2D and 3D CT Radiomics Features Prognostic Performance Comparison in Non-Small Cell Lung Cancer

Integrating radiomics and gene expression by mapping on the image with improved DeepInsight for clear cell renal cell carcinoma

BACKGROUND

Radiomics analysis extracts high-dimensional features from medical images, which are used to predict outcomes in machine learning (ML). Recently, deep-learning methods have become applicable to image data converted from nonimage samples.

PURPOSE

This study conducted a comparative analysis of outcome-prediction performance using radiomics with a conventional ML approach and deep-learning (DL) approach utilising DeepInsight. Furthermore, we enhance the DeepInsight model by integrating radiomics features with gene expression data. This integration aims to improve predictive power and provide a more comprehensive understanding of ccRCC, ultimately contributing to more personalized and effective treatment strategies.

METHODS

A total of 142 patients with clear cell renal cell carcinoma who underwent surgery were divided into training and test datasets. Radiomics features were extracted in the entire tumour region from CT images. The two-year disease-free survival was predicted using ML and DL. ML was used for selective features after LASSO regression. ML algorithms were employed for classification, including the support vector machine, k-nearest neighbour, and neural network classifiers. For DL, radiomics features and gene-expression data were converted into image data with DeepInsight, and classification tasks were performed with DL techniques such as AlexNet, SqueezeNet, and InceptionNet.

RESULTS

For ML, 17 prognosis-related radiomic features were selected from the LASSO regression. The ML accuracy was 76.5 %, 71.4 %, and 78.1 % for the support vector machine, k-nearest neighbour, and neural network models, respectively. For DL, the accuracies were 76.7 %, 83.1 %, and 85.4 % for AlexNet, SqueezeNet, and InceptionNet, respectively. Furthermore, the integrated DeepInsight models exhibited the highest accuracy of 90.9 %.

CONCLUSION

The proposed DL approach utilising DeepInsight demonstrated a significant improvement in outcome-prediction performance compared with the conventional ML approach. Furthermore, the integration of DL with radiomics features and gene-expression data effectively captures the relationship between biological information and image data, rendering it a promising tool for enhancing outcome-prediction capabilities.

Topological radiogenomics based on persistent lifetime images for identification of epidermal growth factor receptor mutation in patients with non-small cell lung tumors

We hypothesized that persistent lifetime (PLT) images could represent tumor imaging traits, locations, and persistent contrasts of topological components (connected and hole components) corresponding to gene mutations such as epidermal growth factor receptor (EGFR) mutant signs. We aimed to develop a topological radiogenomic approach using PLT images to identify EGFR mutation-positive patients with non-small cell lung cancer (NSCLC). The PLT image was newly proposed to visualize the locations and persistent contrasts of the topological components for a sequence of binary images with consecutive thresholding of an original computed tomography (CT) image. This study employed 226 NSCLC patients (94 mutant and 132 wildtype patients) with pretreatment contrast-enhanced CT images obtained from four datasets from different countries for training and testing prediction models. Two-dimensional (2D) and three-dimensional (3D) PLT images were assumed to characterize specific imaging traits (e.g., air bronchogram sign, cavitation, and ground glass nodule) of EGFR-mutant tumors. Seven types of machine learning classification models were constructed to predict EGFR mutations with significant features selected from 2D-PLT, 3D-PLT, and conventional radiogenomic features. Among the means and standard deviations of the test areas under the receiver operating characteristic curves (AUCs) of all radiogenomic approaches in a four-fold cross-validation test, the 2D-PLT features showed the highest AUC with the lowest standard deviation of 0.927 ± 0.08. The best radiogenomic approaches with the highest AUC were the random forest model trained with the Betti number (BN) map features (AUC = 0.984) in the internal test and the adapting boosting model trained with the BN map features (AUC = 0.717) in the external test. PLT features can be used as radiogenomic imaging biomarkers for the identification of EGFR mutation status in patients with NSCLC.

A Comprehensive Model Outperformed the Single Radiomics Model in Noninvasively Predicting the HER2 Status in Patients with Breast Cancer

RATIONALE AND OBJECTIVES

This study aimed to develop predictive models based on conventional magnetic resonance imaging (cMRI) and radiomics features for predicting human epidermal growth factor receptor 2 (HER2) status of breast cancer (BC) and compare their performance.

MATERIALS AND METHODS

A total of 287 patients with invasive BC in our hospital were retrospectively analyzed. All patients underwent preoperative breast MRI consisting of fat-suppressed T2-weighted imaging, axial dynamic contrast-enhanced MRI, and diffusion-weighted imaging sequences. From these sequences, radiomics features were derived. Three distinct models were established utilizing cMRI features, radiomics features, and a comprehensive model that amalgamated both. The predictive capabilities of these models were assessed using the receiver operating characteristic curve analysis. The comparative performance was then determined through the DeLong test and net reclassification improvement (NRI).

RESULTS

In a randomized split, the 287 patients with BC were allotted to either training (234; 46 HER2-zero, 107 HER2-low, 81 HER2-positive) or test (53; 8 HER2-zero, 27 HER2-low, 18 HER2-positive) at an 8:2 ratio. The mean area under the curve (AUCs) for cMRI, radiomics, and comprehensive models predicting HER2 status were 0.705, 0.819, and 0.859 in training set and 0.639, 0.797, and 0.842 in test set, respectively. DeLong's test indicated that the combined model's AUC surpassed the radiomics model significantly (p < 0.05). NRI analysis verified superiority of the combined model over the radiomics for BC HER2 prediction (NRI 25.0) in the test set.

CONCLUSION

The comprehensive model based on the combination of cMRI and radiomics features outperformed the single radiomics model in noninvasively predicting the three-tiered HER2 status in patients with BC.

Machine Learning Based on Digital Mammography to Reduce the Need for Invasive Biopsies of Benign Calcifications Classified in BI-RADS Category 4

This study aims to develop a machine learning model applied on digital mammograms to reduce unnecessary invasive biopsies for suspicious calcifications classified as BI-RADS category 4. This study retrospectively analyzed data from 372 female patients with pathologically confirmed BI-RADS category 4 mammographic calcifications. Patients from the First Affiliated Hospital of Bengbu Medical College (n = 275) were divided chronologically into a training and internal validation set. An external validation set (n = 97) was recruited from Tongde Hospital of Zhejiang Province. We first segmented calcifications using nnUnet, and then built a radiomics model and deep learning model, respectively. Finally, we used an information fusion method to combine the results of the two models to obtain the final prediction. The different models, including the radiomics model, the deep learning model, and the fusion model, were evaluated on the validation set from two hospitals. In the external validation set, the radiomics model yielded an AUC of 0.883 (95% CI, 0.802-0.939), a sensitivity of 0.921, and a specificity of 0.735, and the deep learning model yielded an AUC of 0.873 (95% CI, 0.789-0.932), a sensitivity of 0.905, and a specificity of 0.853. The fusion model achieved an AUC of 0.947 (95% CI, 0.882-0.982), sensitivity of 0.825, and specificity of 0.941 in the external validation set. The fusion model has the potential to reduce the need for invasive biopsies of benign mammographic calcifications classified as BI-RADS category 4, without sacrificing the diagnostic accuracy for malignant cases.

Texture Analysis and Prediction of Response to Neoadjuvant Treatment in Patients with Locally Advanced Rectal Cancer

Background: The purpose of this study is to determine the relationship between the texture analysis extracted from preoperative rectal magnetic resonance (MR) studies and the response to neoadjuvant treatment. Materials and Methods: In total, 88 patients with rectal adenocarcinoma who underwent staging MR between 2017 and 2022 were retrospectively enrolled. After the completion of neoadjuvant treatment, they underwent surgical resection. The tumour regression grade (TRG) was collected. Patients with TRG 1–2 were classified as responders, while patients with TRG 3 to 5 were classified as non-responders. A texture analysis was conducted using LIFEx software (v 7.6.0), where T2-weighted MR sequences on oriented axial planes were uploaded, and a region of interest (ROI) was manually drawn on a single slice. Features with a Spearman correlation index > 0.5 have been discarded, and a LASSO feature selection has been applied. Selected features were trained using bootstrapping. Results: According to the TRG classes, 49 patients (55.8%) were considered responders, while 39 (44.2) were non-responders. Two features were associated with the responder class: GLCM_Homogeneity and Discretized Histo Entropy log 2. Regarding GLCM_Homogeneity, the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were 0.779 (95% CIs = 0.771–0.816), 86% (80–90), and 67% (60–71). Regarding Discretized Histo Entropy log 2, we found 0.775 AUC (0.700–0.801), 80% sensitivity (74–83), and 63% specificity (58–69). Combining both radiomics features the radiomics signature diagnostic accuracy increased (AUC = 0.844). Finally, the AUC of 1000 bootstraps were 0.810. Conclusions: Texture analysis can be considered an advanced tool for determining a possible correlation between pre-surgical MR data and the response to neoadjuvant therapy.

Prediction of lymphovascular invasion in esophageal squamous cell carcinoma by computed tomography-based radiomics analysis: 2D or 3D ?

BACKGROUND

To compare the performance between one-slice two-dimensional (2D) and whole-volume three-dimensional (3D) computed tomography (CT)-based radiomics models in the prediction of lymphovascular invasion (LVI) status in esophageal squamous cell carcinoma (ESCC).

METHODS

Two hundred twenty-four patients with ESCC (158 LVI-absent and 66 LVI-present) were enrolled in this retrospective study. The enrolled patients were randomly split into the training and testing sets with a 7:3 ratio. The 2D and 3D radiomics features were derived from the primary tumors' 2D and 3D regions of interest (ROIs) using 1.0 mm thickness contrast-enhanced CT (CECT) images. The 2D and 3D radiomics features were screened using inter-/intra-class correlation coefficient (ICC) analysis, Wilcoxon rank-sum test, Spearman correlation test, and the least absolute shrinkage and selection operator, and the radiomics models were built by multivariate logistic stepwise regression. The performance of 2D and 3D radiomics models was assessed by the area under the receiver operating characteristic (ROC) curve. The actual clinical utility of the 2D and 3D radiomics models was evaluated by decision curve analysis (DCA).

RESULTS

There were 753 radiomics features from 2D ROIs and 1130 radiomics features from 3D ROIs, and finally, 7 features were retained to construct 2D and 3D radiomics models, respectively. ROC analysis revealed that in both the training and testing sets, the 3D radiomics model exhibited higher AUC values than the 2D radiomics model (0.930 versus 0.852 and 0.897 versus 0.851, respectively). The 3D radiomics model showed higher accuracy than the 2D radiomics model in the training and testing sets (0.899 versus 0.728 and 0.788 versus 0.758, respectively). In addition, the 3D radiomics model has higher specificity and positive predictive value, while the 2D radiomics model has higher sensitivity and negative predictive value. The DCA indicated that the 3D radiomics model provided higher actual clinical utility regarding overall net benefit than the 2D radiomics model.

CONCLUSIONS

Both 2D and 3D radiomics features can be employed as potential biomarkers to predict the LVI in ESCC. The performance of the 3D radiomics model is better than that of the 2D radiomics model for the prediction of the LVI in ESCC.

A combined model using pre-treatment CT radiomics and clinicopathological features of non-small cell lung cancer to predict major pathological responses after neoadjuvant chemoimmunotherapy

OBJECTIVE

To investigate the relationship between clinical pathological characteristics, pretreatment CT radiomics, and major pathologic response (MPR) of non-small cell lung cancer (NSCLC) after neoadjuvant chemoimmunotherapy, and to establish a combined model to predict the major pathologic response of neoadjuvant chemoimmunotherapy.

METHODS

A retrospective study of 211 patients with NSCLC who underwent neoadjuvant chemoimmunotherapy and surgical treatment from January 2019 to April 2021 was conducted. The patients were divided into two groups: the MPR group and the non-MPR group. Pre-treatment CT images were segmented using ITK SNAP software to extract radiomics features using Python software. Then a radiomics model, a clinical model, and a combined model were constructed and validated using a receiver operating characteristic (ROC) curve. Finally, Delong's test was used to compare the three models.

RESULTS

The radiomics model achieved an AUC of 0.70 (95 % CI: 0.62-0.78) in the training group and 0.60 (95 % CI: 0.45-0.76) in the validation group. RECIST assessment results were screened from all clinical characteristics as independent factors for MPR with multivariate logistic regression analysis. The AUC of the clinical model for predicting MPR was 0.66 (95 % CI: 0.59-0.73) in the training group and 0.77 (95 % CI: 0.66-0.87) in the validation group. The combined model with combined radiomics and clinicopathological characteristics achieved an AUC was 0.76 (95 % CI: 0.68-0.84) in the training group, and 0.80 (95 % CI: 0.67-0.92) in the validation group. Delong's test showed that the AUC of the combined model was significantly higher than that of the radiomics model alone in both the training group (P = 0.0067) and the validation group (P = 0.0009).The calibration curve showed good agreement between predicted and actual MPR. Clinical decision curve analysis showed that the combined model was superior to radiomics alone.

CONCLUSIONS

Radiomics model can predict MPR in NSCLC after neoadjuvant chemoimmunotherapy with similar accuracy to RECIST assessment criteria. The combined model based on pretreatment CT radiomics and clinicopathological features showed better predictive power than independent radiomics model or independent clinicopathological features, suggesting that it may be more useful for guiding personalized neoadjuvant chemoimmunotherapy treatment strategies.

Application of radiomics for preoperative prediction of lymph node metastasis in colorectal cancer: a systematic review and meta-analysis

BACKGROUND

Colorectal cancer (CRC) stands as the third most prevalent cancer globally, projecting 3.2 million new cases and 1.6 million deaths by 2040. Accurate lymph node metastasis (LNM) detection is critical for determining optimal surgical approaches, including preoperative neoadjuvant chemoradiotherapy and surgery, which significantly influence CRC prognosis. However, conventional imaging lacks adequate precision, prompting exploration into radiomics, which addresses this shortfall by converting medical images into reproducible, quantitative data.

METHODS

Following PRISMA, Supplemental Digital Content 1 (http://links.lww.com/JS9/C77) and Supplemental Digital Content 2 (http://links.lww.com/JS9/C78), and AMSTAR-2 guidelines, Supplemental Digital Content 3 (http://links.lww.com/JS9/C79), we systematically searched PubMed, Web of Science, Embase, Cochrane Library, and Google Scholar databases until 11 January 2024, to evaluate radiomics models' diagnostic precision in predicting preoperative LNM in CRC patients. The quality and bias risk of the included studies were assessed using the Radiomics Quality Score (RQS) and the modified Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. Subsequently, statistical analyses were conducted.

RESULTS

Thirty-six studies encompassing 8039 patients were included, with a significant concentration in 2022-2023 (20/36). Radiomics models predicting LNM demonstrated a pooled area under the curve (AUC) of 0.814 (95% CI: 0.78-0.85), featuring sensitivity and specificity of 0.77 (95% CI: 0.69, 0.84) and 0.73 (95% CI: 0.67, 0.78), respectively. Subgroup analyses revealed similar AUCs for CT and MRI-based models, and rectal cancer models outperformed colon and colorectal cancers. Additionally, studies utilizing cross-validation, 2D segmentation, internal validation, manual segmentation, prospective design, and single-center populations tended to have higher AUCs. However, these differences were not statistically significant. Radiologists collectively achieved a pooled AUC of 0.659 (95% CI: 0.627, 0.691), significantly differing from the performance of radiomics models (P<0.001).

CONCLUSION

Artificial intelligence-based radiomics shows promise in preoperative lymph node staging for CRC, exhibiting significant predictive performance. These findings support the integration of radiomics into clinical practice to enhance preoperative strategies in CRC management.

Radiomics-based lymph nodes prognostic models from three MRI regions in nasopharyngeal carcinoma

Accurate prediction of the prognosis of nasopharyngeal carcinoma (NPC) is important for treatment. Lymph nodes metastasis is an important predictor for distant failure and regional recurrence in patients with NPC. Traditionally, subjective radiological evaluation increases concerns regarding the accuracy and consistency of predictions. Radiomics is an objective and quantitative evaluation algorithm for medical images. This retrospective analysis was conducted based on the data of 729 patients newly diagnosed with NPC without distant metastases to evaluate the performance of radiomics pretreatment using magnetic resonance imaging (MRI)-determined metastatic lymph nodes models to predict NPC prognosis with three delineation methods. Radiomics features were extracted from all lymph nodes (ALN), largest lymph node (LLN), and largest slice of the largest lymph node (LSLN) to generate three radiomics signatures. The radiomics signatures, clinical model, and radiomics-clinic merged models were developed in training cohort for predicting overall survival (OS). The results showed that LSLN signature with clinical factors predicted OS with high accuracy and robustness using pretreatment MR-determined metastatic lymph nodes (C-index [95 % confidence interval]: 0.762[0.760-0.763]), providing a new tool for treatment planning in NPC.

Establishment and verification of a prediction model based on clinical characteristics and computed tomography radiomics parameters for distinguishing benign and malignant pulmonary nodules

Background

The radiographic classification of pulmonary nodules into benign versus malignant categories is a pivotal component of early lung cancer diagnosis. The present study aimed to investigate clinical and computed tomography (CT) clinical-radiomics nomogram for preoperative differentiation of benign and malignant pulmonary nodules.

Methods

This retrospective study included 342 patients with pulmonary nodules who underwent high-resolution CT (HRCT) examination. We assigned them to a training dataset (n=239) and a validation dataset (n=103). There are 1781 tumor characteristics quantified by extracted features from the lesion segmented from patients' CT images. The features with poor reproducibility and high redundancy were removed. Then a least absolute shrinkage and selection operator (LASSO) logistic regression model with 10-fold cross-validation was used to further select features and build radiomics signatures. The independent predictive factors were identified by multivariate logistic regression. A radiomics nomogram was developed to predict the malignant probability. The performance and clinical utility of the clinical-radiomics nomogram was evaluated by receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA).

Results

After dimension reduction by the LASSO algorithm and multivariate logistic regression, four radiomic features were selected, including original_shape_Sphericity, exponential_glcm_Maximum Probability, log_sigma_2_0_mm_3D_glcm_Maximum Probability, and ogarithm_firstorder_90Percentile. Multivariate logistic regression showed that carcinoembryonic antigen (CEA) [odds ratio (OR) 95% confidence interval (CI): 1.40 (1.09-1.88)], CT rad score [OR (95% CI): 2.74 (2.03-3.85)], and cytokeratin-19-fragment (CYFRA21-1) [OR (95% CI): 1.80 (1.14-2.94)] were independent influencing factors of malignant pulmonary nodule (all P<0.05). The clinical-radiomics nomogram combining CEA, CYFRA21-1 and radiomics features achieved an area of curve (AUC) of 0.85 and 0.76 in the training group and verification group for the prediction of malignant pulmonary nodules. The clinical-radiomics nomogram demonstrated excellent agreement and practicality, as evidenced by the calibration curve and DCA.

Conclusions

The clinical-radiomics nomogram combined of CT-based radiomics signature, along with CYFRA21-1 and CEA, demonstrated strong predictive ability, calibration, and clinical usefulness in distinguishing between benign and malignant pulmonary nodules. The use of CT-based radiomics has the potential to assist clinicians in making informed decisions prior to biopsy or surgery while avoiding unnecessary treatment for non-cancerous lesions.

Computed tomography radiomics identification of T1-2 and T3-4 stages of esophageal squamous cell carcinoma: two-dimensional or three-dimensional?

BACKGROUND

To evaluate two-dimensional (2D) and three-dimensional (3D) computed tomography (CT) radiomics analysis for the T stage of esophageal squamous cell carcinoma (ESCC).

METHODS

398 patients with pathologically confirmed ESCC were divided into training and testing sets. All patients underwent chest CT scans preoperatively. For each tumor, based on CT images, a 2D region of interest (ROI) was outlined on the largest cross-sectional area, and a 3D ROI was outlined layer by layer on each section of the tumor. The radiomics platform was used for feature extraction. For feature selection, stepwise logistic regression was used. The receiver operating characteristic (ROC) curve was used to assess the diagnostic performance of the 2D radiomics model versus the 3D radiomics model. The differences were compared using the DeLong test. The value of the clinical utility of the two radiomics models was evaluated.

RESULTS

1595 radiomics features were extracted. After screening, two radiomics models were constructed. In the training set, the difference between the area under the curve (AUC) of the 2D radiomics model (AUC = 0.831) and the 3D radiomics model (AUC = 0.830) was not statistically significant (p = 0.973). In the testing set, the difference between the AUC of the 2D radiomics model (AUC = 0.807) and the 3D radiomics model (AUC = 0.797) was also not statistically significant (p = 0.748). A 2D model was equally useful as a 3D model in clinical situations.

CONCLUSION

The performance of 2D radiomics model is comparable to that of 3D radiomics model in distinguishing between the T1-2 and T3-4 stages of ESCC. In addition, 2D radiomics model may be a more feasible option due to the shorter time required for segmenting the ROI.

Machine-learning-based performance comparison of two-dimensional (2D) and three-dimensional (3D) CT radiomics features for intracerebral haemorrhage expansion

AIM

To investigate the value of non-contrast CT (NCCT)-based two-dimensional (2D) radiomics features in predicting haematoma expansion (HE) after spontaneous intracerebral haemorrhage (ICH) and compare its predictive ability with the three-dimensional (3D) signature.

MATERIALS AND METHODS

Three hundred and seven ICH patients who received baseline NCCT within 6 h of ictus from two stroke centres were analysed retrospectively. 2D and 3D radiomics features were extracted in the manner of one-to-one correspondence. The 2D and 3D models were generated by four different machine-learning algorithms (regularised L1 logistic regression, decision tree, support vector machine and AdaBoost), and the receiver operating characteristic (ROC) curve was used to compare their predictive performance. A robustness analysis was performed according to baseline haematoma volume.

RESULTS

Each feature type of 2D and 3D modalities used for subsequent analyses had excellent consistency (mean ICC >0.9). Among the different machine-learning algorithms, pairwise comparison showed no significant difference in both the training (mean area under the ROC curve [AUC] 0.858 versus 0.802, all p>0.05) and validation datasets (mean AUC 0.725 versus 0.678, all p>0.05), and the 10-fold cross-validation evaluation yielded similar results. The AUCs of the 2D and 3D models were comparable either in the binary or tertile volume analysis (all p>0.5).

CONCLUSION

NCCT-derived 2D radiomics features exhibited acceptable and similar performance to the 3D features in predicting HE, and this comparability seemed unaffected by initial haematoma volume. The 2D signature may be preferred in future HE-related radiomic works given its compatibility with emergency condition of ICH.

On-cloud decision-support system for non-small cell lung cancer histology characterization from thorax computed tomography scans

Non-Small Cell Lung Cancer (NSCLC) accounts for about 85% of all lung cancers. Developing non-invasive techniques for NSCLC histology characterization may not only help clinicians to make targeted therapeutic treatments but also prevent subjects from undergoing lung biopsy, which is challenging and could lead to clinical implications. The motivation behind the study presented here is to develop an advanced on-cloud decision-support system, named LUCY, for non-small cell LUng Cancer histologY characterization directly from thorax Computed Tomography (CT) scans. This aim was pursued by selecting thorax CT scans of 182 LUng ADenocarcinoma (LUAD) and 186 LUng Squamous Cell carcinoma (LUSC) subjects from four openly accessible data collections (NSCLC-Radiomics, NSCLC-Radiogenomics, NSCLC-Radiomics-Genomics and TCGA-LUAD), in addition to the implementation and comparison of two end-to-end neural networks (the core layer of whom is a convolutional long short-term memory layer), the performance evaluation on test dataset (NSCLC-Radiomics-Genomics) from a subject-level perspective in relation to NSCLC histological subtype location and grade, and the dynamic visual interpretation of the achieved results by producing and analyzing one heatmap video for each scan. LUCY reached test Area Under the receiver operating characteristic Curve (AUC) values above 77% in all NSCLC histological subtype location and grade groups, and a best AUC value of 97% on the entire dataset reserved for testing, proving high generalizability to heterogeneous data and robustness. Thus, LUCY is a clinically-useful decision-support system able to timely, non-invasively and reliably provide visually-understandable predictions on LUAD and LUSC subjects in relation to clinically-relevant information.

Lack of incremental value of three-dimensional measurement in assessing invasiveness for lung cancer

OBJECTIVES

The aim of this study was to evaluate the performance of consolidation-to-tumour ratio (CTR) and the radiomic models in two- and three-dimensional modalities for assessing radiological invasiveness in early-stage lung adenocarcinoma.

METHODS

A retrospective analysis was conducted on patients with early-stage lung adenocarcinoma from Guangdong Provincial People's Hospital and Shenzhen People's Hospital. Manual delineation of pulmonary nodules along the boundary was performed on cross-sectional images to extract radiomic features. Clinicopathological characteristics and radiomic signatures were identified in both cohorts. CTR and radiomic score for every patient were calculated. The performance of CTR and radiomic models were tested and validated in the respective cohorts.

RESULTS

A total of 818 patients from Guangdong Provincial People's Hospital were included in the primary cohort, while 474 patients from Shenzhen People's Hospital constituted an independent validation cohort. Both CTR and radiomic score were identified as independent factors for predicting pathological invasiveness. CTR in two- and three-dimensional modalities exhibited comparable results with areas under the receiver operating characteristic curves and were demonstrated in the validation cohort (area under the curve: 0.807 vs 0.826, P = 0.059) Furthermore, both CTR in two- and three-dimensional modalities was able to stratify patients with significant relapse-free survival (P < 0.000 vs P < 0.000) and overall survival (P = 0.003 vs P = 0.001). The radiomic models in two- and three-dimensional modalities demonstrated favourable discrimination and calibration in independent cohorts (P = 0.189).

CONCLUSIONS

Three-dimensional measurement provides no additional clinical benefit compared to two-dimensional.

Evaluating the Effectiveness of 2D and 3D CT Image Features for Predicting Tumor Response to Chemotherapy

Background and Objective: 2D and 3D tumor features are widely used in a variety of medical image analysis tasks. However, for chemotherapy response prediction, the effectiveness between different kinds of 2D and 3D features are not comprehensively assessed, especially in ovarian-cancer-related applications. This investigation aims to accomplish such a comprehensive evaluation. Methods: For this purpose, CT images were collected retrospectively from 188 advanced-stage ovarian cancer patients. All the metastatic tumors that occurred in each patient were segmented and then processed by a set of six filters. Next, three categories of features, namely geometric, density, and texture features, were calculated from both the filtered results and the original segmented tumors, generating a total of 1403 and 1595 features for the 2D and 3D tumors, respectively. In addition to the conventional single-slice 2D and full-volume 3D tumor features, we also computed the incomplete-3D tumor features, which were achieved by sequentially adding one individual CT slice and calculating the corresponding features. Support vector machine (SVM)-based prediction models were developed and optimized for each feature set. Five-fold cross-validation was used to assess the performance of each individual model. Results: The results show that the 2D feature-based model achieved an AUC (area under the ROC curve (receiver operating characteristic)) of 0.84 ± 0.02. When adding more slices, the AUC first increased to reach the maximum and then gradually decreased to 0.86 ± 0.02. The maximum AUC was yielded when adding two adjacent slices, with a value of 0.91 ± 0.01. Conclusions: This initial result provides meaningful information for optimizing machine learning-based decision-making support tools in the future.

Machine Learning and Radiomics of Bone Scintigraphy: Their Role in Predicting Recurrence of Localized or Locally Advanced Prostate Cancer

BACKGROUND

Machine-learning (ML) and radiomics features have been utilized for survival outcome analysis in various cancers. This study aims to investigate the application of ML based on patients' clinical features and radiomics features derived from bone scintigraphy (BS) and to evaluate recurrence-free survival in local or locally advanced prostate cancer (PCa) patients after the initial treatment.

METHODS

A total of 354 patients who met the eligibility criteria were analyzed and used to train the model. Clinical information and radiomics features of BS were obtained. Survival-related clinical features and radiomics features were included in the ML model training. Using the pyradiomics software, 128 radiomics features from each BS image's region of interest, validated by experts, were extracted. Four textural matrices were also calculated: GLCM, NGLDM, GLRLM, and GLSZM. Five training models (Logistic Regression, Naive Bayes, Random Forest, Support Vector Classification, and XGBoost) were applied using K-fold cross-validation. Recurrence was defined as either a rise in PSA levels, radiographic progression, or death. To assess the classifier's effectiveness, the ROC curve area and confusion matrix were employed.

RESULTS

Of the 354 patients, 101 patients were categorized into the recurrence group with more advanced disease status compared to the non-recurrence group. Key clinical features including tumor stage, radical prostatectomy, initial PSA, Gleason Score primary pattern, and radiotherapy were used for model training. Random Forest (RF) was the best-performing model, with a sensitivity of 0.81, specificity of 0.87, and accuracy of 0.85. The ROC curve analysis showed that predictions from RF outperformed predictions from other ML models with a final AUC of 0.94 and a p-value of <0.001. The other models had accuracy ranges from 0.52 to 0.78 and AUC ranges from 0.67 to 0.84.

CONCLUSIONS

The study showed that ML based on clinical features and radiomics features of BS improves the prediction of PCa recurrence after initial treatment. These findings highlight the added value of ML techniques for risk classification in PCa based on clinical features and radiomics features of BS.

Multi-dimensional radiomics analysis to predict visceral pleural invasion in lung adenocarcinoma of ≤3 cm maximum diameter

AIM

To explore the value of radiomics analysis in preoperatively predicting visceral pleural invasion (VPI) of lung adenocarcinoma (LAC) with ≤3 cm maximum diameter and to compare the performance of two-dimensional (2D) and three-dimensional (3D) computed tomography (CT) radiomics models.

MATERIALS AND METHODS

A total of 391 LAC patients were enrolled retrospectively, of whom 142 were VPI (+) and 249 were VPI (-). Radiomics features were extracted from 2D and 3D regions of interest (ROIs) of tumours in CT images. 2D and 3D radiomics models were developed combining the optimal radiomics features by using the logistic regression machine-learning method and radiomics scores (rad-scores) were calculated. Nomograms were constructed by integrating independent risk factors and rad-scores. The performance of each model was evaluated by using the receiver operator characteristic (ROC) curve, decision curve analysis (DCA), clinical impact curve (CIC), and calculating the area under the curve (AUC).

RESULTS

There was no difference in the VPI prediction between 2D and 3D radiomics models (training group: 2D AUC=0.835, 3D AUC=0.836, p=0.896; validation group: 2D AUC=0.803, 3D AUC=0.794, p=0.567). The 2D and 3D nomograms performed similarly regarding discrimination (training group: 2D AUC=0.867, 3D AUC=0.862, p=0.409, validation group: 2D AUC=0.835, 3D AUC=0.827, p=0.558), and outperformed their corresponding radiomics models and the clinical model. DCA and CIC revealed that the 2D nomogram had slightly better clinical utility.

CONCLUSION

The 2D radiomics model has a similar discrimination capability compared with the 3D radiomics model. The 2D nomogram performs slightly better for individual VPI prediction in LAC.

Construction of 3D and 2D contrast-enhanced CT radiomics for prediction of CGB3 expression level and clinical prognosis in bladder cancer

Objective

The purpose of this study was to construct a 3D and 2D contrast-enhanced computed tomography (CECT) radiomics model to predict CGB3 levels and assess its prognostic abilities in bladder cancer (Bca) patients.

Methods

Transcriptome data and CECT images of Bca patients were downloaded from The Cancer Imaging Archive (TCIA) and The Cancer Genome Atlas (TCGA) database. Clinical data of 43 cases from TCGA and TCIA were used for radiomics model evaluation. The Volume of interest (VOI) (3D) and region of interest (ROI) (2D) radiomics features were extracted. For the construction of predicting radiomics models, least absolute shrinkage and selection operator regression were used, and the filtered radiomics features were fitted using the logistic regression algorithm (LR). The model's effectiveness was measured using 10-fold cross-validation and the area under the receiver operating characteristic curve (AUC of ROC).

Result

CGB3 was a differential expressed prognosis-related gene and involved in the immune response process of plasma cells and T cell gamma delta. The high levels of CGB3 are a risk element for overall survival (OS). The AUCs of VOI and ROI radiomics models in the training set were 0.841 and 0.776, while in the validation set were 0.815 and 0.754, respectively. The Delong test revealed that the AUCs of the two models were not statistically different, and both models had good predictive performance.

Conclusion

The CGB3 expression level is an important prognosis factor for Bca patients. Both 3D and 2D CECT radiomics are effective in predicting CGB3 expression levels.

The predictive value of a computed tomography-based radiomics model for the surgical separability of thymic epithelial tumors from the superior vena cava and the left innominate vein

Background

The aim of this study was to develop a radiomics machine learning model based on computed tomography (CT) that can predict whether thymic epithelial tumors (TETs) can be separated from veins during surgery and to compare the accuracy of the radiomics model to that of radiologists.

Methods

Patients who underwent thymectomy at our hospital from 2009 to 2017 were included in the screening process. After the selection of patients according to the inclusion and exclusion criteria, the cohort was randomly divided into training and testing groups, and CT images of these patients were collected. Subsequently, two-dimensional (2D) and three-dimensional (3D) regions of interest were labelled using ITK-SNAP 3.8.0 software, and Radiomics features were extracted using Python software (Python Software Foundation) and selected through the least absolute shrinkage and selection operator (LASSO) regression model. To construct the classifier, a support vector machine (SVM) was employed, and a nomogram was created using logistic regression to predict vascular inseparable TETs based on the radiomics score (radscore) and image features. To assess the accuracy of these models, area under receiver operating characteristic (ROC) curves of these models were calculated, and differences among the models were identified using the Delong test.

Results

In this retrospective study, 204 patients with TETs were included, among whom 21 were diagnosed with surgical vascularly inseparable TETs. The area under ROC curve (AUC) of the 2D model, 3D model, 2D + 3D model, and radiologist diagnoses were 0.94, 0.92, 0.95, and 0.87 in the training cohort and 0.95, 0.92, 0.98, and 0.78 in testing cohort, respectively. The Delong test revealed a significant improvement in the performance of the radiomics models compared to radiologists' diagnoses. The logistic regression selected 3 image features, namely maximum diameter of the tumor, degree of abutment of vessel circumference >50%, and absence of the mediastinal fat layer or space between the tumor and surrounding structures. These features, along with the radscore, were included to develop a nomogram. The AUCs of this nomogram were 0.99 in both the training set and testing set, and the Delong test did not find a significant difference between ROC plots of the nomogram and radiomics models.

Conclusions

The proposed radiomics model could accurately predict surgical vascularly inseparable TETs preoperatively and was shown to have a higher predictive value than the radiologists.

Deep learning nomogram based on Gd-EOB-DTPA MRI for predicting early recurrence in hepatocellular carcinoma after hepatectomy

OBJECTIVES

The accurate prediction of post-hepatectomy early recurrence in patients with hepatocellular carcinoma (HCC) is crucial for decision-making regarding postoperative adjuvant treatment and monitoring. We aimed to explore the feasibility of deep learning (DL) features derived from gadoxetate disodium (Gd-EOB-DTPA) MRI, qualitative features, and clinical variables for predicting early recurrence.

METHODS

In this bicentric study, 285 patients with HCC who underwent Gd-EOB-DTPA MRI before resection were divided into training (n = 195) and validation (n = 90) sets. DL features were extracted from contrast-enhanced MRI images using VGGNet-19. Three feature selection methods and five classification methods were combined for DL signature construction. Subsequently, an mp-MR DL signature fused with multiphase DL signatures of contrast-enhanced images was constructed. Univariate and multivariate logistic regression analyses were used to identify early recurrence risk factors including mp-MR DL signature, microvascular invasion (MVI), and tumor number. A DL nomogram was built by incorporating deep features and significant clinical variables to achieve early recurrence prediction.

RESULTS

MVI (p = 0.039), tumor number (p = 0.001), and mp-MR DL signature (p < 0.001) were independent risk factors for early recurrence. The DL nomogram outperformed the clinical nomogram in the training set (AUC: 0.949 vs. 0.751; p < 0.001) and validation set (AUC: 0.909 vs. 0.715; p = 0.002). Excellent DL nomogram calibration was achieved in both training and validation sets. Decision curve analysis confirmed the clinical usefulness of DL nomogram.

CONCLUSION

The proposed DL nomogram was superior to the clinical nomogram in predicting early recurrence for HCC patients after hepatectomy.

KEY POINTS

• Deep learning signature based on Gd-EOB-DTPA MRI was the predominant independent predictor of early recurrence for hepatocellular carcinoma (HCC) after hepatectomy. • Deep learning nomogram based on clinical factors and Gd-EOB-DTPA MRI features is promising for predicting early recurrence of HCC. • Deep learning nomogram outperformed the conventional clinical nomogram in predicting early recurrence.

Artificial intelligence-driven radiomics study in cancer: the role of feature engineering and modeling

Modern medicine is reliant on various medical imaging technologies for non-invasively observing patients' anatomy. However, the interpretation of medical images can be highly subjective and dependent on the expertise of clinicians. Moreover, some potentially useful quantitative information in medical images, especially that which is not visible to the naked eye, is often ignored during clinical practice. In contrast, radiomics performs high-throughput feature extraction from medical images, which enables quantitative analysis of medical images and prediction of various clinical endpoints. Studies have reported that radiomics exhibits promising performance in diagnosis and predicting treatment responses and prognosis, demonstrating its potential to be a non-invasive auxiliary tool for personalized medicine. However, radiomics remains in a developmental phase as numerous technical challenges have yet to be solved, especially in feature engineering and statistical modeling. In this review, we introduce the current utility of radiomics by summarizing research on its application in the diagnosis, prognosis, and prediction of treatment responses in patients with cancer. We focus on machine learning approaches, for feature extraction and selection during feature engineering and for imbalanced datasets and multi-modality fusion during statistical modeling. Furthermore, we introduce the stability, reproducibility, and interpretability of features, and the generalizability and interpretability of models. Finally, we offer possible solutions to current challenges in radiomics research.

Early Detection of Lung Cancer in the NLST Dataset

Lung Cancer is the leading cause of cancer mortality in the U.S. The effectiveness of standard treatments, including surgery, chemotherapy or radiotherapy, depends on several factors like type and stage of cancer, with the survival rate being much worse for later cancer stages. The National Lung Screening Trial (NLST) established that patients screened using low-dose Computed Tomography (CT) had a 15 to 20 percent lower risk of dying from lung cancer than patients screened using chest X-rays. While CT excelled at detecting small early stage malignant nodules, a large proportion of patients (> 25%) screened positive and only a small fraction (< 10%) of these positive screens actually had or developed cancer in the subsequent years. We developed a model to distinguish between high and low risk patients among the positive screens, predicting the likelihood of having or developing lung cancer at the current time point or in subsequent years non-invasively, based on current and previous CT imaging data. However, most of the nodules in NLST are very small, and nodule segmentations or even precise locations are unavailable. Our model comprises two stages: the first stage is a neural network model trained on the Lung Image Database Consortium (LIDC-IDRI) cohort which detects nodules and assigns them malignancy scores. The second part of our model is a boosted tree which outputs a cancer probability for a patient based on the nodule information (location and malignancy score) predicted by the first stage. Our model, built on a subset of the NLST cohort (n = 1138) shows excellent performance, achieving an area under the receiver operating characteristics curve (ROC AUC) of 0.85 when predicting based on CT images from all three time points available in the NLST dataset.

Multisequence magnetic resonance imaging-based radiomics models for the prediction of microsatellite instability in endometrial cancer

PURPOSE

To evaluate the performance of multisequence magnetic resonance imaging (MRI)-based radiomics models in the assessment of microsatellite instability (MSI) status in endometrial cancer (EC).

MATERIALS AND METHODS

This retrospective multicentre study included 338 EC patients with available MSI status and preoperative MRI scans, divided into training (37 MSI, 123 microsatellite stability [MSS]), internal validation (15 MSI, 52 MSS), and external validation cohorts (30 MSI, 81 MSS). Radiomics features were extracted from T2-weighted images, diffusion-weighted images, and contrast-enhanced T1-weighted images. The ComBat harmonisation method was applied to remove intrascanner variability. The Boruta wrapper algorithm was used for key feature selection. Three classification algorithms, logistic regression (LR), random forest (RF), and support vector machine (SVM), were applied to build the radiomics models. The area under the receiver operating characteristic curve (AUC) was calculated to compare the diagnostic performance of the models. Decision curve analysis (DCA) was conducted to determine the clinical usefulness of the models.

RESULTS

Among the 1980 features, Boruta finally selected nine radiomics features. A higher MSI prediction performance was achieved after running the ComBat harmonisation method. The SVM algorithm had the best performance, with AUCs of 0.921, 0.903, and 0.937 in the training, internal validation, and external validation cohorts, respectively. The DCA results showed that the SVM algorithm achieved higher net benefits than the other classifiers over a threshold range of 0.581-0.783.

CONCLUSION

The multisequence MRI-based radiomics models showed promise in preoperatively predicting the MSI status in EC in this multicentre setting.

Feature selection methods and predictive models in CT lung cancer radiomics

Radiomics is a technique that extracts quantitative features from medical images using data-characterization algorithms. Radiomic features can be used to identify tissue characteristics and radiologic phenotyping that is not observable by clinicians. A typical workflow for a radiomics study includes cohort selection, radiomic feature extraction, feature and predictive model selection, and model training and validation. While there has been increasing attention given to radiomic feature extraction, standardization, and reproducibility, currently, there is a lack of rigorous evaluation of feature selection methods and predictive models. Herein, we review the published radiomics investigations in CT lung cancer and provide an overview of the commonly used radiomic feature selection methods and predictive models. We also compare limitations of various methods in clinical applications and present sources of uncertainty associated with those methods. This review is expected to help raise awareness of the impact of radiomic feature and model selection methods on the integrity of radiomics studies.

Performance comparison of 2D and 3D MRI radiomics features in meningioma grade prediction: A preliminary study

Objectives

The objective of this study was to compare the predictive performance of 2D and 3D radiomics features in meningioma grade based on enhanced T1 WI images.

Methods

There were 170 high grade meningioma and 170 low grade meningioma were selected randomly. The 2D and 3D features were extracted from 2D and 3D ROI of each meningioma. The Spearman correlation analysis and least absolute shrinkage and selection operator (LASSO) regression were used to select the valuable features. The 2D and 3D predictive models were constructed by naive Bayes (NB), gradient boosting decision tree (GBDT), and support vector machine (SVM). The ROC curve was drawn and AUC was calculated. The 2D and 3D models were compared by Delong test of AUCs and decision curve analysis (DCA) curve.

Results

There were 1143 features extracted from each ROI. Six and seven features were selected. The AUC of 2D and 3D model in NB, GBDT, and SVM was 0.773 and 0.771, 0.722 and 0.717, 0.733 and 0.743. There was no significant difference in two AUCs (p=0.960, 0.913, 0.830) between 2D and 3D model. The 2D features had a better performance than 3D features in NB models and the 3D features had a better performance than 2D features in GBDT models. The 2D features and 3D features had an equal performance in SVM models.

Conclusions

The 2D and 3D features had a comparable performance in predicting meningioma grade. Considering the issue of time and labor, 2D features could be selected for radiomics study in meningioma.

Key points

There was a comparable performance between 2D and 3D features in meningioma grade prediction. The 2D features was a proper selection in meningioma radiomics study because of its time and labor saving.

Radiomics in Abdominopelvic Solid-Organ Oncologic Imaging: Current Status

Radiomics is the process of extraction of high-throughput quantitative imaging features from medical images. These features represent noninvasive quantitative biomarkers that go beyond the traditional imaging features visible to the human eye. This article first reviews the steps of the radiomics pipeline, including image acquisition, ROI selection and image segmentation, image preprocessing, feature extraction, feature selection, and model development and application. Current evidence for the application of radiomics in abdominopelvic solid-organ cancers is then reviewed. Applications including diagnosis, subtype determination, treatment response assessment, and outcome prediction are explored within the context of hepatobiliary and pancreatic cancer, renal cell carcinoma, prostate cancer, gynecologic cancer, and adrenal masses. This literature review focuses on the strongest available evidence, including systematic reviews, meta-analyses, and large multicenter studies. Limitations of the available literature are highlighted, including marked heterogeneity in radiomics methodology, frequent use of small sample sizes with high risk of overfitting, and lack of prospective design, external validation, and standardized radiomics workflow. Thus, although studies have laid a foundation that supports continued investigation into radiomics models, stronger evidence is needed before clinical adoption.

Prediction of Individual Lymph Node Metastatic Status in Esophageal Squamous Cell Carcinoma Using Routine Computed Tomography Imaging: Comparison of Size-Based Measurements and Radiomics-Based Models

BACKGROUND

Lymph node status is vital for prognosis and treatment decisions for esophageal squamous cell carcinoma (ESCC). This study aimed to construct and evaluate an optimal radiomics-based method for a more accurate evaluation of individual regional lymph node status in ESCC and to compare it with traditional size-based measurements.

METHODS

The study consecutively collected 3225 regional lymph nodes from 530 ESCC patients receiving upfront surgery from January 2011 to October 2015. Computed tomography (CT) scans for individual lymph nodes were analyzed. The study evaluated the predictive performance of machine-learning models trained on features extracted from two-dimensional (2D) and three-dimensional (3D) radiomics by different contouring methods. Robust and important radiomics features were selected, and classification models were further established and validated.

RESULTS

The lymph node metastasis rate was 13.2% (427/3225). The average short-axis diameter was 6.4 mm for benign lymph nodes and 7.9 mm for metastatic lymph nodes. The division of lymph node stations into five regions according to anatomic lymph node drainage (cervical, upper mediastinal, middle mediastinal, lower mediastinal, and abdominal regions) improved the predictive performance. The 2D radiomics method showed optimal diagnostic results, with more efficient segmentation of nodal lesions. In the test set, this optimal model achieved an area under the receiver operating characteristic curve of 0.841-0.891, an accuracy of 84.2-94.7%, a sensitivity of 65.7-83.3%, and a specificity of 84.4-96.7%.

CONCLUSIONS

The 2D radiomics-based models noninvasively predicted the metastatic status of an individual lymph node in ESCC and outperformed the conventional size-based measurement. The 2D radiomics-based model could be incorporated into the current clinical workflow to enable better decision-making for treatment strategies.

A radiomics nomogram for predicting the meningioma grade based on enhanced T1WI images

OBJECTIVES

The objective of this study was to develop a radiomics nomogram for predicting the meningioma grade based on enhanced T1 weighted imaging (T1WI) images.

METHODS

188 patients with meningioma were analyzed retrospectively. There were 94 high-grade meningioma to form high-grade group and 94 low-grade meningioma were selected randomly to form low-grade group. Clinical data and MRI features were analyzed and compared. The clinical model was built by using the significant variables. The least absolute shrinkage and selection operator regression was used to select the most valuable radiomics feature. The radiomics signature was built and the Rad-score was calculated. The radiomics nomogram was developed by the significant variables of the clinical factors and Rad-score. The calibration curve and the Hosmer-Lemeshow test were used to evaluate the radiomics nomogram. Different models were compared by Delong test and decision curve analysis curve.

RESULTS

The sex, size and surrounding invasion were used to build clinical model. The area under the receiver operator characteristic curve (AUC) of clinical model was 0.870 (95% CI: 0.782-0.959). Nine features were used to construct the radiomics signature. The AUC of the radiomics signature was 0.885 (95% CI: 0.802-0.968). The AUC of radiomics nomogram was 0.952 (95% CI: 0.904-1). The AUC of radiomics nomogram was higher than that of clinical model and radiomics signature with a significant difference (p＜0.05). The decision curve analysis curve showed that the radiomics nomogram had a larger net benefit than the clinical model and radiomics signature.

CONCLUSION

The radiomics nomogram based on enhanced T1 weighted imaging images for predicting the meningioma grade showed high predictive value and might contribute to the diagnosis and treatment of meningioma.

ADVANCES IN KNOWLEDGE

1. We first constructed a radiomic nomogram to predict the meningioma grade.2. We compared the results of the clinical model, radiomics signature and radiomics nomogram.

A primer on texture analysis in abdominal radiology

The number of publications on texture analysis (TA), radiomics, and radiogenomics has been growing exponentially, with abdominal radiologists aiming to build new prognostic or predictive biomarkers for a wide range of clinical applications including the use of oncological imaging to advance the field of precision medicine. TA is specifically concerned with the study of the variation of pixel intensity values in radiological images. Radiologists aim to capture pixel variation in radiological images to deliver new insights into tumor biology that cannot be derived from visual inspection alone. TA remains an active area of investigation and requires further standardization prior to its clinical acceptance and applicability. This review is for radiologists interested in this rapidly evolving field, who are thinking of performing research or want to better interpret results in this arena. We will review the main concepts in TA, workflow processes, and existing challenges and steps to overcome them, as well as look at publications in body imaging with external validation.

Repeatability and reproducibility of magnetic resonance imaging-based radiomic features in rectal cancer

Purpose: Radiomics of magnetic resonance images (MRIs) in rectal cancer can non-invasively characterize tumor heterogeneity with potential to discover new imaging biomarkers. However, for radiomics to be reliable, the imaging features measured must be stable and reproducible. The aim of this study is to quantify the repeatability and reproducibility of MRI-based radiomic features in rectal cancer. Approach: An MRI radiomics phantom was used to measure the longitudinal repeatability of radiomic features and the impact of post-processing changes related to image resolution and noise. Repeatability measurements in rectal cancers were also quantified in a cohort of 10 patients with test-retest imaging among two observers. Results: We found that many radiomic features, particularly from texture classes, were highly sensitive to changes in image resolution and noise. About 49% of features had coefficient of variations ≤ 10 % in longitudinal phantom measurements. About 75% of radiomic features in in vivo test-retest measurements had an intraclass correlation coefficient of ≥ 0.8 . We saw excellent interobserver agreement with mean Dice similarity coefficient of 0.95 ± 0.04 for test and retest scans. Conclusions: The results of this study show that even when using a consistent imaging protocol many radiomic features were unstable. Therefore, caution must be taken when selecting features for potential imaging biomarkers.

Automatic Feature Construction Based on Genetic Programming for Survival Prediction in Lung Cancer Using CT Images

In the radiomics workflow, machine learning builds classification models from a set of input features. However, some features can be irrelevant and redundant, reducing the classification performance. This paper proposes using the Genetic Programming (GP) algorithm to automatically construct a reduced number of independent and relevant radiomic features. The proposed method is applied to patients affected by Non-Small Cell Lung Cancer (NSCLC) with pre-operative computed tomography (CT) images to predict the two-year survival by the use of linear classifiers. The model built using GP features is compared with benchmark models built using traditional features. The use of the GP algorithm increased classification performance: [Formula: see text] for the proposed model vs. [Formula: see text] and 0.64 for the benchmark models. Hence, the proposed approach better stratifies patients at high and low risk according to their overall postoperative survival time.

Lung nodule diagnosis and cancer histology classification from computed tomography data by convolutional neural networks: A survey

Lung cancer is among the deadliest cancers. Besides lung nodule classification and diagnosis, developing non-invasive systems to classify lung cancer histological types/subtypes may help clinicians to make targeted treatment decisions timely, having a positive impact on patients' comfort and survival rate. As convolutional neural networks have proven to be responsible for the significant improvement of the accuracy in lung cancer diagnosis, with this survey we intend to: show the contribution of convolutional neural networks not only in identifying malignant lung nodules but also in classifying lung cancer histological types/subtypes directly from computed tomography data; point out the strengths and weaknesses of slice-based and scan-based approaches employing convolutional neural networks; and highlight the challenges and prospective solutions to successfully apply convolutional neural networks for such classification tasks. To this aim, we conducted a comprehensive analysis of relevant Scopus-indexed studies involved in lung nodule diagnosis and cancer histology classification up to January 2022, dividing the investigation in convolutional neural network-based approaches fed with planar or volumetric computed tomography data. Despite the application of convolutional neural networks in lung nodule diagnosis and cancer histology classification is a valid strategy, some challenges raised, mainly including the lack of publicly-accessible annotated data, together with the lack of reproducibility and clinical interpretability. We believe that this survey will be helpful for future studies involved in lung nodule diagnosis and cancer histology classification prior to lung biopsy by means of convolutional neural networks.

Multiplexed imaging in oncology

In oncology, technologies for clinical molecular imaging are used to diagnose patients, establish the efficacy of treatments and monitor the recurrence of disease. Multiplexed methods increase the number of disease-specific biomarkers that can be detected simultaneously, such as the overexpression of oncogenic proteins, aberrant metabolite uptake and anomalous blood perfusion. The quantitative localization of each biomarker could considerably increase the specificity and the accuracy of technologies for clinical molecular imaging to facilitate granular diagnoses, patient stratification and earlier assessments of the responses to administered therapeutics. In this Review, we discuss established techniques for multiplexed imaging and the most promising emerging multiplexing technologies applied to the imaging of isolated tissues and cells and to non-invasive whole-body imaging. We also highlight advances in radiology that have been made possible by multiplexed imaging.

Automated segmentation of the fractured vertebrae on CT and its applicability in a radiomics model to predict fracture malignancy

Although CT radiomics has shown promising results in the evaluation of vertebral fractures, the need for manual segmentation of fractured vertebrae limited the routine clinical implementation of radiomics. Therefore, automated segmentation of fractured vertebrae is needed for successful clinical use of radiomics. In this study, we aimed to develop and validate an automated algorithm for segmentation of fractured vertebral bodies on CT, and to evaluate the applicability of the algorithm in a radiomics prediction model to differentiate benign and malignant fractures. A convolutional neural network was trained to perform automated segmentation of fractured vertebral bodies using 341 vertebrae with benign or malignant fractures from 158 patients, and was validated on independent test sets (internal test, 86 vertebrae [59 patients]; external test, 102 vertebrae [59 patients]). Then, a radiomics model predicting fracture malignancy on CT was constructed, and the prediction performance was compared between automated and human expert segmentations. The algorithm achieved good agreement with human expert segmentation at testing (Dice similarity coefficient, 0.93-0.94; cross-sectional area error, 2.66-2.97%; average surface distance, 0.40-0.54 mm). The radiomics model demonstrated good performance in the training set (AUC, 0.93). In the test sets, automated and human expert segmentations showed comparable prediction performances (AUC, internal test, 0.80 vs 0.87, p = 0.044; external test, 0.83 vs 0.80, p = 0.37). In summary, we developed and validated an automated segmentation algorithm that showed comparable performance to human expert segmentation in a CT radiomics model to predict fracture malignancy, which may enable more practical clinical utilization of radiomics.

Radiomics Based on Digital Mammography Helps to Identify Mammographic Masses Suspicious for Cancer

Objectives

This study aims to build radiomics model of Breast Imaging Reporting and Data System (BI-RADS) category 4 and 5 mammographic masses extracted from digital mammography (DM) for mammographic masses characterization by using a sensitivity threshold similar to that of biopsy.

Materials and Methods

This retrospective study included 288 female patients (age, 52.41 ± 10.31) who had BI-RADS category 4 or 5 mammographic masses with an indication for biopsy. The patients were divided into two temporal set (training set, 82 malignancies and 110 benign lesions; independent test set, 48 malignancies and 48 benign lesions). A total of 188 radiomics features were extracted from mammographic masses on the combination of craniocaudal (CC) position images and mediolateral oblique (MLO) position images. For the training set, Pearson’s correlation and the least absolute shrinkage and selection operator (LASSO) were used to select non-redundant radiomics features and useful radiomics features, respectively, and support vector machine (SVM) was applied to construct a radiomics model. The receiver operating characteristic curve (ROC) analysis was used to evaluate the classification performance of the radiomics model and to determine a threshold value with a sensitivity higher than 98% to predict the mammographic masses malignancy. For independent test set, identical threshold value was used to validate the classification performance of the radiomics model. The stability of the radiomics model was evaluated by using a fivefold cross-validation method, and two breast radiologists assessed the diagnostic agreement of the radiomics model.

Results

In the training set, the radiomics model obtained an area under the receiver operating characteristic curve (AUC) of 0.934 [95% confidence intervals (95% CI), 0.898–0.971], a sensitivity of 98.8% (81/82), a threshold of 0.22, and a specificity of 60% (66/110). In the test set, the radiomics model obtained an AUC of 0.901 (95% CI, 0.835–0.961), a sensitivity of 95.8% (46/48), and a specificity of 66.7% (32/48). The radiomics model had relatively stable sensitivities in fivefold cross-validation (training set, 97.39% ± 3.9%; test set, 98.7% ± 4%).

Conclusion

The radiomics method based on DM may help reduce the temporarily unnecessary invasive biopsies for benign mammographic masses over-classified in BI-RADS category 4 and 5 while providing similar diagnostic performance for malignant mammographic masses as biopsies.

Application Values of 2D and 3D Radiomics Models Based on CT Plain Scan in Differentiating Benign from Malignant Ovarian Tumors

Background

Accurate identification of ovarian tumors as benign or malignant is highly crucial. Radiomics is a new branch of imaging that has emerged in recent years to replace the traditional naked eye qualitative diagnosis.

Objective

This study is aimed at exploring the difference in the application potential of two- (2D) and three-dimensional (3D) radiomics models based on CT plain scan in differentiating benign from malignant ovarian tumors.

Method

A retrospective analysis was performed on 140 patients with ovarian tumors confirmed by surgery and pathology in our hospital from July 2017 to August 2020. These 140 patients were divided into benign group and malignant group according to the pathological results. The ITK-SNAP software was used to outline the regions-of-interest (ROI) of 2D or 3D tumors on the CT plain scan image of each patient; the texture features were extracted through analysis kit (AK), and the cases were randomly divided into training groups (n = 99) and validation group (n = 41) in a ratio of 7 : 3. The least absolute shrinkage and selection operator (LASSO) algorithm was used to perform dimensionality reduction, followed by the construction of the radiomics nomogram model using the logistic regression method. The receiver operating characteristic (ROC) curve was drawn, and the calibration curve and decision curve analysis (DCA) were used to evaluate and verify the results of the radiomics nomogram and compare the differences between 2D and 3D diagnostic performance.

Results

There were 396 quantitative radiomics feature parameters extracted from 2D group and the 3D group, respectively. The area under the curve (AUC) of the radiomics nomogram of the 2D training group and the validation group were 0.96 and 0.97, respectively. The accuracy, specificity, and sensitivity of the training set were 92.9%, 88.9%, and 96.3%, respectively, and those of the validation set were 90.2%, 82.6%, and 100.0%, respectively. The AUCs of the radiomics nomogram of the 3D training group and validation group were 0.96% and 0.99%, respectively. The accuracy, sensitivity, and specificity of the training set were 92.9%, 96.3%, and 88.9%, respectively, and those of the validation set were 97.6%, 95.7%, and 100.0%, respectively. DeLong's test indicated that there was no statistical significance between the two sets (P > 0.05).

Conclusions

For the differential diagnosis of benign and malignant ovarian tumors, the 2D and 3D radiomics nomogram models exhibited comparable diagnostic performance. Considering that the 2D model was cost-effective and time-efficient, it was more recommended to use 2D features in future research.

Two-dimensional and three-dimensional T2 weighted imaging-based radiomic signatures for the preoperative discrimination of ovarian borderline tumors and malignant tumors

BACKGROUND

Ovarian cancer is the most women malignancy in the whole world. It is difficult to differentiate ovarian cancers from ovarian borderline tumors because of some similar imaging findings.Radiomics study may help clinicians to make a proper diagnosis before invasive surgery.

PURPOSE

To evaluate the ability of T2-weighted imaging (T2WI)-based radiomics to discriminate ovarian borderline tumors (BOTs) from malignancies based on two-dimensional (2D) and three-dimensional (3D) lesion segmentation methods.

METHODS

A total of 95 patients with pathologically proven ovarian BOTs and 101 patients with malignancies were retrospectively included in this study. We evaluated the diagnostic performance of the signatures derived from T2WI-based radiomics in their ability to differentiate between BOTs and malignancies and compared the performance differences in the 2D and 3D segmentation models. The least absolute shrinkage and selection operator method (Lasso) was used for radiomics feature selection and machine learning processing.

RESULTS

The radiomics score between BOTs and malignancies in four types of selected T2WI-based radiomics models differed significantly at the statistical level (p < 0.0001). For the classification between BOTs and malignant masses, the 2D and 3D coronal T2WI-based radiomics models yielded accuracy values of 0.79 and 0.83 in the testing group, respectively; the 2D and 3D sagittal fat-suppressed (fs) T2WI-based radiomics models yielded an accuracy of 0.78 and 0.99, respectively.

CONCLUSIONS

Our results suggest that T2WI-based radiomic features were highly correlated with ovarian tumor subtype classification. 3D-sagittal MRI radiomics features may help clinicians differentiate ovarian BOTs from malignancies with high ACC.

Learning-based Cancer Treatment Outcome Prognosis using Multimodal Biomarkers

Predicting early in treatment whether a tumor is likely to be responsive is a difficult yet important task to support clinical decision-making. Studies have shown that multimodal biomarkers could provide complementary information and lead to more accurate treatment outcome prognosis than unimodal biomarkers. However, the prognosis accuracy could be affected by multimodal data heterogeneity and incompleteness. The small-sized and imbalance datasets also bring additional challenges for training a designed prognosis model. In this study, a modular framework employing multimodal biomarkers for cancer treatment outcome prediction was proposed. It includes four modules of synthetic data generation, deep feature extraction, multimodal feature fusion, and classification to address the challenges described above. The feasibility and advantages of the designed framework were demonstrated through an example study, in which the goal was to stratify oropharyngeal squamous cell carcinoma (OPSCC) patients with low- and high-risks of treatment failures by use of positron emission tomography (PET) image data and microRNA (miRNA) biomarkers. The superior prognosis performance and the comparison with other methods demonstrated the efficiency of the proposed framework and its ability of enabling seamless integration, validation and comparison of various algorithms in each module of the framework. The limitation and future work was discussed as well.

MR-based radiomics-clinical nomogram in epithelial ovarian tumor prognosis prediction: tumor body texture analysis across various acquisition protocols

Background

Epithelial ovarian cancer (EOC) is the most malignant gynecological tumor in women. This study aimed to construct and compare radiomics-clinical nomograms based on MR images in EOC prognosis prediction.

Methods

A total of 186 patients with pathologically proven EOC were enrolled and randomly divided into a training cohort (n = 130) and a validation cohort (n = 56). Clinical characteristics of each patient were retrieved from the hospital information system. A total of 1116 radiomics features were extracted from tumor body on T2-weighted imaging (T2WI), T1-weighted imaging (T1WI), diffusion weighted imaging (DWI) and contrast-enhanced T1-weighted imaging (CE-T1WI). Paired sequence signatures were constructed, selected and trained to build a prognosis prediction model. Radiomic-clinical nomogram was constructed based on multivariate logistic regression analysis with radiomics score and clinical features. The predictive performance was evaluated by receiver operating characteristic curve (ROC) analysis, decision curve analysis (DCA) and calibration curve.

Results

The T2WI radiomic-clinical nomogram achieved a favorable prediction performance in the training and validation cohort with an area under ROC curve (AUC) of 0.866 and 0.818, respectively. The DCA showed that the T2WI radiomic-clinical nomogram was better than other models with a greater clinical net benefit.

Conclusion

MR-based radiomics analysis showed the high accuracy in prognostic estimation of EOC patients and could help to predict therapeutic outcome before treatment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13048-021-00941-7.

Non-contrast-enhanced CT texture analysis of primary and metastatic pancreatic ductal adenocarcinomas: value in assessment of histopathological grade and differences between primary and metastatic lesions

PURPOSE

To evaluate the utility of non-contrast-enhanced CT texture analysis (CTTA) for predicting the histopathological differentiation of pancreatic ductal adenocarcinomas (PDAC) and to compare non-contrast-enhanced CTTA texture features between primary PDAC and hepatic metastases of PDAC.

METHODS

This retrospective study included 120 patients with histopathologically confirmed PDAC. Sixty-five patients underwent CT-guided biopsy of primary PDAC, while 55 patients underwent CT-guided biopsy of hepatic PDAC metastasis. All lesions were segmented in non-contrast-enhanced CT scans for CTTA based on histogram analysis, co-occurrence matrix, and run-length matrix. Statistical analysis was conducted for 372 texture features using Mann-Whitney U test, Bonferroni-Holm correction, and receiver operating characteristic (ROC) analysis. A p value < 0.05 was considered statistically significant.

RESULTS

Three features were identified that differed significantly between histopathological G2 and G3 primary tumors. Of these, "low gray-level zone emphasis" yielded the largest AUC (0.87 ± 0.04), reaching a sensitivity and specificity of 0.76 and 0.83, respectively, when a cut-off value of 0.482 was applied. Fifty-four features differed significantly between primary and hepatic metastatic PDAC.

CONCLUSION

Non-contrast-enhanced CTTA of PDAC identified differences in texture features between primary G2 and G3 tumors that could be used for non-invasive tumor assessment. Extensive differences between the features of primary and metastatic PDAC on CTTA suggest differences in tumor microenvironment.

Evaluation of the Risk of Recurrence in Patients with Local Advanced Rectal Tumours by Different Radiomic Analysis Approaches

The word radiomics, like all domains of type omics, assumes the existence of a large amount of data. Using artificial intelligence, in particular, different machine learning techniques, is a necessary step for better data exploitation. Classically, researchers in this field of radiomics have used conventional machine learning techniques (random forest, for example). More recently, deep learning, a subdomain of machine learning, has emerged. Its applications are increasing, and the results obtained so far have demonstrated their remarkable effectiveness. Several previous studies have explored the potential applications of radiomics in colorectal cancer. These potential applications can be grouped into several categories like evaluation of the reproducibility of texture data, prediction of response to treatment, prediction of the occurrence of metastases, and prediction of survival. Few studies, however, have explored the potential of radiomics in predicting recurrence-free survival. In this study, we evaluated and compared six conventional learning models and a deep learning model, based on MRI textural analysis of patients with locally advanced rectal tumours, correlated with the risk of recidivism; in traditional learning, we compared 2D image analysis models vs. 3D image analysis models, models based on a textural analysis of the tumour versus models taking into account the peritumoural environment in addition to the tumour itself. In deep learning, we built a 16-layer convolutional neural network model, driven by a 2D MRI image database comprising both the native images and the bounding box corresponding to each image.

Prediction of BRCA gene mutation status in epithelial ovarian cancer by radiomics models based on 2D and 3D CT images

Background

The objective of this study is to explore the value of two-dimensional (2D) and three-dimensional (3D) radiomics models based on enhanced computed tomography (CT) images in predicting BRCA gene mutations in patients with epithelial ovarian cancer.

Methods

The clinical and imaging data of 106 patients with ovarian cancer confirmed by surgery and pathology were retrospectively analyzed and genetic testing was performed. Radiomics features extracted from the 2D and 3D regions of interest of the patients’ primary tumor lesions were selected in the training set using the maximum correlation and minimum redundancy method. Then, the best features were selected through Lasso tenfold cross-validation. Feature subsets were employed to establish a radiomics model. The model’s performance was evaluated via area under the receiver operating characteristic curve analysis and its clinical validity was assessed by using the model’s decision curve.

Results

On the validation set, the area under the curve values of the 2D, 3D, and 2D + 3D combined models were 0.78 (0.61–0.96), 0.75 (0.55–0.92), and 0.82 (0.61–0.96), respectively. However, the DeLong test P values between the three pairs of models were all > 0.05. The decision curve analysis showed that the radiomics model had a high net benefit across all high-risk threshold probabilities.

Conclusions

The three radiomics models can predict the BRCA gene mutation in ovarian cancer, and there were no statistically significant differences between the prediction performance of the three models.

Diagnostic Performance of 2D and 3D T2WI-Based Radiomics Features With Machine Learning Algorithms to Distinguish Solid Solitary Pulmonary Lesion

OBJECTIVE

To evaluate the performance of 2D and 3D radiomics features with different machine learning approaches to classify SPLs based on magnetic resonance(MR) T2 weighted imaging (T2WI).

MATERIAL AND METHODS

A total of 132 patients with pathologically confirmed SPLs were examined and randomly divided into training (n = 92) and test datasets (n = 40). A total of 1692 3D and 1231 2D radiomics features per patient were extracted. Both radiomics features and clinical data were evaluated. A total of 1260 classification models, comprising 3 normalization methods, 2 dimension reduction algorithms, 3 feature selection methods, and 10 classifiers with 7 different feature numbers (confined to 3-9), were compared. The ten-fold cross-validation on the training dataset was applied to choose the candidate final model. The area under the receiver operating characteristic curve (AUC), precision-recall plot, and Matthews Correlation Coefficient were used to evaluate the performance of machine learning approaches.

RESULTS

The 3D features were significantly superior to 2D features, showing much more machine learning combinations with AUC greater than 0.7 in both validation and test groups (129 vs. 11). The feature selection method Analysis of Variance(ANOVA), Recursive Feature Elimination(RFE) and the classifier Logistic Regression(LR), Linear Discriminant Analysis(LDA), Support Vector Machine(SVM), Gaussian Process(GP) had relatively better performance. The best performance of 3D radiomics features in the test dataset (AUC = 0.824, AUC-PR = 0.927, MCC = 0.514) was higher than that of 2D features (AUC = 0.740, AUC-PR = 0.846, MCC = 0.404). The joint 3D and 2D features (AUC=0.813, AUC-PR = 0.926, MCC = 0.563) showed similar results as 3D features. Incorporating clinical features with 3D and 2D radiomics features slightly improved the AUC to 0.836 (AUC-PR = 0.918, MCC = 0.620) and 0.780 (AUC-PR = 0.900, MCC = 0.574), respectively.

CONCLUSIONS

After algorithm optimization, 2D feature-based radiomics models yield favorable results in differentiating malignant and benign SPLs, but 3D features are still preferred because of the availability of more machine learning algorithmic combinations with better performance. Feature selection methods ANOVA and RFE, and classifier LR, LDA, SVM and GP are more likely to demonstrate better diagnostic performance for 3D features in the current study.

Radiomics for Predicting Lung Cancer Outcomes Following Radiotherapy: A Systematic Review

Lung cancer's radiomic phenotype may potentially inform clinical decision-making with respect to radical radiotherapy. At present there are no validated biomarkers available for the individualisation of radical radiotherapy in lung cancer and the mortality rate of this disease remains the highest of all other solid tumours. MEDLINE was searched using the terms 'radiomics' and 'lung cancer' according to the Preferred Reporting Items for Systematic Reviews and Met-Analyses (PRISMA) guidance. Radiomics studies were defined as those manuscripts describing the extraction and analysis of at least 10 quantifiable imaging features. Only those studies assessing disease control, survival or toxicity outcomes for patients with lung cancer following radical radiotherapy ± chemotherapy were included. Study titles and abstracts were reviewed by two independent reviewers. The Radiomics Quality Score was applied to the full text of included papers. Of 244 returned results, 44 studies met the eligibility criteria for inclusion. End points frequently reported were local (17%), regional (17%) and distant control (31%), overall survival (79%) and pulmonary toxicity (4%). Imaging features strongly associated with clinical outcomes include texture features belonging to the subclasses Gray level run length matrix, Gray level co-occurrence matrix and kurtosis. The median cohort size for model development was 100 (15-645); in the 11 studies with external validation in a separate independent population, the median cohort size was 84 (21-295). The median number of imaging features extracted was 184 (10-6538). The median Radiomics Quality Score was 11% (0-47). Patient-reported outcomes were not incorporated within any studies identified. No studies externally validated a radiomics signature in a registered prospective study. Imaging-derived indices attained through radiomic analyses could equip thoracic oncologists with biomarkers for treatment response, patterns of failure, normal tissue toxicity and survival in lung cancer. Based on routine scans, their non-invasive nature and cost-effectiveness are major advantages over conventional pathological assessment. Improved tools are required for the appraisal of radiomics studies, as significant barriers to clinical implementation remain, such as standardisation of input scan data, quality of reporting and external validation of signatures in randomised, interventional clinical trials.

Radiomic Signature as a Diagnostic Factor for Classification of Histologic Subtypes of Lung Cancer

<sec> <title>Objectives:</title> To discuss the application of radiomics using Computerized Tomography (CT) analysis, for improving its diagnostic efficacy in lung, specifically in distinguishing Squamous Cell Carcinoma (SCC), lung Adenocarcinoma (ADC), and Small Cell Lung Cancer (SCLC). </sec> <sec> <title>Methods:</title> The pathology of 189 identified cases of lung cancer was analyzed, retrospectively (60 patients with SCC, 69 patients with lung ADC and 60 patients with SCLC). A neural network was used to determine whether the pulmonary or mediastinal window was selected to extract effective radiomic features. The key features of radiomic signature were retrieved by a Least Absolute Shrinkage and Selection Operator (LASSO) multiple logistic regression model. Next, receiver operating characteristic curve and Area Under the Curve (AUC) analysis were used to evaluate the performance of the radiomic signature in both, training(129 patients) and validation cohorts (60 patients). </sec> <sec> <title>Results:</title> About 295 features were extracted from a manually outlined tumor region. Features extracted from mediastinal window CT scans had a better prognostic ability than pulmonary window scans. The average accuracy for mediastinal window scans was 0.933. Our analysis revealed that the radiomic features extracted from mediastinal window scans had the potential to build a prediction model for distinguishing between SCC, lung ADC, and SCLC. The performance of the radiomic signature to diagnose SCC and SCLC in validation cohorts proved effective, with AUC values of 0.869 and 0.859, respectively. </sec> <sec> <title>Conclusions:</title> A unique radiomic signature was constructed as a diagnostic factor for different histologic subtypes of lung cancer. Patients with lung cancer may benefit from this proposed radiomic signature. </sec>

Robustness and performance of radiomic features in diagnosing cystic renal masses

PURPOSE

We study the inter-reader variability in manual delineation of cystic renal masses (CRMs) presented in computerized tomography (CT) images and its effect on the classification performance of a machine learning algorithm in distinguishing benign from potentially malignant CRMs. In addition, we assessed whether the inclusion of higher-order robust radiomic features improves the classification performance over the use of first-order features.

METHODS

230 CRMs were independently delineated by two radiologists. Through a combination of random fluctuations, dilation, and erosion operations over the original region of interests (ROIs), we generated four additional sets of synthetic ROIs to capture the inter-reader variability realistically, as confirmed by dice coefficient measurements and visual assessment. We then identified the robust features based on the intra-class coefficient (ICC > 0.85) across these datasets. We applied a tenfold stratified cross-validation (CV) to train and test the performance of the random forest model for the classification of CRMs into benign and potentially malignant.

RESULTS

The mean area under the curve (AUC), sensitivity, specificity, positive predictive value, and negative predictive value were 0.87, 0.82, 0.90, 0.85, and 0.93, respectively. With the usage of first-order features alone, the corresponding values were nearly identical.

CONCLUSION

AUC ranged for the robust and uncorrelated features from 0.83 ± 0.09 to 0.93 ± 0.04 and for the first-order features from 0.84 ± 0.09 to 0.91 ± 0.04. Our study indicates that the first-order features alone are sufficient for the classification of CRMs, and that inclusion of higher-order features does not necessarily improve performance.

Predicting response to immunotherapy plus chemotherapy in patients with esophageal squamous cell carcinoma using non-invasive Radiomic biomarkers

Objectives

To develop and validate a radiomics model for evaluating treatment response to immune-checkpoint inhibitor plus chemotherapy (ICI + CT) in patients with advanced esophageal squamous cell carcinoma (ESCC).

Methods

A total of 64 patients with advance ESCC receiving first-line ICI + CT at two centers between January 2019 and June 2020 were enrolled in this study. Both 2D ROIs and 3D ROIs were segmented. ComBat correction was applied to minimize the potential bias on the results due to different scan protocols. A total of 788 features were extracted and radiomics models were built on corrected/uncorrected 2D and 3D features by using 5-fold cross-validation. The performance of the radiomics models was assessed by its discrimination, calibration and clinical usefulness with independent validation.

Results

Five features and support vector machine algorithm were selected to build the 2D uncorrected, 2D corrected, 3D uncorrected and 3D corrected radiomics models. The 2D radiomics models significantly outperformed the 3D radiomics models in both primary and validation cohorts. When ComBat correction was used, the performance of 2D models was better (p = 0.0059) in the training cohort, and significantly better (p < 0.0001) in the validation cohort. The 2D corrected radiomics model yielded the optimal performance and was used to build the nomogram. The calibration curve of the radiomics model demonstrated good agreement between prediction and observation and the decision curve analysis confirmed the clinical utility.

Conclusions

The easy-to-use 2D corrected radiomics model could facilitate noninvasive preselection of ESCC patients who would benefit from ICI + CT.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-021-08899-x.