.045, 0.016, and 0.047, respectively). Energy of the T1 map [odds ratio (OR) =1.683, 95% confidence interval (CI): 1.207-2.346, P=0.002] and extramural venous invasion on MRI (mrEMVI) (OR =10.853, 95% CI: 2.339-50.364, P=0.002) were significant predictors of LN metastasis. Moreover, the T1 map energy significantly improved the predictive performance compared to morphological features and CSE (P=0.0002 and 0.0485).

Conclusions

The histogram parameters derived from SyMRI of the primary tumor were associated with LN metastasis in rectal cancer and could significantly improve the predictive performance compared with morphological features and CSE.

An MRI-based multi-objective radiomics model predicts lymph node status in patients with rectal cancer

PURPOSE

To apply a multi-objective radiomics model based on pre-operative magnetic resonance imaging (MRI) for improving diagnostic accuracy of LN metastasis in rectal cancer patients.

METHODS

This study consisted of 91 patients diagnosed with rectal cancer from April 2018 to March 2019. All patients underwent rectal MRI before surgery without any other treatment. Clinical data, subjective radiologist assessments, and radiomic features of LNs were obtained. A total of 1409 radiomic features were extracted from T2WI LN images. Multi-objective optimization with the iterative multi-objective immune algorithm (IMIA) was used to select radiomic features to build prediction models. Predictive performances of radiomic, radiologist, and combined radiomic and radiologist models were assessed for accuracy by receiver operating characteristics (ROC) curves.

RESULTS

For the radiologist analysis, heterogeneity was the only significant independent predictor of LN status. The sensitivity, specificity, and accuracy of the subjective radiologist analysis were 72.09%, 73.81%, and 78.12%, respectively. The sensitivity, specificity, and accuracy of the solitary radiomic model consisting of 10 features were 89.81%, 82.57%, and 87.77%, respectively. The sensitivity, specificity, and accuracy of the combined model, which consisted of 12 radiomic and radiologist features, were 92.23%, 84.69%, and 89.88%, respectively. The combined model had the best prediction performance with an AUC of 0.94.

CONCLUSIONS

The multi-objective radiomics model based on T2WI images was very useful in predicting pre-operative LN status in rectal cancer patients.

MRI Based Radiomics Approach With Deep Learning for Prediction of Vessel Invasion in Early-Stage Cervical Cancer

This article aims to build deep learning-based radiomic methods in differentiating vessel invasion from non-vessel invasion in cervical cancer with multi-parametric MRI data. A set of 1,070 dynamic T1 contrast-enhanced (DCE-T1) and 986 T2 weighted imaging (T2WI) MRI images from 167 early-stage cervical cancer patients (January 2014 - August 2018) were used to train and validate deep learning models. Predictive performances were evaluated using receiver operating characteristic (ROC) curve and confusion matrix analysis, with the DCE-T1 showing more discriminative results than T2WI MRI. By adopting an attention ensemble learning strategy that integrates both MRI sequences, the highest average area was obtained under the ROC curve (AUC) of 0.911 (Sensitivity = 0.881 and Specificity = 0.752). The superior performances in this article, when compared to existing radiomic methods, indicate that a wealth of deep learning-based radiomics could be developed to aid radiologists in preoperatively predicting vessel invasion in cervical cancer patients.

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

Purpose

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

Materials and Methods

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

Results

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

Conclusion

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

Multiregional-Based Magnetic Resonance Imaging Radiomics Combined With Clinical Data Improves Efficacy in Predicting Lymph Node Metastasis of Rectal Cancer

Objective

To develop and validate a multiregional-based magnetic resonance imaging (MRI) radiomics model and combine it with clinical data for individual preoperative prediction of lymph node (LN) metastasis in rectal cancer patients.

Methods

186 rectal adenocarcinoma patients from our retrospective study cohort were randomly selected as the training (n = 123) and testing cohorts (n = 63). Spearman’s rank correlation coefficient and the least absolute shrinkage and selection operator were used for feature selection and dimensionality reduction. Five support vector machine (SVM) classification models were built using selected clinical and semantic variables, single-regional radiomics features, multiregional radiomics features, and combinations, for predicting LN metastasis in rectal cancer. The performance of the five SVM models was evaluated via the area under the receiver operator characteristic curve (AUC), accuracy, sensitivity, and specificity in the testing cohort. Differences in the AUCs among the five models were compared using DeLong’s test.

Results

The clinical, single-regional radiomics and multiregional radiomics models showed moderate predictive performance and diagnostic accuracy in predicting LN metastasis with an AUC of 0.725, 0.702, and 0.736, respectively. A model with improved performance was created by combining clinical data with single-regional radiomics features (AUC = 0.827, (95% CI, 0.711–0.911), P = 0.016). Incorporating clinical data with multiregional radiomics features also improved the performance (AUC = 0.832 (95% CI, 0.717–0.915), P = 0.015).

Conclusion

Multiregional-based MRI radiomics combined with clinical data can improve efficacy in predicting LN metastasis and could be a useful tool to guide surgical decision-making in patients with rectal cancer.

Artificial intelligence for the diagnosis of lymph node metastases in patients with abdominopelvic malignancy: A systematic review and meta-analysis

PURPOSE

Accurate clinical diagnosis of lymph node metastases is of paramount importance in the treatment of patients with abdominopelvic malignancy. This review assesses the diagnostic performance of deep learning algorithms and radiomics models for lymph node metastases in abdominopelvic malignancies.

METHODOLOGY

Embase (PubMed, MEDLINE), Science Direct and IEEE Xplore databases were searched to identify eligible studies published between January 2009 and March 2019. Studies that reported on the accuracy of deep learning algorithms or radiomics models for abdominopelvic malignancy by CT or MRI were selected. Study characteristics and diagnostic measures were extracted. Estimates were pooled using random-effects meta-analysis. Evaluation of risk of bias was performed using the QUADAS-2 tool.

RESULTS

In total, 498 potentially eligible studies were identified, of which 21 were included and 17 offered enough information for a quantitative analysis. Studies were heterogeneous and substantial risk of bias was found in 18 studies. Almost all studies employed radiomics models (n = 20). The single published deep-learning model out-performed radiomics models with a higher AUROC (0.912 vs 0.895), but both radiomics and deep-learning models outperformed the radiologist's interpretation in isolation (0.774). Pooled results for radiomics nomograms amongst tumour subtypes demonstrated the highest AUC 0.895 (95 %CI, 0.810-0.980) for urological malignancy, and the lowest AUC 0.798 (95 %CI, 0.744-0.852) for colorectal malignancy.

CONCLUSION

Radiomics models improve the diagnostic accuracy of lymph node staging for abdominopelvic malignancies in comparison with radiologist's assessment. Deep learning models may further improve on this, but data remain limited.

Radiomics for the Prediction of Treatment Outcome and Survival in Patients With Colorectal Cancer: A Systematic Review

Prediction of outcome in patients with colorectal cancer (CRC) is challenging as a result of lack of a robust biomarker and heterogeneity between and within tumors. The aim of this review was to assess the current possibilities and limitations of radiomics (on computed tomography [CT], magnetic resonance imaging [MRI], and positron emission tomography [PET]) for the prediction of treatment outcome and long-term outcome in CRC. Medline/PubMed was searched up to August 2020 for studies that used radiomics for the prediction of response to treatment and survival in patients with CRC (based on pretreatment imaging). The Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool and Radiomics Quality Score (RQS) were used for quality assessment. A total of 76 studies met the inclusion criteria and were included for further analysis. Radiomics analyses were performed on MRI in 41 studies, on CT in 30 studies, and on ¹⁸F-FDG-PET/CT in 10 studies. Heterogeneous results were reported regarding radiomics methods and included features. High-quality studies (n = 13), consisting mainly of MRI-based radiomics to predict response in rectal cancer, were able to predict response with good performance. Radiomics literature in CRC is highly heterogeneous, but it nonetheless holds promise for the prediction of outcome. The most evidence is available for MRI-based radiomics in rectal cancer. Future radiomics research in CRC should focus on independent validation of existing models rather than on developing new models.

Prediction Model Combining Clinical and MR Data for Diagnosis of Lymph Node Metastasis in Patients With Rectal Cancer

BACKGROUND

Determining the status of lymph node (LN) metastasis in rectal cancer patients preoperatively is crucial for the treatment option. However, the diagnostic accuracy of current imaging methods is low.

PURPOSE

To develop and test a model for predicting metastatic LNs of rectal cancer patients based on clinical data and MR images to improve the diagnosis of metastatic LNs.

STUDY TYPE

Retrospective.

SUBJECTS

In all, 341 patients with histologically confirmed rectal cancer were divided into one training set (120 cases) and three validation sets (69, 103, 49 cases).

FIELD STRENGTH/SEQUENCE

3.0T, axial and sagittal T₂ -weighted turbo spin echo and diffusion-weighted imaging (b = 0 s/mm² , 800 s/mm² ) ASSESSMENT: In the training dataset, univariate logistic regression was used to identify the clinical factors (age, gender, and tumor markers) and MR data that correlated with LN metastasis. Then we developed a prediction model with these factors by multiple logistic regression analysis. The accuracy of the model was verified using three validation sets and compared with the traditional MRI method.

STATISTICAL TESTS

Univariate and multivariate logistic regression. The area under the curve (AUC) value was used to quantify the diagnostic accuracy of the model.

RESULTS

Eight factors (CEA, CA199, ADCmean, mriT stage, mriN stage, CRM, EMVI, and differentiation degree) were significantly associated with LN metastasis in rectal cancer patients (P<0.1). In the training set (120) and the three validation sets (69, 103, 49), the AUC values of the model were much higher than the diagnosis by MR alone (training set, 0.902 vs. 0.580; first validation set, 0.789 vs. 0.743; second validation set, 0.774 vs. 0.573; third validation set, 0.761 vs. 0.524).

DATA CONCLUSION

For the diagnosis of metastatic LNs in rectal cancer patients, our proposed logistic regression model, combining clinical and MR data, demonstrated higher diagnostic efficiency than MRI alone.

LEVEL OF EVIDENCE

4 TECHNICAL EFFICACY STAGE: 2.

Biplane transrectal ultrasonography plus ultrasonic elastosonography and contrast-enhanced ultrasonography in T staging of rectal cancer

Background

The aim of this study is to assess biplane transrectal ultrasonography (TRUS) plus ultrasonic elastosonography (UE) and contrast-enhanced ultrasonography (CEUS) in T staging of rectal cancer.

Methods

Between March 2016 and January 2019, 66 rectal cancer patients who completed biplane TRUS plus UE and CEUS for preoperative workup and were treated by primary total mesorectal excision (TME) were retrospectively analyzed.

Results

The accuracy of TRUS plus UE and CEUS in all T staging of rectal cancer was 69.7%. The highest accuracy was achieved in the T3 stage (87.5%), while it was 71.4 and 50.0% in the T1 and T2 stage, respectively. The mean sizes of uT1-T2 lesions and uT3-T4 lesions were 30.0 ± 10.6 mm (range, 10.0–55.0) and 40.2 ± 11.2 mm (range, 14.0–57.0), respectively (p < 0.001). According to the receiver operating characteristic (ROC) curve to predict pT stages (pT1,2 vs. pT3), the optimal cut-off value of lesions in greatest dimension was 28.5 mm by TRUS with areas under the curve (AUC) of 0.769, and the optimal cut-off values of peak systolic velocity (PSV) and resistive index (RI) were 18.8 cm/sec and 0.645, respectively. The AUCs of PSV and RI were 0.588 and 0.555, respectively.

Conclusions

Diagnostic accuracy of TRUS plus UE and CEUS in T staging of rectal cancer does not reach the excellent published study results, especially for patients with early rectal cancer. Tumor sizes, PSV and RI are useful additions for TRUS in T staging of rectal cancer.

A Novel Multimodal Radiomics Model for Preoperative Prediction of Lymphovascular Invasion in Rectal Cancer

Objective: To explore a new predictive model of lymphatic vascular infiltration (LVI) in rectal cancer based on magnetic resonance (MR) and computed tomography (CT).

Methods: A retrospective study was conducted on 94 patients with histologically confirmed rectal cancer, they were randomly divided into training cohort (n = 65) and validation cohort (n = 29). All patients underwent MR and CT examination within 2 weeks before treatment. On each slice of the tumor, we delineated the volume of interest on T2-weighted imaging, diffusion weighted imaging, and enhanced CT images, respectively. A total of 1,188 radiological features were extracted from each patient. Then, we used the student t-test or Mann–Whitney U-test, Spearman's rank correlation and least absolute shrinkage and selection operator (LASSO) algorithm to select the strongest features to establish a single and multimodal logic model for predicting LVI. Receiver operating characteristic (ROC) curves and calibration curves were plotted to determine how well they explored LVI prediction performance in the training and validation cohorts.

Results: An optimal multi-mode radiology nomogram for LVI estimation was established, which had significant predictive power in training (AUC, 0.884; 95% CI, 0.803–0.964) and validation (AUC, 0.876; 95% CI, 0.721–1.000). Calibration curve and decision curve analysis showed that the multimodal radiomics model provides greater clinical benefits.

Conclusion: Multimodal (MR/CT) radiomics models can serve as an effective visual prognostic tool for predicting LVI in rectal cancer. It demonstrated great potential of preoperative prediction to improve treatment decisions.

CT-based radiomics for preoperative prediction of early recurrent hepatocellular carcinoma: technical reproducibility of acquisition and scanners

PURPOSE

To test the technical reproducibility of acquisition and scanners of CT image-based radiomics model for early recurrent hepatocellular carcinoma (HCC).

METHODS

We included primary HCC patient undergone curative therapies, using early recurrence as endpoint. Four datasets were constructed: 109 images from hospital #1 for training (set 1: 1-mm image slice thickness), 47 images from hospital #1 for internal validation (sets 2 and 3: 1-mm and 10-mm image slice thicknesses, respectively), and 47 images from hospital #2 for external validation (set 4: vastly different from training dataset). A radiomics model was constructed. Radiomics technical reproducibility was measured by overfitting and calibration deviation in external validation dataset. The influence of slice thickness on reproducibility was evaluated in two internal validation datasets.

RESULTS

Compared with set 1, the model in set 2 indicated favorable prediction efficiency (the area under the curve 0.79 vs. 0.80, P = 0.47) and good calibration (unreliability statistic U: P = 0.33). However, in set 4, significant overfitting (0.63 vs. 0.80, P < 0.01) and calibration deviation (U: P < 0.01) were observed. Similar poor performance was also observed in set 3 (0.56 vs. 0.80, P = 0.02; U: P < 0.01).

CONCLUSIONS

CT-based radiomics has poor reproducibility between centers. Image heterogeneity, such as slice thickness, can be a significant influencing factor.

Development and assessment of an individualized nomogram to predict colorectal cancer liver metastases

Background

This article aims to develop and assess the radiomics paradigm for predicting colorectal cancer liver metastasis (CRLM) from the primary tumor.

Methods

This retrospective study included 100 patients from the First Hospital of Jilin University from June 2017 to December 2017. The 100 patients comprised 50 patients with and 50 without CRLM. The maximum-level enhanced computed tomography (CT) image of primary cancer in the portal venous phase of each patient was selected as the original image data. To automatically implement radiomics-related paradigms, we developed a toolkit called Radiomics Intelligent Analysis Toolkit (RIAT).

Results

With RIAT, the model based on logistic regression (LR) using both the radiomics and clinical information signatures showed the maximum net benefit. The area under the curve (AUC) value was 0.90±0.02 (sensitivity =0.85±0.02, specificity =0.79±0.04) for the training set, 0.86±0.11 (sensitivity =0.85±0.09, specificity =0.75±0.19) for the verification set, 0.906 (95% CI, 0.840-0.971; sensitivity =0.81, specificity =0.84) for the cross-validation set, and 0.899 (95% CI, 0.761-1.000; sensitivity =0.78, specificity =0.91) for the test set.

Conclusions

The radiomics nomogram-based LR with clinical risk and radiomics features allows for a more accurate classification of CRLM using CT images with RIAT.

Preoperative prediction of tumour deposits in rectal cancer by an artificial neural network-based US radiomics model

OBJECTIVE

To develop a machine learning-based ultrasound (US) radiomics model for predicting tumour deposits (TDs) preoperatively.

METHODS

From December 2015 to December 2017, 127 patients with rectal cancer were prospectively enrolled and divided into training and validation sets. Endorectal ultrasound (ERUS) and shear-wave elastography (SWE) examinations were conducted for each patient. A total of 4176 US radiomics features were extracted for each patient. After the reduction and selection of US radiomics features , a predictive model using an artificial neural network (ANN) was constructed in the training set. Furthermore, two models (one incorporating clinical information and one based on MRI radiomics) were developed. These models were validated by assessing their diagnostic performance and comparing the areas under the curve (AUCs) in the validation set.

RESULTS

The training and validation sets included 29 (33.3%) and 11 (27.5%) patients with TDs, respectively. A US radiomics ANN model was constructed. The model for predicting TDs showed an accuracy of 75.0% in the validation cohort. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and AUC were 72.7%, 75.9%, 53.3%, 88.0% and 0.743, respectively. For the model incorporating clinical information, the AUC improved to 0.795. Although the AUC of the US radiomics model was improved compared with that of the MRI radiomics model (0.916 vs. 0.872) in the 90 patients with both ultrasound and MRI data (which included both the training and validation sets), the difference was nonsignificant (p = 0.384).

CONCLUSIONS

US radiomics may be a potential model to accurately predict TDs before therapy.

KEY POINTS

• We prospectively developed an artificial neural network model for predicting tumour deposits based on US radiomics that had an accuracy of 75.0%. • The area under the curve of the US radiomics model was improved than that of the MRI radiomics model (0.916 vs. 0.872), but the difference was not significant (p = 0.384). • The US radiomics-based model may potentially predict TDs accurately before therapy, but this model needs further validation with larger samples.

Application of radiomics signature captured from pretreatment thoracic CT to predict brain metastases in stage III/IV ALK-positive non-small cell lung cancer patients

Background

The purpose of this study is to develop a radiomics approach to predict brain metastasis (BM) for stage III/IV ALK-positive non-small cell lung cancer (NSCLC) patients.

Methods

Patients with ALK-positive III/IV NSCLC from 2014 to 2017 were enrolled retrospectively. Their pretreatment thoracic CT images were collected, and the gross tumor volume (GTV) was defined by two experienced radiation oncologists. An in-house feature extraction code-set was performed based on MATLAB 2015b (Mathworks, Natick, MA, USA) in patients' CT images to extract features. Patients were randomly divided into training set and test set (4:1) by using createDataPartition function in caret package. A test-retest in RIDER NSCLC dataset was performed to identify stable radiomics features. LASSO Cox regression and a leave-one-out cross-validation were conducted to identify optimal features for the logistic regression model to evaluate the predictive value of radiomics feature(s) for BM. Furthermore, extended validation for the radiomics feature(s) and Cox regression analyses which combined radiomics feature(s) and treatment elements were implemented to predict the risk of BM during follow-up.

Results

In total, 132 patients were included, among which 27 patients had pretreatment BM. The median follow-up time was 11.8 (range, 0.1-65.2) months. In the training set, one radiomics feature (W_GLCM_LH_Correlation) showed discrimination ability of BM (P value =0.014, AUC =0.687, 95% CI: 0.551-0.824, specificity =83.5%, sensitivity =57.1%). It also exhibited reposeful performance in the test set (AUC =0.642, 95% CI: 0.501-0.783, specificity =60.0%, sensitivity =83.3%). Those 105 patients without pretreatment BM were divided into stage III (n=57) and stage IV (n=48) groups. The radiomics feature (W_GLCM_LH_Correlation) had moderate performance to predict BM during/after treatment in separate groups (stage III: AUC =0.682, 95% CI: 0.537-0.826, specificity =64.4%, sensitivity =75.0%; stage IV: AUC =0.653, 95% CI: 0.503-0.804, specificity =70.4%, sensitivity =75.0%). Meanwhile, stage III patients could be divided into low risk and high risk groups for BM during surveillance according to Cox regression analysis (log-rank P value =0.021).

Conclusions

We identified one wavelet texture feature derived from pretreatment thoracic CT that presented potential in predicting BM in stage III/IV ALK-positive NSCLC patients. This could be beneficial to risk stratification for such patients. Further investigation is necessary to include expanded sample size investigation and external multicenter validation.

Novel imaging techniques of rectal cancer: what do radiomics and radiogenomics have to offer? A literature review

INTRODUCTION

As computational capabilities have advanced, radiologists and their collaborators have looked for novel ways to analyze diagnostic images. This has resulted in the development of radiomics and radiogenomics as new fields in medical imaging. Radiomics and radiogenomics may change the practice of medicine, particularly for patients with colorectal cancer. Radiomics corresponds to the extraction and analysis of numerous quantitative imaging features from conventional imaging modalities in correlation with several endpoints, including the prediction of pathology, genomics, therapeutic response, and clinical outcome. In radiogenomics, qualitative and/or quantitative imaging features are extracted and correlated with genetic profiles of the imaged tissue. Thus far, several studies have evaluated the use of radiomics and radiogenomics in patients with colorectal cancer; however, there are challenges to be overcome before its routine implementation including challenges related to sample size, model design and interpretability, and the lack of robust multicenter validation set.

MATERIAL AND METHODS

In this article, we will review the concepts of radiomics and radiogenomics and their potential applications in rectal cancer.

CONCLUSION

Radiologists should be aware of the basic concepts, benefits, pitfalls, and limitations of new radiomic and radiogenomics techniques to achieve a balanced interpretation of the results.

CT radiomics and PET radiomics: ready for clinical implementation?

INTRODUCTION

Today, rapid technical and clinical developments result in an increasing number of treatment options for oncological diseases. Thus, decision support systems are needed to offer the right treatment to the right patient. Imaging biomarkers hold great promise in patient-individual treatment guidance. Routinely performed for diagnosis and staging, imaging datasets are expected to hold more information than used in the clinical practice. Radiomics describes the extraction of a large number of meaningful quantitative features from medical images, such as computed tomography (CT) and positron emission tomography (PET). Due to the non-invasive nature and ability to capture 3D image-based heterogeneity, radiomic features are potential surrogate markers of the cancer phenotype. Several radiomic studies are published per day, owing to encouraging results of many radiomics-based patient outcome models. Despite this comparably large number of studies, radiomics is mainly studied in proof of principle concept. Hence, a translation of radiomics from a hot topic research field into an essential clinical decision-making tool is lacking, but of high clinical interest.

EVIDENCE ACQUISITION

Herein, we present a literature review addressing the clinical evidence of CT and PET radiomics. An extensive literature review was conducted in PubMed, including papers on robustness and clinical applications.

EVIDENCE SYNTHESIS

We summarize image-modality related influences on the robustness of radiomic features and provide an overview of clinical evidence reported in the literature. Today, more evidence has been provided for CT imaging, however, PET imaging offers the promise of direct imaging of biological processes and functions. We provide a summary of future research directions, which needs to be addressed in order to successfully introduce radiomics into clinical medicine. In comparison to CT, more focus should be directed towards harmonization of PET acquisition and reconstruction protocols, which is important for transferable modelling.

CONCLUSIONS

Both CT and PET radiomics are promising pre-treatment and intra-treatment biomarkers for outcome prediction. Most studies are performed in retrospective setting, however their validation in prospective data collections is ongoing.

A nomogram for determining the disease-specific survival in Ewing sarcoma: a population study

Background

We aimed to develop and validate a nomogram for predicting the disease-specific survival of Ewing sarcoma (ES) patients.

Methods

The Surveillance, Epidemiology, and End Results (SEER) program database was used to identify ES from 1990 to 2015, in which the data was extracted from 18 registries in the US. Multivariate analysis performed using Cox proportional hazards regression models was performed on the training set to identify independent prognostic factors and construct a nomogram for the prediction of the 3-, 5-, and 10-year survival rates of patients with ES. The predictive values were compared by using concordance indexes (C-indexes), calibration plots, integrated discrimination improvement (IDI), net reclassification improvement (NRI), and decision curve analysis (DCA).

Results

A total of 2,643 patients were identified. After multivariate Cox regression, a nomogram was established based on a new model containing the predictive variables of age, race, extent of disease, tumor size, and therapy of surgery. The new model provided better C-indexes (0.684 and 0.704 in the training and validation cohorts, respectively) than the model without therapy of surgery (0.661 and 0.668 in the training and validation cohorts, respectively). The good discrimination and calibration of the nomogram were demonstrated for both the training and validation cohorts. NRI and IDI were also improved. Finally, DCA demonstrated that the nomogram was clinically useful.

Conclusion

We developed a reliable nomogram for determining the prognosis and treatment outcomes of patients with ES in the US. However, the proposed nomogram still requires external data verification in future applications, especially for regions outside the US.

Development and validation of a nomogram containing the prognostic determinants of chondrosarcoma based on the Surveillance, Epidemiology, and End Results database

BACKGROUND

We aimed to develop and validate a reliable nomogram for predicting the disease-specific survival (DSS) of chondrosarcoma patients.

METHODS

The Surveillance, Epidemiology, and End Results (SEER) database was queried from 2004 to 2015 to identify cases of histologically confirmed chondrosarcoma. Multivariate Cox regression analysis was performed to identify independent prognostic factors and construct a nomogram for predicting the 3- and 5-year DSS rates. Predictive values were compared between the new model and the American Joint Committee on Cancer (AJCC) staging system using concordance indexes (C-indexes), calibration plots, integrated discrimination improvement (IDI), net reclassification improvement (NRI), and decision curve analysis (DCA).

RESULTS

Multivariate Cox regression identified 1180 patients, who were used to establish a nomogram based on a new model containing the predictive variables of age, socioeconomic status, tumor size, surgery status, chemotherapy status, and AJCC staging. In the nomogram, age at diagnosis is the factor with the highest risk, followed by AJCC stage IV and tumor size > 100 mm. Both the C-index and the calibration plots demonstrated the good performance of the nomogram. Moreover, both NRI and IDI were improved compared to the AJCC staging system, and also DCA demonstrated that the nomogram is clinically useful.

CONCLUSION

We have developed a reliable nomogram for determining the prognosis and treatment outcomes of chondrosarcoma patients that is superior to the traditional AJCC staging system.

Development and validation of a nomogram for osteosarcoma-specific survival: A population-based study

This study aimed to establish a comprehensive prognostic system for osteosarcoma based on a large population database with high quality.The Surveillance, Epidemiology, and End Results (SEER) Program database was used to identify patients with osteosarcoma from 1973 to 2015. Multivariate analysis was performed to screen statistically significant variables. A nomogram was constructed by R software to predict the 3-, 5- and 10-year survival rates. Predictive abilities were compared by C-indexes, calibration plots, integrated discrimination improvement (IDI), net reclassification improvement (NRI), as well as decision curve analysis (DCA).In total, 4505 osteosarcoma patients were identified. They were divided into training (70%, n = 3153) and validating (30%, n = 1352) groups. Multivariate analyses identified independent predictors. Subsequently, the nomogram system of a new model was established, which comprised 7 variables as age, sex, site, decade of diagnosis (DOD), extent of disease (EOD), tumor size and patients undergoing tri-modality therapy (surgery, radiotherapy and chemotherapy). It provided better C-indexes than the model without therapies (0.727, 0.712 vs 0.705, 0.668) in the 2 cohort, respectively. As well, the new model had good performances in the calibration plots. Moreover, both IDI and NRI improved for 3-, 5- and 10-year follow-up of C-indexes. Finally, DCA demonstrated that the nomogram of new model was clinically meaningful.We developed a reliable nomogram for prognostic determinants and treatment outcome analysis of osteosarcoma, thus helping better choose medical examinations and optimize therapeutic regimen under the cooperation among oncologists and surgeons.