Prediction of Preoperative Synchronous Distant Metastasis of Rectal Cancer Based on MRI Radiomics Model

Prediction of Preoperative Synchronous Distant Metastasis of Rectal Cancer Based on MRI Radiomics Model | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Prediction of Preoperative Synchronous Distant Metastasis of Rectal Cancer Based on MRI Radiomics Model Hao Jiang, Wei Guo, Xue Lin, Zhuo Yu, Yudie Qin, Zhongqi Sun, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5041812/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose The objective of this study was to develop and validate a new non-invasive artificial intelligence (AI) model based on preoperative magnetic resonance imaging (MRI) data to predict the presence of synchronous distant metastasis (SDM) in rectal cancer (RC). Methods 169 eligible RC patients were enrolled, and T2WI and DWI sequence images were collected. The radiomics features were extracted through the PyRadiomics package of Python language, and a total of 1688 radiomics features were extracted, including first-order features, shape features, texture features, and Baud signs. One clinical model and three comprehensive models of clinical imaging were constructed. Five indexes including receiver operating characteristic (ROC), area under curve (AUC), accuracy, sensitivity, specificity, and 95% confidence interval (CI) were selected to evaluate the model. The clinical model using four independent risk factors (CEA, age, CA199, and T stage). Combining the clinical factors and imaging characteristics of different sequences, we established three clinically-imaging models: the DWI + clinical model, the T2W + clinical model, and the nomogram (radiomics + clinical) model. Results This nomogram model performed the best in predicting rectal cancer SDM. In the training set, the AUC, accuracy, sensitivity, specificity and 95%CI of the nomogram model were 0.93, 0.85, 0.85, 0.86, 0.89–0.96, respectively. In the test set, five indexes of the nomogram model were 0.94, 0.89, 0.88, 0.89, and 0.79 ~ 0.97, respectively. The correction plots were consistent between the predictions of the clinical radiomics model and the actual observed probabilities. Decision curve analysis showed that the nomogram model achieved the highest net benefit on the training set and the test set compared to the clinical model and the radiomics model. Conclusion Our predictive model is valuable for guiding and managing patients with rectal cancer SDM, providing options for improving patient treatment decisions and guiding personalized treatment regimens. magnetic resonance imaging rectal cancer synchronous distant metastasis radiomics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 INTRODUCTION Rectal cancer is the third most common malignant cause of morbidity and mortality. Despite total mesorectal excision and neoadjuvant chemoradiotherapy, the local recurrence rate of rectal cancer has been significantly reduced to 5–10%[ 1 ]. Distant metastasis remains the main cause of treatment failure in patients with rectal cancer [ 2 , 3 ]. In those metastatic cases, the liver is the most frequently involved organ, followed by the lung [ 4 ]. Surgical excision is the primary treatment strategy for early detection of metastasis. It has a better prognosis and survival rate compared to other treatments, offering these patients a chance of cure. For localized colorectal liver or lung metastases, resection has been considered the treatment of choice for improving longterm survival. The 5-year survival rates for patients treated with surgical resection of colorectal liver or lung metastasis can be enhanced to 58.0% or 56.2%, respectively[ 5 , 6 ]. Therefore, preoperative diagnosis of colorectal cancer patients at high risk for SDM is essential for personalized treatment strategies. The traditional standard in China defines simultaneous metastasis as colorectal cancer found at the time of diagnosis or metastasis occurring within 6 months after radical resection of the primary focus of colorectal cancer. However, some indicators are only available after radical resection and cannot be used as a basis for preoperative treatment strategies. Therefore, developing a preoperative, non-invasive, and accurate method to predict SDM is necessary. Radiomics uses automated high-throughput extraction techniques for many quantitative features and can capture intra-tumor heterogeneity in a non-invasive manner, so it can be used in personalized medicine. Tumor heterogeneity can be reflected in imaging, creating the opportunity to identify imaging biomarkers that correlate with the tumor’s biological behavior. In this context, radiomics can play a key role, providing minable data from standard radiological images and exploring quantitative features that can describe tumor heterogeneity and other intrinsic characteristics that could correlate with its biological behavior[ 7 ]. Radiomics models based on computed tomography (CT)/MRI have been reported to predict metastatic lymph nodes and distant metastases of lung adenocarcinomas in bladder cancer [ 8 ] and breast cancer [ 9 – 10 ]. Furthermore, some studies have used the image analysis approach to identify the association between tumor tissue and distant metastases in different malignancies [ 11 , 12 ]. Specifically, Chen et al. showed that a CT-based radiomics model demonstrated good performance in the prediction of brain metastasis in lung cancer patients [ 13 ]. Other studies also reported that radiomics signature has the potential to predict liver metastases (LM) in rectal and esophagogastric cancer [ 14 , 15 ]. In addition, the imaging features of primary rectal cancer are often less affected and more stable than those of metastases. However, few studies have been carried out to predict SDM by the imaging features of primary rectal cancer, which is worthy of further study. Gaitanidis et al [ 16 ] recently demonstrated the feasibility of predictive nomograms for evaluating the probability of synchronous liver, lung, and bone-distant disease in 46,785 rectal cancer patients. The results were promising with clinical and pathologic features in the proposed nomograms, yet the pathologic information is available only after surgery, which cannot be used to guide preoperative treatment strategy. Several studies have shown that radiomics models can predict distant metastases of different primary tumors [ 12 , 17 , 18 ]. However, the role of radiomics nomograms from primary lesions in predicting SDM in RC patients has not been clearly defined. Therefore, it is necessary to develop preoperative non-invasive biomarkers to predict SDM. MRI was the noninvasive imaging modality of choice for preoperative rectal cancer staging. It can also provide more than just morphological information, as images are data more than pictures. For rectal cancer, the application of radiomics mainly focused on the treatment response to chemoradiotherapy [ 19 – 21 ]. For the applications of radiomics in colorectal cancer, Huang et al. [ 22 ] reported that a CT-based radiomics nomogram facilitated the prediction of malignant lymph nodes in colorectal cancer. Other studies [ 19 , 21 ] on rectal cancer have also indicated that the MRI radiomics signature might be used as an effective biomarker for the prediction of long-term outcomes. However, in these rectal cancer radiomics studies, there are few studies on distant synchronous metastasis of rectal cancer. MRI is well-established in the local staging of rectal cancer because of its superior efficacy compared to CT [ 23 ]. In this study, we aimed to evaluate the value of MRI radiology based on T2-weighted imaging (T2WI) and DWI imaging in pre-operative identification of rectal cancer patients at high SDM risk and to establish a predictive clinical imaging combination model that can help improve decision-making and guide personalized treatment. MATERIALS AND METHODS Patient Characteristics This retrospective study was approved by the Ethics Committee of the ××× Hospital. We reviewed clinicopathological and imaging data collected from 465 RC patients confirmed by endoscopic biopsy or surgical pathology between January 2013 and December 2020. Including rectal cancer MRI examination, chest CT and abdominal CT contrast-enhanced examination, age, gender, pre-treated Carcinoembryonic antigen (CEA) data, pre-treated Carbohydrate antigen 199 (CA199) imaging, and clinical data, tumor size, tumor location, T stage and N stage predicted by MR. SDM inclusion criteria were as follows: (a) Patients with a pathological diagnosis of RC, (b) there is no prior history or co-existing history of other malignancies, (c) Patients undergoing preoperative high-resolution rectal MRI, chest CT, and contrast-enhanced abdominal CT, (d) Regular review and follow-up imaging data in our hospital were complete. Of these patients, SDM excluded based on the following exclusion criteria: (a) Patients who have received treatment (radiotherapy, chemotherapy, or chemoradiotherapy) before the MRI examination; (b) Patients whose images are of poor quality and cannot be analyzed; (c) Lack of clinical data; (d) Patients with mucinous adenocarcinoma. Finally, 169 patients were included according to the above criteria and were randomly divided into the training dataset (n = 134) and the test dataset (n = 35) with a segmentation ratio of 8:2. The detailed selection process is shown in Fig. 1 . MRI Acquisition: All patients underwent rectal MRI using the GE Discovery MR750w 3.0T MRI scanner in the supine position using a 32-channel phased array human coil. Take a low-residue diet one day before the MRI examination, and fast for 4–6 hours before the MRI examination. A 60 mL enema was performed 1 h before the examination to better observe the primary tumor. Rectal dilation with air or water was not performed, and spasmolytics were not provided. The MRI scan sequences included: (1) T2-weighted images (T2WI) in high-resolution oblique (perpendicular to the long axis of the tumor; (2) Sagittal and coronal plane (parallel to the long axis of the rectum) T2WI spin echoes; (3) T1-weighted image (T1WI); (4) diffusion-weighted image (DWI). All patients also underwent contra-enhanced CT imaging of the chest and abdomen using a dual-source 64-MDCT (Somatom Definition Flash, Siemens Healthineers) scanner. Preparation and Scanning Methods Before the Examination Before the examination, confirm that the patient is stable and free of contraindications (no metal implants, no claustrophobia, etc). Explain to the patients before the examination and patients’ informed consent. At least half an hour before the examination, the bowel was cleaned without air or any contrast agent. Radiomic Feature Extraction Two radiologists (with > 5 years of experience in abdominal radiology diagnosis) examined all MRI data images and evaluated the T and N stages as defined by the American Cancer Federation's TN Staging System for Rectal Cancer, Version 8. The first radiologist manually mapped the tumor area of interest (ROI), and the second radiologist validated the ROI for the tumor tissue. Two other radiologists (with > 5 years of experience in chest and abdominal radiology diagnosis) consistently analyzed chest and abdominal CT images for distant metastasis assessment. All radiologists knew that these tumors were biopsies-confirmed adenocarcinomas of the rectum, but they did not know the histopathological stage of the patient. When observers cannot reach a consensus, ask another experienced radiologist (with > 10 years of experience in chest and abdominal diagnosis) to seek a final opinion. Radiomics all work in the ××× medical research platform to complete. The radiomics workflow includes four steps: data collection and segmentation, radiomics feature extraction, feature screening, model building, and effect evaluation. Region of interest (ROIs) was segmented by doctors with more than 10 years of clinical experience on the ITK-SNAP software. The extraction of radiomics features are extracted through the PyRadiomics package of the Python language. For each image phase, 1688 radiomics features were extracted, including first-order features, shape features, texture features, wavelet features from 10 image types, which are original image, Laplacian of Gaussian (LoG, based on SimpleITK function), Wavelet ( Using the PyWavelets package), Square, Square Root, Logarithm, Exponential, Gradient, Local Binary Pattern (2D), Local Binary Pattern (3D). Radiomics Signature Construction We sequentially use Variance Threshold, SelectKBest, and Least Absolute Shrinkage and Selection Operator (LASSO) for feature screening. The variance threshold method is used to eliminate features with small variance (< 0.8) because features with small variance have small changes in value and have little effect on the result identification. For the variance threshold method, values with variances smaller than the variance threshold (0.8) are removed. The SelectKBest method is based on the chi-square test leaving the features of P < 0.05. The chi-square test is a method for the degree of deviation between the actual observed value of the statistical sample and the theoretically inferred value. The LASSO method obtains the sparseness of the feature weight The matrix realizes the screening of features. This topic uses the L2 regularizer as the loss function. Finally, a radiomics model was constructed based on Support Vector Machines (SVM). According to the above process, the features were extracted from different phases of images and the pooling features of the two phases were screened, and then based on these screened features, the radiomics models of different phases and the fusion phases were constructed. To evaluate the performance of these models, five indicators of receiver operating characteristic (ROC), area under the curve (AUC), accuracy, sensitivity, specificity, and 95% confidence interval (CI) were selected for evaluation. Development of the Clinical Model and the Nomogram To construct the clinical model, we used univariate analysis to identify independent risk factors between the NR group and the R group. Clinical models were built using significant factors ( P < 0.05) from univariate analysis. The clinical-radiomics nomogram was built using the selected pooling features of the two phases and independent clinical-radiologic risk factors in the clinical model. The fusion characteristics of DWI and T2WI were combined with clinically independent risk factors for synchronous liver metastases of rectal cancer for Logistic regression analysis, and the fusion model was established and a nomogram of the fusion model was mapped to provide a quantitative guidance tool for clinical diagnosis and prediction of synchronous liver metastases of rectal cancer. The receiver operating characteristic curve (ROC) was drawn, and the diagnostic effect of the three models was compared by the Delong test. In addition, a calibration curve test was used to assess the consistency between the predicted risks of the model and the actual risks. The statistical analysis method of decision curve analysis (DCA) was used to verify the clinical effectiveness of the model。 Experimental Environment All experiments are completed on the PyCharm program, the programming language used is Python (v. 3.6), mainly using sklearn, numpy, scipy, pandas, and other packages. Statistical Analysis Normal test was performed on all measurement data, mean ± standard deviation (SD) was used to represent consistent with normal distribution, and median was used to represent inconsistent with normal distribution. Independent sample t test or Mann-Whitney U test were used for measurement data of two groups with metastasis and non-metastasis. The statistical data were expressed by example (%), chi square test was used between the two groups, the theoretical frequency was less than 5, and Fisher’s exact test was used to test. ROC curves were drawn, and the area under the curve (AUC) was calculated to assess the diagnostic performance of each model. The DeLong test was used to compare the AUCs of the different classification models. Decision curve analysis (DCA) was performed to compare the differences in net benefits between the different models at the threshold probability. SPSS 26.0 statistical software was used to analyze the data. The reported statistical significance levels were all two sided, with the statistical significance level set at 0.05. Results Patient Characteristics Baseline clinical characteristics are shown in Table 1 . There is no significant difference between the data distribution of the training set and the test set (69/65 vs 18/17, P = 0.900). On the training set and the test set, the clinical factors that have significant differences at the same time are ( P ), age ( P < 0.05), CA199( P < 0.05), and T stage( P < 0.05). Tumor location was only significantly different on the test set. Other clinical variables were not significantly different on the training and test sets. Table 1 Characteristics of patients and associations with synchronous distant metastasis Characteristics Non-metastasis (n = 87) Metastasis (n = 82) χ 2 /t P value Age, mean ± SD, years 57.9 ± 11.8 62.3 ± 8.4 -2.82 0.005 Gender (%) 0.85 0.356 Male 62 (71.3) 53 (64.6) Female 25 (28.7) 29 (35.4) CEA, µg/L 42.72 0.000 < 5 69 (79.3) 24 (29.3) ≥ 5 18 (20.7) 58 (70.7) CA199, U/ml 22.92 0.000 < 37 78(89.7) 47(57.3) ≥ 37 9(10.3) 35(42.7) Tumor diameter, mean ± SD, cm 4.3 ± 1.6 4.6 ± 1.2 -1.73 0.086 Tumor location (%) 2.55 0.279 Proximal rectum 31(35.6) 23(28.1) Middle rectum 50(57.5) 48(58.5) Distal rectum 6(6.9) 11(13.4) Degree of differentiation (%) 1.39 0.498 Well 7(8.0) 10(12.2) Moderate 74(85.1) 64(78.0) Poor 6(6.9) 8(9.8) T staging (%) 4.23 0.040 T1-2 14(16.1) 5(6.1) T3-4 73(83.9) 77(93.9) N staging (%) 2.85 0.093 N0 23(26.4) 13(15.9) N1-2 64(73.6) 69(84.1) CEA carcinoembryonic antigen, CA199 carbohydrate antigen 199 P < 0.05 indicates a statistically significant difference Development of the Radiomics Model 1688 radiomics features were extracted from each imaging phase. After feature screening, DWI phase images retained 8 features; T2W phase images retained 8 features; Radiomics model(DWI phase and T2W phase images) retained 6 features, among which, there are 4 features of DWI phase images and 2 features of T2W phase images. In terms of the retained characteristics of each phase, the SVM model was selected to construct the radiomics model of each phase. The experimental results show that none of radiomics models has a good discriminative ability for SDM(Table 2 , Fig. 2 ). On the training set, the DWI model performed best, and its AUC, accuracy, sensitivity, specificity, and 95%CI were 0.88, 0.82, 0.55, 0.92, 0.84–0.92, respectively; on the test set, the Radiomics model performed the best, and its AUC, accuracy, sensitivity, specificity, and 95%CI are 0.82, 0.71, 0.76, 0.67, 0.62–0.86, respectively. Table 2 Performance of models based on different phase images. Classifiers Train set Test set AUC ACC SEN SPE 95%CI AUC ACC SEN SPE 95%CI DWI model 0.88 0.82 0.55 0.92 0.84–0.92 0.81 0.71 0.76 0.67 0.59–0.83 T2W model 0.52 0.73 0.05 0.99 0.5–0.55 0.68 0.6 0.59 0.61 0.44–0.73 Radiomics model 0.87 0.81 0.48 0.94 0.83–0.9 0.82 0.71 0.76 0.67 0.62–0.86 Development of the Clinical Model and Nomogram The clinical model developed using four independent risk factors (CEA, age, CA199, and T stage) showed AUCs of 0.81 and 0.83 in the training and test datasets, respectively (Fig. 4 A, 4 B). Combining clinical factors and radiomics characteristics of different sequence images, we established three clinical-radiomics models, namely the DWI + clinical model, the T2W + clinical model, and the nomogram (radiomics + clinical) model (Fig. 3 ). In terms of indicators, the nomogram model has the best discriminative performance on SDM(Table 3 , Fig. 4 A, 4 B). On the training set, the AUC, accuracy, sensitivity, specificity, and 95% CI of the nomogram model are 0.93, respectively. 0.85, 0.85, 0.86, 0.89–0.96. On the test set, the AUC, accuracy, sensitivity, specificity, and 95% CI of the nomogram model are 0.94, 0.89, 0.88, 0.89, and 0.79–0.97, respectively. The calibration plots (Fig. 4 C, 4 D) were consistent between the clinical radiomics model predicted and observed probabilities. Decision curve analysis (Fig. 5 ) showed that the nomogram model achieved the highest net benefit compared with the clinical model and the radiomics model on the training and test sets. Table 3 Performance of radiomics model, clinical model, DWI + clinical model, T2W + clinical model, and nomogram. Classifiers Train set Test set AUC ACC SEN SPE 95%CI AUC ACC SEN SPE 95%CI Radiomics model 0.77 0.66 0.66 0.67 0.7–0.83 0.82 0.71 0.76 0.67 0.62–0.86 Clinical model 0.81 0.7 0.69 0.71 0.74–0.86 0.83 0.83 0.82 0.83 0.72–0.92 DWI + clinical model 0.94 0.87 0.88 0.87 0.9–0.97 0.86 0.8 0.71 0.89 0.68–0.89 T2W + clinical model 0.95 0.88 0.88 0.88 0.91–0.98 0.83 0.74 0.82 0.67 0.62–0.86 Nomogram model 0.93 0.85 0.85 0.86 0.89–0.96 0.94 0.89 0.88 0.89 0.79–0.97 Discussion Our previous experience confirmed that MRI staging-based radiomics models provide a relevant predictive tool to identify the oncological behavior of pCR after nCRT [ 24 ]. Despite the important effort made in terms of treatment response prediction, few experiences reported the relation between radiomics predictors and early distant recurrence [ 25 , 26 ]. Our results show that adding radiomics features to the clinical model can better predict SDM performance in rectal cancer patients, increasing the AUC from 0.83 to 0.94, with high sensitivity and specificity in the validation cohort. The high specificity indicates that the model is reliable and can eliminate more false positive and false negative patients. We developed a clinical-radiomics nomogram as an individual and visualization tool to provide an estimated probability of SDM for newly diagnosed rectal cancer patients. Decision curve analysis was used to determine its clinical benefit. Radiomics, integrating many high-dimensional imaging features used to quantify tumor heterogeneity, could facilitate oncologic diagnosis and prognosis prediction. Huang et al [ 22 ] revealed that radiomics signature including 24 selected features could help predict LN-positive patients with a C-index of 0.773 in the validation cohort and the proposed clinical-radiomics nomogram was useful for predicting LN involvement. In our study, radiomics features were obtained using high-resolution axial T2WI and DWI, which are the most critical sequences for evaluating primary rectal cancer[ 27 ]. Functional imaging techniques such as DWI can analyze lesions from the perspective of the microenvironment and molecular pathology, to quantitatively study many small lesions that cannot be quantified and identified by the naked eye. Our proposed clinical-radiomics predictive model confirms the feasibility of imaging analysis based on T2WI and DWI, providing a potentially effective and easy-to-use model in clinical practice. Using our clinical-radiomics nomogram, an estimated probability of SDM could be calculated after referring to the selected T2WI-based radiomics features, DWI-based radiomics features as well as other clinical information. Our results show that the radiomics nomogram provides predictive information about SDM in primary rectal cancer. Contrast-enhanced CT, MRI, and PET-CT are common imaging examinations for the diagnosis of SLM in RC preoperatively. However, the sensitivity and accuracy of these imaging techniques are not satisfactory[ 28 – 30 ]. According to one meta-analysis, the detection sensitivity of colorectal LM in contrast-enhanced CT, routine MRI, and FDG PET-CT were 63%-80%, 76%-85.7%, and 51%-90%, respectively[ 31 ]. Many studies have shown that MRI had a higher accuracy compared to CT in diagnosing SLM of rectal cancer patients[ 32 – 34 ], and recent consensus guidelines from the radiologic community recommend MRI for the preoperative evaluation of SLM [ 35 , 36 ]. The other studies had shown that some adverse features found on rectal MRI identified patients with rectal cancer at higher risk of distant metastasis[ 37 – 39 ]. Therefore, we developed a radiomics nomogram based on multi-sequence MRI to predict the risk of SDM. Therefore, screening for high-risk predictors would improve the probability of early detection of SDM in RC patients. Typically, the clinicopathological predictors of SDM in RC patients include the histological type, pathological grade, depth of tumor invasion, lymph node status, vascular invasion, and tumor markers[ 40 ]. However, some of these predictors can only be obtained postoperatively, and hence, inappropriate to guide preoperative treatment. Other studies have demonstrated that some features of rectal MRI, such as extramural vascular invasion, higher T stage, and regional lymph node metastasis are potential predictors[ 38 , 39 ]. However, these image features are subjective and qualitative, lacking quantitative assessment. In recent years, radiomics has been regarded as an advanced tool for evaluating tumor heterogeneity in tumor diagnosis and prognosis prediction. In this study, factors such as radiomics features, CEA, and CA19-9 levels were incorporated into multivariate logistic regression to build a prediction model and a nomogram, and the research results are promising. Therefore, our analysis indicates that the radiomics nomogram combined with tumor markers was superior to the radiomics signature alone. It exhibited a high predictive performance for SDM in RC patients, and the AUC improved from 0.82 to 0.94. Moreover, the results were better than those reported in a previous study on a per-patient basis, wherein the AUC was 0.92 and 0.88 (MRI readers), 0.80 and 0.82 (CT readers), and an AUC of 0.83 and 0.84 (PET-CT readers)[ 41 ]. In this study, we constructed a primary RC-based radiomics nomogram from high-resolution T2WI and DWI of the recta to predict SDM. On the training set, the AUC, accuracy, sensitivity, specificity, and 95%CI of the nomogram model were 0.93, 0.85, 0.85, 0.86, and 0.89–0.96, respectively. The AUC, accuracy, sensitivity, specificity, and 95% CI of the nomogram model were 0.94, 0.89, 0.89, and 0.79 ~ 0.97 respectively. Therefore, based on clinical risk factors and radiomics characteristics, the proposed nomogram may be a valuable predictive tool for SDM in RC patients. It can be easily used to identify patients who need further whole-body imaging. Radiomics features are composed of multiple features, which is of great significance for selecting the optimal feature collection through dimensionality reduction. In this study, we used the LASSO method for feature screening and built a radiomics model based on a support vector machine (SVM). In our study, the method of noose dimension reduction is chosen to increase the stability of the nomogram and to carry out the overall analysis. The nomogram we constructed is easier for clinicians to use because it can derive information from T2WI, and DWI, and includes clinical risk factors. The radiomics nomogram is an auxiliary tool that can be used to identify and follow those patients with rectal cancer. When the group was divided into a low-risk group and a high-risk group, the high-risk group had a higher probability of developing SDM. Therefore, in a certain sense, the radiomics nomogram can be used as an accurate and reliable detection tool for SDM in rectal cancer patients. It is quick and easy to perform and helps to determine which patients will benefit from further imaging of distant metastases. However, the current study still has some limitations. First of all, the sample size based on single-institution retrospective analysis is relatively small, and selection bias is inevitable. Second, this study lacks external validation, so a large multicenter trial is needed to improve the generalizability of the results. In summary, we developed a clinical imaging nomogram by combining tumor markers with imaging features to accurately predict the presence of SDM in RC patients. This visualization tool will detect the probability of SDM and help doctors make clinical decisions. Declarations Author Contribution H.J. and W.G. carried out the studies, participated in collecting data, and drafted the manuscript. H.J. and Z.Y. performed the statistical analysis and participated in its design. Y.D.Q.、Z.Q.S. and X.L. participated in the acquisition and analysis of data. H.J.J. designed the study and carefully revised the manuscript. H.B.H.、J.P.L.、Q.W. and L.H.Z. examined the manuscript carefully. All authors read and approved the final manuscript. References Bosset JF, Collette L, Calais G et al (2006) Chemotherapy with preoperative radiotherapy in rectal cancer. N Engl J Med 355(11):1114–1123. 10.1056/NEJMoa060829 Ceelen W, Fierens K, Van Nieuwenhove Y, Pattyn P (2009) Preoperative chemoradiation versus radiation alone for stage II and III resectable rectal cancer: a systematic review and meta-analysis. Int J Cancer 124(12):2966–2972. 10.1002/ijc.24247 Rodel C, Liersch T, Becker H et al (2012) Preoperative chemoradiotherapy and postoperative chemotherapy with fluorouracil and oxaliplatin versus fluorouracil alone in locally advanced rectal cancer: initial results of the German CAO/ARO/AIO-04 randomised phase 3 trial. Lancet Oncol 13(7):679–687. 10.1016/S1470-2045(12)70187-0 van der Geest LG, Lam-Boer J, Koopman M, Verhoef C, Elferink MA, de Wilt JH (2015) Nationwide trends in incidence, treatment and survival of colorectal cancer patients with synchronous metastases. Clin Exp Metastasis 32(5):457–465. 10.1007/s10585-015-9719-0 Hur H, Ko YT, Min BS et al (2009) Comparative study of resection and radiofrequency ablation in the treatment of solitary colorectal liver metastases. Am J Surg 197(6):728–736. 10.1016/j.amjsurg.2008.04.013 Kanas GP, Taylor A, Primrose JN et al (2012) Survival after liver resection in metastatic colorectal cancer: review and meta-analysis of prognostic factors. Clin Epidemiol 4:283–301. 10.2147/CLEP.S34285 O'Connor JPB (2017) Cancer heterogeneity and imaging. Semin Cell Dev Biol 64:48–57. 10.1016/j.semcdb.2016.10.001 Wu S, Zheng J, Li Y et al (2017) A Radiomics Nomogram for the Preoperative Prediction of Lymph Node Metastasis in Bladder Cancer. Clin Cancer Res 23(22):6904–6911. 10.1158/1078-0432.CCR-17-1510 Dong Y, Feng Q, Yang W et al (2018) Preoperative prediction of sentinel lymph node metastasis in breast cancer based on radiomics of T2-weighted fat-suppression and diffusion-weighted MRI. Eur Radiol 28(2):582–591. 10.1007/s00330-017-5005-7 Coroller TP, Grossmann P, Hou Y et al (2015) CT-based radiomic signature predicts distant metastasis in lung adenocarcinoma. Radiother Oncol 114(3):345–350. 10.1016/j.radonc.2015.02.015 Liu H, Zhang C, Wang L et al (2019) MRI radiomics analysis for predicting preoperative synchronous distant metastasis in patients with rectal cancer. Eur Radiol 29(8):4418–4426. 10.1007/s00330-018-5802-7 Zhang L, Dong D, Li H et al (2019) Development and validation of a magnetic resonance imaging-based model for the prediction of distant metastasis before initial treatment of nasopharyngeal carcinoma: A retrospective cohort study. EBioMedicine 40:327–335. 10.1016/j.ebiom.2019.01.013 Chen A, Lu L, Pu X et al (2019) CT-Based Radiomics Model for Predicting Brain Metastasis in Category T1 Lung Adenocarcinoma. AJR Am J Roentgenol 213(1):134–139. 10.2214/AJR.18.20591 Liang M, Cai Z, Zhang H et al (2019) Machine Learning-based Analysis of Rectal Cancer MRI Radiomics for Prediction of Metachronous Liver Metastasis. Acad Radiol 26(11):1495–1504. 10.1016/j.acra.2018.12.019 Klaassen R, Larue R, Mearadji B et al (2018) Feasibility of CT radiomics to predict treatment response of individual liver metastases in esophagogastric cancer patients. PLoS ONE 13(11):e0207362. 10.1371/journal.pone.0207362 Gaitanidis A, Alevizakos M, Tsaroucha A, Tsalikidis C, Pitiakoudis M (2018) Predictive Nomograms for Synchronous Distant Metastasis in Rectal Cancer. J Gastrointest Surg 22(7):1268–1276. 10.1007/s11605-018-3767-0 Li Y, Eresen A, Shangguan J et al (2019) Establishment of a new non-invasive imaging prediction model for liver metastasis in colon cancer. Am J Cancer Res 9(11):2482–2492 Fan L, Fang M, Tu W et al (2019) Radiomics Signature: A Biomarker for the Preoperative Distant Metastatic Prediction of Stage I Nonsmall Cell Lung Cancer. Acad Radiol 26(9):1253–1261. 10.1016/j.acra.2018.11.004 Meng Y, Zhang Y, Dong D et al (2018) Novel radiomic signature as a prognostic biomarker for locally advanced rectal cancer. J Magn Reson Imaging. 10.1002/jmri.25968 Horvat N, Veeraraghavan H, Khan M et al (2018) MR Imaging of Rectal Cancer: Radiomics Analysis to Assess Treatment Response after Neoadjuvant Therapy. Radiology 287(3):833–843. 10.1148/radiol.2018172300 Nie K, Shi L, Chen Q et al (2016) Rectal Cancer: Assessment of Neoadjuvant Chemoradiation Outcome based on Radiomics of Multiparametric MRI. Clin Cancer Res 22(21):5256–5264. 10.1158/1078-0432.CCR-15-2997 Huang YQ, Liang CH, He L et al (2016) Development and Validation of a Radiomics Nomogram for Preoperative Prediction of Lymph Node Metastasis in Colorectal Cancer. J Clin Oncol 34(18):2157–2164. 10.1200/JCO.2015.65.9128 Maizlin ZV, Brown JA, So G et al (2010) Can CT replace MRI in preoperative assessment of the circumferential resection margin in rectal cancer? Dis Colon Rectum 53(3):308–314. 10.1007/DCR.0b013e3181c5321e Jiang H, Guo W, Yu Z et al (2023) A Comprehensive Prediction Model Based on MRI Radiomics and Clinical Factors to Predict Tumor Response After Neoadjuvant Chemoradiotherapy in Rectal Cancer. Acad Radiol 30(Suppl 1):S185–S98. 10.1016/j.acra.2023.04.032 Dinapoli N, Barbaro B, Gatta R et al (2018) Magnetic Resonance, Vendor-independent, Intensity Histogram Analysis Predicting Pathologic Complete Response After Radiochemotherapy of Rectal Cancer. Int J Radiat Oncol Biol Phys 102(4):765–774. 10.1016/j.ijrobp.2018.04.065 Cusumano D, Dinapoli N, Boldrini L et al (2018) Fractal-based radiomic approach to predict complete pathological response after chemo-radiotherapy in rectal cancer. Radiol Med 123(4):286–295. 10.1007/s11547-017-0838-3 Jhaveri KS, Hosseini-Nik H (2015) MRI of Rectal Cancer: An Overview and Update on Recent Advances. AJR Am J Roentgenol 205(1):W42–55. 10.2214/AJR.14.14201 Niekel MC, Bipat S, Stoker J (2010) Diagnostic imaging of colorectal liver metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: a meta-analysis of prospective studies including patients who have not previously undergone treatment. Radiology 257(3):674–684. 10.1148/radiol.10100729 Lee SJ, Zea R, Kim DH, Lubner MG, Deming DA, Pickhardt PJ (2018) CT texture features of liver parenchyma for predicting development of metastatic disease and overall survival in patients with colorectal cancer. Eur Radiol 28(4):1520–1528. 10.1007/s00330-017-5111-6 Kijima S, Sasaki T, Nagata K, Utano K, Lefor AT, Sugimoto H (2014) Preoperative evaluation of colorectal cancer using CT colonography, MRI, and PET/CT. World J Gastroenterol 20(45):16964–16975. 10.3748/wjg.v20.i45.16964 Fowler KJ, Linehan DC, Menias CO (2013) Colorectal liver metastases: state of the art imaging. Ann Surg Oncol 20(4):1185–1193. 10.1245/s10434-012-2730-7 Lee KH, Lee JM, Park JH et al (2013) MR imaging in patients with suspected liver metastases: value of liver-specific contrast agent gadoxetic acid. Korean J Radiol 14(6):894–904. 10.3348/kjr.2013.14.6.894 Tsurusaki M, Sofue K, Murakami T (2016) Current evidence for the diagnostic value of gadoxetic acid-enhanced magnetic resonance imaging for liver metastasis. Hepatol Res 46(9):853–861. 10.1111/hepr.12646 Floriani I, Torri V, Rulli E et al (2010) Performance of imaging modalities in diagnosis of liver metastases from colorectal cancer: a systematic review and meta-analysis. J Magn Reson Imaging 31(1):19–31. 10.1002/jmri.22010 Jhaveri K, Cleary S, Audet P et al (2015) Consensus statements from a multidisciplinary expert panel on the utilization and application of a liver-specific MRI contrast agent (gadoxetic acid). AJR Am J Roentgenol 204(3):498–509. 10.2214/AJR.13.12399 Merkle EM, Zech CJ, Bartolozzi C et al (2016) Consensus report from the 7th International Forum for Liver Magnetic Resonance Imaging. Eur Radiol 26(3):674–682. 10.1007/s00330-015-3873-2 Bugg WG, Andreou AK, Biswas D, Toms AP, Williams SM (2014) The prognostic significance of MRI-detected extramural venous invasion in rectal carcinoma. Clin Radiol 69(6):619–623. 10.1016/j.crad.2014.01.010 Kim YC, Kim JK, Kim MJ, Lee JH, Kim YB, Shin SJ (2016) Feasibility of mesorectal vascular invasion in predicting early distant metastasis in patients with stage T3 rectal cancer based on rectal MRI. Eur Radiol 26(2):297–305. 10.1007/s00330-015-3837-6 Sohn B, Lim JS, Kim H et al (2015) MRI-detected extramural vascular invasion is an independent prognostic factor for synchronous metastasis in patients with rectal cancer. Eur Radiol 25(5):1347–1355. 10.1007/s00330-014-3527-9 Chuang SC, Su YC, Lu CY et al (2011) Risk factors for the development of metachronous liver metastasis in colorectal cancer patients after curative resection. World J Surg 35(2):424–429. 10.1007/s00268-010-0881-x Sivesgaard K, Larsen LP, Sorensen M et al (2018) Diagnostic accuracy of CE-CT, MRI and FDG PET/CT for detecting colorectal cancer liver metastases in patients considered eligible for hepatic resection and/or local ablation. Eur Radiol 28(11):4735–4747. 10.1007/s00330-018-5469-0 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5041812","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":358627244,"identity":"9af68ebf-334d-42e7-ae51-cd5b595b42bb","order_by":0,"name":"Hao Jiang","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Jiang","suffix":""},{"id":358627247,"identity":"681a3d72-3c32-46ee-9ff7-c119645c919c","order_by":1,"name":"Wei Guo","email":"","orcid":"","institution":"Harbin Medical University Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Guo","suffix":""},{"id":358627248,"identity":"734d18bf-2bcf-4268-a4e8-f554b94f3ab6","order_by":2,"name":"Xue Lin","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Xue","middleName":"","lastName":"Lin","suffix":""},{"id":358627249,"identity":"890623e5-79de-4d9b-a90b-d1ebb0a08dc4","order_by":3,"name":"Zhuo Yu","email":"","orcid":"","institution":"Hangzhou Linping research medical film technical service studio","correspondingAuthor":false,"prefix":"","firstName":"Zhuo","middleName":"","lastName":"Yu","suffix":""},{"id":358627250,"identity":"7726a4a2-c4c0-4d89-b190-cdd7cf871a9b","order_by":4,"name":"Yudie Qin","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yudie","middleName":"","lastName":"Qin","suffix":""},{"id":358627251,"identity":"6575a5ec-4cb3-4ae6-a704-2841a7dfa31b","order_by":5,"name":"Zhongqi Sun","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Zhongqi","middleName":"","lastName":"Sun","suffix":""},{"id":358627252,"identity":"14fad51b-f418-4450-ac75-e9d4ddf866fc","order_by":6,"name":"Hongbo Hu","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hongbo","middleName":"","lastName":"Hu","suffix":""},{"id":358627253,"identity":"2cfe0620-f083-419c-aa48-b25f82214478","order_by":7,"name":"Jinping Li","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Jinping","middleName":"","lastName":"Li","suffix":""},{"id":358627254,"identity":"ce64a18a-fcab-4eb6-811b-ea4d7c770dc0","order_by":8,"name":"Linhan Zhang","email":"","orcid":"","institution":"First Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Linhan","middleName":"","lastName":"Zhang","suffix":""},{"id":358627255,"identity":"20b5ff70-d8f8-4347-82fc-657259c74b47","order_by":9,"name":"Qiong Wu","email":"","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"prefix":"","firstName":"Qiong","middleName":"","lastName":"Wu","suffix":""},{"id":358627256,"identity":"7cde1b6d-8861-41a8-a1e4-280550078110","order_by":10,"name":"Huijie Jiang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvUlEQVRIiWNgGAWjYBACxhkMCUBKQoafmfnwA5K08Ei2s6UZEGeNBITiMTjPoyBBlA7m2Q0PH92oseAxPszDYMBQY0OEw+YcSDbOOSbBY3aY98ADhmNpRGiZkZAmncMG0sKXYMDYcJgoLem/c/5J8Bg38xhIEKsljTm3TYLHgJkELcnSuX0SPBKHgYGcQIxfDGfkJH7O+VYnx99/+PCDD8SEmGEDTwKCl4BLGTKQZ2A/QIy6UTAKRsEoGMkAAIxjNLmcFGcdAAAAAElFTkSuQmCC","orcid":"","institution":"Second Affiliated Hospital of Harbin Medical University","correspondingAuthor":true,"prefix":"","firstName":"Huijie","middleName":"","lastName":"Jiang","suffix":""}],"badges":[],"createdAt":"2024-09-06 05:55:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5041812/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5041812/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":67195127,"identity":"3716a5b0-24ed-4cd5-91f4-c994afde501c","added_by":"auto","created_at":"2024-10-22 09:05:39","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":209500,"visible":true,"origin":"","legend":"\u003cp\u003eFlow chart of patients’ recruitment pathway.\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5041812/v1/67d50c097c153ff5cc9d0ad5.jpg"},{"id":67194463,"identity":"f803a2ab-d002-4616-a81e-2f04ccaca320","added_by":"auto","created_at":"2024-10-22 08:57:39","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":52604,"visible":true,"origin":"","legend":"\u003cp\u003ePerformance of models based on different phase images. (A) Training set, (B) Test set.\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5041812/v1/12a34f375be6adb9e94ba3a4.jpg"},{"id":67194460,"identity":"3211ac07-eb89-44a0-8958-c749571eab6d","added_by":"auto","created_at":"2024-10-22 08:57:39","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":52389,"visible":true,"origin":"","legend":"\u003cp\u003eNomogram, constructed based on CEA, age, CA199, and T stage and radscore.\u003c/p\u003e","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5041812/v1/a56e18cb4b839dd5ea962508.jpg"},{"id":67194465,"identity":"674dd49b-b820-4ed1-84fd-d20acb64a0d2","added_by":"auto","created_at":"2024-10-22 08:57:39","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":114911,"visible":true,"origin":"","legend":"\u003cp\u003eThe performance of the nomogram on the training set and the test set. Figures (A) and (B) show the ROC of the radiomics model, clinical model, DWI+clinical model, T2W+clicincal model and nomogram model on the training set and the test set, respectively. Figures (C) and (D) show the calibration curves of the nomogram model on the training set and the test set, respectively.\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5041812/v1/9d1eef9d3c6b2c7026f94217.jpg"},{"id":67194462,"identity":"eac8e961-559c-42d7-bd6e-ef7dfce625ca","added_by":"auto","created_at":"2024-10-22 08:57:39","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":59468,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analysis.\u003c/p\u003e","description":"","filename":"Fig5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5041812/v1/9e09cf11f3ee20a5f2e97840.jpg"},{"id":75372611,"identity":"d6211144-fe5e-4a07-9d0a-e412a8f8a7a8","added_by":"auto","created_at":"2025-02-03 23:16:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1308016,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5041812/v1/4658a293-a8fb-4315-9a8a-a8434054f1ed.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Prediction of Preoperative Synchronous Distant Metastasis of Rectal Cancer Based on MRI Radiomics Model","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eRectal cancer is the third most common malignant cause of morbidity and mortality. Despite total mesorectal excision and neoadjuvant chemoradiotherapy, the local recurrence rate of rectal cancer has been significantly reduced to 5\u0026ndash;10%[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Distant metastasis remains the main cause of treatment failure in patients with rectal cancer [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In those metastatic cases, the liver is the most frequently involved organ, followed by the lung [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Surgical excision is the primary treatment strategy for early detection of metastasis. It has a better prognosis and survival rate compared to other treatments, offering these patients a chance of cure. For localized colorectal liver or lung metastases, resection has been considered the treatment of choice for improving longterm survival. The 5-year survival rates for patients treated with surgical resection of colorectal liver or lung metastasis can be enhanced to 58.0% or 56.2%, respectively[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Therefore, preoperative diagnosis of colorectal cancer patients at high risk for SDM is essential for personalized treatment strategies. The traditional standard in China defines simultaneous metastasis as colorectal cancer found at the time of diagnosis or metastasis occurring within 6 months after radical resection of the primary focus of colorectal cancer. However, some indicators are only available after radical resection and cannot be used as a basis for preoperative treatment strategies. Therefore, developing a preoperative, non-invasive, and accurate method to predict SDM is necessary.\u003c/p\u003e \u003cp\u003eRadiomics uses automated high-throughput extraction techniques for many quantitative features and can capture intra-tumor heterogeneity in a non-invasive manner, so it can be used in personalized medicine. Tumor heterogeneity can be reflected in imaging, creating the opportunity to identify imaging biomarkers that correlate with the tumor\u0026rsquo;s biological behavior. In this context, radiomics can play a key role, providing minable data from standard radiological images and exploring quantitative features that can describe tumor heterogeneity and other intrinsic characteristics that could correlate with its biological behavior[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Radiomics models based on computed tomography (CT)/MRI have been reported to predict metastatic lymph nodes and distant metastases of lung adenocarcinomas in bladder cancer [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] and breast cancer [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Furthermore, some studies have used the image analysis approach to identify the association between tumor tissue and distant metastases in different malignancies [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Specifically, Chen et al. showed that a CT-based radiomics model demonstrated good performance in the prediction of brain metastasis in lung cancer patients [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Other studies also reported that radiomics signature has the potential to predict liver metastases (LM) in rectal and esophagogastric cancer [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. In addition, the imaging features of primary rectal cancer are often less affected and more stable than those of metastases. However, few studies have been carried out to predict SDM by the imaging features of primary rectal cancer, which is worthy of further study. Gaitanidis et al [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] recently demonstrated the feasibility of predictive nomograms for evaluating the probability of synchronous liver, lung, and bone-distant disease in 46,785 rectal cancer patients. The results were promising with clinical and pathologic features in the proposed nomograms, yet the pathologic information is available only after surgery, which cannot be used to guide preoperative treatment strategy. Several studies have shown that radiomics models can predict distant metastases of different primary tumors [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. However, the role of radiomics nomograms from primary lesions in predicting SDM in RC patients has not been clearly defined. Therefore, it is necessary to develop preoperative non-invasive biomarkers to predict SDM.\u003c/p\u003e \u003cp\u003eMRI was the noninvasive imaging modality of choice for preoperative rectal cancer staging. It can also provide more than just morphological information, as images are data more than pictures. For rectal cancer, the application of radiomics mainly focused on the treatment response to chemoradiotherapy [\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. For the applications of radiomics in colorectal cancer, Huang et al. [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] reported that a CT-based radiomics nomogram facilitated the prediction of malignant lymph nodes in colorectal cancer. Other studies [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] on rectal cancer have also indicated that the MRI radiomics signature might be used as an effective biomarker for the prediction of long-term outcomes. However, in these rectal cancer radiomics studies, there are few studies on distant synchronous metastasis of rectal cancer.\u003c/p\u003e \u003cp\u003eMRI is well-established in the local staging of rectal cancer because of its superior efficacy compared to CT [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. In this study, we aimed to evaluate the value of MRI radiology based on T2-weighted imaging (T2WI) and DWI imaging in pre-operative identification of rectal cancer patients at high SDM risk and to establish a predictive clinical imaging combination model that can help improve decision-making and guide personalized treatment.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePatient Characteristics\u003c/h2\u003e \u003cp\u003eThis retrospective study was approved by the Ethics Committee of the \u003cb\u003e\u0026times;\u0026times;\u0026times;\u003c/b\u003e Hospital. We reviewed clinicopathological and imaging data collected from 465 RC patients confirmed by endoscopic biopsy or surgical pathology between January 2013 and December 2020. Including rectal cancer MRI examination, chest CT and abdominal CT contrast-enhanced examination, age, gender, pre-treated Carcinoembryonic antigen (CEA) data, pre-treated Carbohydrate antigen 199 (CA199) imaging, and clinical data, tumor size, tumor location, T stage and N stage predicted by MR. SDM inclusion criteria were as follows: (a) Patients with a pathological diagnosis of RC, (b) there is no prior history or co-existing history of other malignancies, (c) Patients undergoing preoperative high-resolution rectal MRI, chest CT, and contrast-enhanced abdominal CT, (d) Regular review and follow-up imaging data in our hospital were complete. Of these patients, SDM excluded based on the following exclusion criteria: (a) Patients who have received treatment (radiotherapy, chemotherapy, or chemoradiotherapy) before the MRI examination; (b) Patients whose images are of poor quality and cannot be analyzed; (c) Lack of clinical data; (d) Patients with mucinous adenocarcinoma. Finally, 169 patients were included according to the above criteria and were randomly divided into the training dataset (n\u0026thinsp;=\u0026thinsp;134) and the test dataset (n\u0026thinsp;=\u0026thinsp;35) with a segmentation ratio of 8:2. The detailed selection process is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eMRI Acquisition:\u003c/h2\u003e \u003cp\u003eAll patients underwent rectal MRI using the GE Discovery MR750w 3.0T MRI scanner in the supine position using a 32-channel phased array human coil. Take a low-residue diet one day before the MRI examination, and fast for 4\u0026ndash;6 hours before the MRI examination. A 60 mL enema was performed 1 h before the examination to better observe the primary tumor. Rectal dilation with air or water was not performed, and spasmolytics were not provided. The MRI scan sequences included: (1) T2-weighted images (T2WI) in high-resolution oblique (perpendicular to the long axis of the tumor; (2) Sagittal and coronal plane (parallel to the long axis of the rectum) T2WI spin echoes; (3) T1-weighted image (T1WI); (4) diffusion-weighted image (DWI). All patients also underwent contra-enhanced CT imaging of the chest and abdomen using a dual-source 64-MDCT (Somatom Definition Flash, Siemens Healthineers) scanner.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003ePreparation and Scanning Methods Before the Examination\u003c/h2\u003e \u003cp\u003eBefore the examination, confirm that the patient is stable and free of contraindications (no metal implants, no claustrophobia, etc). Explain to the patients before the examination and patients\u0026rsquo; informed consent. At least half an hour before the examination, the bowel was cleaned without air or any contrast agent.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eRadiomic Feature Extraction\u003c/h2\u003e \u003cp\u003eTwo radiologists (with \u0026gt;\u0026thinsp;5 years of experience in abdominal radiology diagnosis) examined all MRI data images and evaluated the T and N stages as defined by the American Cancer Federation's TN Staging System for Rectal Cancer, Version 8. The first radiologist manually mapped the tumor area of interest (ROI), and the second radiologist validated the ROI for the tumor tissue. Two other radiologists (with \u0026gt;\u0026thinsp;5 years of experience in chest and abdominal radiology diagnosis) consistently analyzed chest and abdominal CT images for distant metastasis assessment. All radiologists knew that these tumors were biopsies-confirmed adenocarcinomas of the rectum, but they did not know the histopathological stage of the patient. When observers cannot reach a consensus, ask another experienced radiologist (with \u0026gt;\u0026thinsp;10 years of experience in chest and abdominal diagnosis) to seek a final opinion.\u003c/p\u003e \u003cp\u003eRadiomics all work in the \u003cb\u003e\u0026times;\u0026times;\u0026times;\u003c/b\u003e medical research platform to complete. The radiomics workflow includes four steps: data collection and segmentation, radiomics feature extraction, feature screening, model building, and effect evaluation. Region of interest (ROIs) was segmented by doctors with more than 10 years of clinical experience on the ITK-SNAP software. The extraction of radiomics features are extracted through the PyRadiomics package of the Python language. For each image phase, 1688 radiomics features were extracted, including first-order features, shape features, texture features, wavelet features from 10 image types, which are original image, Laplacian\u0026ensp;of\u0026ensp;Gaussian (LoG, based on SimpleITK function), Wavelet ( Using the PyWavelets package), Square, Square\u0026ensp;Root, Logarithm, Exponential, Gradient, Local\u0026ensp;Binary\u0026ensp;Pattern (2D), Local\u0026ensp;Binary\u0026ensp;Pattern\u0026ensp;(3D).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eRadiomics Signature Construction\u003c/h2\u003e \u003cp\u003eWe sequentially use Variance Threshold, SelectKBest, and Least Absolute Shrinkage and Selection Operator (LASSO) for feature screening. The variance threshold method is used to eliminate features with small variance (\u0026lt;\u0026thinsp;0.8) because features with small variance have small changes in value and have little effect on the result identification. For the variance threshold method, values with variances smaller than the variance threshold (0.8) are removed. The SelectKBest method is based on the chi-square test leaving the features of \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05. The chi-square test is a method for the degree of deviation between the actual observed value of the statistical sample and the theoretically inferred value. The LASSO method obtains the sparseness of the feature weight The matrix realizes the screening of features. This topic uses the L2 regularizer as the loss function. Finally, a radiomics model was constructed based on Support Vector Machines (SVM).\u003c/p\u003e \u003cp\u003eAccording to the above process, the features were extracted from different phases of images and the pooling features of the two phases were screened, and then based on these screened features, the radiomics models of different phases and the fusion phases were constructed. To evaluate the performance of these models, five indicators of receiver operating characteristic (ROC), area under the curve (AUC), accuracy, sensitivity, specificity, and 95% confidence interval (CI) were selected for evaluation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment of the Clinical Model and the Nomogram\u003c/h2\u003e \u003cp\u003eTo construct the clinical model, we used univariate analysis to identify independent risk factors between the NR group and the R group. Clinical models were built using significant factors (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) from univariate analysis. The clinical-radiomics nomogram was built using the selected pooling features of the two phases and independent clinical-radiologic risk factors in the clinical model. The fusion characteristics of DWI and T2WI were combined with clinically independent risk factors for synchronous liver metastases of rectal cancer for Logistic regression analysis, and the fusion model was established and a nomogram of the fusion model was mapped to provide a quantitative guidance tool for clinical diagnosis and prediction of synchronous liver metastases of rectal cancer. The receiver operating characteristic curve (ROC) was drawn, and the diagnostic effect of the three models was compared by the Delong test. In addition, a calibration curve test was used to assess the consistency between the predicted risks of the model and the actual risks. The statistical analysis method of decision curve analysis (DCA) was used to verify the clinical effectiveness of the model。\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eExperimental Environment\u003c/h2\u003e \u003cp\u003eAll experiments are completed on the PyCharm program, the programming language used is Python (v. 3.6), mainly using sklearn, numpy, scipy, pandas, and other packages.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eNormal test was performed on all measurement data, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (SD) was used to represent consistent with normal distribution, and median was used to represent inconsistent with normal distribution. Independent sample t test or Mann-Whitney \u003cem\u003eU\u003c/em\u003e test were used for measurement data of two groups with metastasis and non-metastasis. The statistical data were expressed by example (%), chi square test was used between the two groups, the theoretical frequency was less than 5, and Fisher\u0026rsquo;s exact test was used to test. ROC curves were drawn, and the area under the curve (AUC) was calculated to assess the diagnostic performance of each model. The DeLong test was used to compare the AUCs of the different classification models. Decision curve analysis (DCA) was performed to compare the differences in net benefits between the different models at the threshold probability. SPSS 26.0 statistical software was used to analyze the data. The reported statistical significance levels were all two sided, with the statistical significance level set at 0.05.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003ePatient Characteristics\u003c/h2\u003e \u003cp\u003eBaseline clinical characteristics are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. There is no significant difference between the data distribution of the training set and the test set (69/65 vs 18/17, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.900). On the training set and the test set, the clinical factors that have significant differences at the same time are (\u003cem\u003eP\u003c/em\u003e), age (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), CA199(\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), and T stage(\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Tumor location was only significantly different on the test set. Other clinical variables were not significantly different on the training and test sets.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCharacteristics of patients and associations with synchronous distant metastasis\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNon-metastasis (n\u0026thinsp;=\u0026thinsp;87)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMetastasis (n\u0026thinsp;=\u0026thinsp;82)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eχ\u003csup\u003e2\u003c/sup\u003e/t\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD, years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e57.9\u0026thinsp;\u0026plusmn;\u0026thinsp;11.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e62.3\u0026thinsp;\u0026plusmn;\u0026thinsp;8.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-2.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.356\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62 (71.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e53 (64.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e25 (28.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29 (35.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCEA, \u0026micro;g/L\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e42.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e69 (79.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24 (29.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 (20.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58 (70.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCA199, U/ml\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e22.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e78(89.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e47(57.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9(10.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35(42.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor diameter, mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD, cm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.3\u0026thinsp;\u0026plusmn;\u0026thinsp;1.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.6\u0026thinsp;\u0026plusmn;\u0026thinsp;1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.086\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor location (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.279\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProximal rectum\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31(35.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23(28.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMiddle rectum\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e50(57.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48(58.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDistal rectum\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6(6.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11(13.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDegree of differentiation (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.498\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWell\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7(8.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e10(12.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e74(85.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64(78.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePoor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6(6.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8(9.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT staging (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.040\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT1-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14(16.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5(6.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT3-4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e73(83.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e77(93.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN staging (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.093\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23(26.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13(15.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eN1-2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e64(73.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e69(84.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003eCEA carcinoembryonic antigen, CA199 carbohydrate antigen 199\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"5\"\u003e\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicates a statistically significant difference\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment of the Radiomics Model\u003c/h2\u003e \u003cp\u003e1688 radiomics features were extracted from each imaging phase. After feature screening, DWI phase images retained 8 features; T2W phase images retained 8 features; Radiomics model(DWI phase and T2W phase images) retained 6 features, among which, there are 4 features of DWI phase images and 2 features of T2W phase images. In terms of the retained characteristics of each phase, the SVM model was selected to construct the radiomics model of each phase. The experimental results show that none of radiomics models has a good discriminative ability for SDM(Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). On the training set, the DWI model performed best, and its AUC, accuracy, sensitivity, specificity, and 95%CI were 0.88, 0.82, 0.55, 0.92, 0.84\u0026ndash;0.92, respectively; on the test set, the Radiomics model performed the best, and its AUC, accuracy, sensitivity, specificity, and 95%CI are 0.82, 0.71, 0.76, 0.67, 0.62\u0026ndash;0.86, respectively.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePerformance of models based on different phase images.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"12\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eClassifiers\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c6\" namest=\"c2\"\u003e \u003cp\u003eTrain set\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c12\" namest=\"c8\"\u003e \u003cp\u003eTest set\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eACC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSEN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e95%CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eACC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eSEN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003eSPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c12\"\u003e \u003cp\u003e95%CI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDWI model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.84\u0026ndash;0.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.59\u0026ndash;0.83\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT2W model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.5\u0026ndash;0.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.44\u0026ndash;0.73\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.83\u0026ndash;0.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.62\u0026ndash;0.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment of the Clinical Model and Nomogram\u003c/h2\u003e \u003cp\u003eThe clinical model developed using four independent risk factors (CEA, age, CA199, and T stage) showed AUCs of 0.81 and 0.83 in the training and test datasets, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA, \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). Combining clinical factors and radiomics characteristics of different sequence images, we established three clinical-radiomics models, namely the DWI\u0026thinsp;+\u0026thinsp;clinical model, the T2W\u0026thinsp;+\u0026thinsp;clinical model, and the nomogram (radiomics\u0026thinsp;+\u0026thinsp;clinical) model (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In terms of indicators, the nomogram model has the best discriminative performance on SDM(Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA, \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). On the training set, the AUC, accuracy, sensitivity, specificity, and 95% CI of the nomogram model are 0.93, respectively. 0.85, 0.85, 0.86, 0.89\u0026ndash;0.96. On the test set, the AUC, accuracy, sensitivity, specificity, and 95% CI of the nomogram model are 0.94, 0.89, 0.88, 0.89, and 0.79\u0026ndash;0.97, respectively. The calibration plots (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC, \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD) were consistent between the clinical radiomics model predicted and observed probabilities. Decision curve analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) showed that the nomogram model achieved the highest net benefit compared with the clinical model and the radiomics model on the training and test sets.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePerformance of radiomics model, clinical model, DWI\u0026thinsp;+\u0026thinsp;clinical model, T2W\u0026thinsp;+\u0026thinsp;clinical model, and nomogram.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"12\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eClassifiers\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c6\" namest=\"c2\"\u003e \u003cp\u003eTrain set\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c12\" namest=\"c8\"\u003e \u003cp\u003eTest set\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eACC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSEN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e95%CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eACC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eSEN\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c11\"\u003e \u003cp\u003eSPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c12\"\u003e \u003cp\u003e95%CI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiomics model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.7\u0026ndash;0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.62\u0026ndash;0.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.74\u0026ndash;0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.72\u0026ndash;0.92\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDWI\u0026thinsp;+\u0026thinsp;clinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.9\u0026ndash;0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.68\u0026ndash;0.89\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT2W\u0026thinsp;+\u0026thinsp;clinical model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.91\u0026ndash;0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.62\u0026ndash;0.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNomogram model\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.89\u0026ndash;0.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c11\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c12\"\u003e \u003cp\u003e0.79\u0026ndash;0.97\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur previous experience confirmed that MRI staging-based radiomics models provide a relevant predictive tool to identify the oncological behavior of pCR after nCRT [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Despite the important effort made in terms of treatment response prediction, few experiences reported the relation between radiomics predictors and early distant recurrence [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Our results show that adding radiomics features to the clinical model can better predict SDM performance in rectal cancer patients, increasing the AUC from 0.83 to 0.94, with high sensitivity and specificity in the validation cohort. The high specificity indicates that the model is reliable and can eliminate more false positive and false negative patients. We developed a clinical-radiomics nomogram as an individual and visualization tool to provide an estimated probability of SDM for newly diagnosed rectal cancer patients. Decision curve analysis was used to determine its clinical benefit. Radiomics, integrating many high-dimensional imaging features used to quantify tumor heterogeneity, could facilitate oncologic diagnosis and prognosis prediction. Huang et al [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] revealed that radiomics signature including 24 selected features could help predict LN-positive patients with a C-index of 0.773 in the validation cohort and the proposed clinical-radiomics nomogram was useful for predicting LN involvement. In our study, radiomics features were obtained using high-resolution axial T2WI and DWI, which are the most critical sequences for evaluating primary rectal cancer[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Functional imaging techniques such as DWI can analyze lesions from the perspective of the microenvironment and molecular pathology, to quantitatively study many small lesions that cannot be quantified and identified by the naked eye. Our proposed clinical-radiomics predictive model confirms the feasibility of imaging analysis based on T2WI and DWI, providing a potentially effective and easy-to-use model in clinical practice. Using our clinical-radiomics nomogram, an estimated probability of SDM could be calculated after referring to the selected T2WI-based radiomics features, DWI-based radiomics features as well as other clinical information.\u003c/p\u003e \u003cp\u003eOur results show that the radiomics nomogram provides predictive information about SDM in primary rectal cancer. Contrast-enhanced CT, MRI, and PET-CT are common imaging examinations for the diagnosis of SLM in RC preoperatively. However, the sensitivity and accuracy of these imaging techniques are not satisfactory[\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. According to one meta-analysis, the detection sensitivity of colorectal LM in contrast-enhanced CT, routine MRI, and FDG PET-CT were 63%-80%, 76%-85.7%, and 51%-90%, respectively[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Many studies have shown that MRI had a higher accuracy compared to CT in diagnosing SLM of rectal cancer patients[\u003cspan additionalcitationids=\"CR33\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], and recent consensus guidelines from the radiologic community recommend MRI for the preoperative evaluation of SLM [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. The other studies had shown that some adverse features found on rectal MRI identified patients with rectal cancer at higher risk of distant metastasis[\u003cspan additionalcitationids=\"CR38\" citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. Therefore, we developed a radiomics nomogram based on multi-sequence MRI to predict the risk of SDM.\u003c/p\u003e \u003cp\u003eTherefore, screening for high-risk predictors would improve the probability of early detection of SDM in RC patients. Typically, the clinicopathological predictors of SDM in RC patients include the histological type, pathological grade, depth of tumor invasion, lymph node status, vascular invasion, and tumor markers[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. However, some of these predictors can only be obtained postoperatively, and hence, inappropriate to guide preoperative treatment. Other studies have demonstrated that some features of rectal MRI, such as extramural vascular invasion, higher T stage, and regional lymph node metastasis are potential predictors[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. However, these image features are subjective and qualitative, lacking quantitative assessment. In recent years, radiomics has been regarded as an advanced tool for evaluating tumor heterogeneity in tumor diagnosis and prognosis prediction. In this study, factors such as radiomics features, CEA, and CA19-9 levels were incorporated into multivariate logistic regression to build a prediction model and a nomogram, and the research results are promising. Therefore, our analysis indicates that the radiomics nomogram combined with tumor markers was superior to the radiomics signature alone. It exhibited a high predictive performance for SDM in RC patients, and the AUC improved from 0.82 to 0.94. Moreover, the results were better than those reported in a previous study on a per-patient basis, wherein the AUC was 0.92 and 0.88 (MRI readers), 0.80 and 0.82 (CT readers), and an AUC of 0.83 and 0.84 (PET-CT readers)[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. In this study, we constructed a primary RC-based radiomics nomogram from high-resolution T2WI and DWI of the recta to predict SDM. On the training set, the AUC, accuracy, sensitivity, specificity, and 95%CI of the nomogram model were 0.93, 0.85, 0.85, 0.86, and 0.89\u0026ndash;0.96, respectively. The AUC, accuracy, sensitivity, specificity, and 95% CI of the nomogram model were 0.94, 0.89, 0.89, and 0.79\u0026thinsp;~\u0026thinsp;0.97 respectively. Therefore, based on clinical risk factors and radiomics characteristics, the proposed nomogram may be a valuable predictive tool for SDM in RC patients. It can be easily used to identify patients who need further whole-body imaging.\u003c/p\u003e \u003cp\u003eRadiomics features are composed of multiple features, which is of great significance for selecting the optimal feature collection through dimensionality reduction. In this study, we used the LASSO method for feature screening and built a radiomics model based on a support vector machine (SVM). In our study, the method of noose dimension reduction is chosen to increase the stability of the nomogram and to carry out the overall analysis. The nomogram we constructed is easier for clinicians to use because it can derive information from T2WI, and DWI, and includes clinical risk factors. The radiomics nomogram is an auxiliary tool that can be used to identify and follow those patients with rectal cancer. When the group was divided into a low-risk group and a high-risk group, the high-risk group had a higher probability of developing SDM. Therefore, in a certain sense, the radiomics nomogram can be used as an accurate and reliable detection tool for SDM in rectal cancer patients. It is quick and easy to perform and helps to determine which patients will benefit from further imaging of distant metastases.\u003c/p\u003e \u003cp\u003eHowever, the current study still has some limitations. First of all, the sample size based on single-institution retrospective analysis is relatively small, and selection bias is inevitable. Second, this study lacks external validation, so a large multicenter trial is needed to improve the generalizability of the results.\u003c/p\u003e \u003cp\u003eIn summary, we developed a clinical imaging nomogram by combining tumor markers with imaging features to accurately predict the presence of SDM in RC patients. This visualization tool will detect the probability of SDM and help doctors make clinical decisions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eH.J. and W.G. carried out the studies, participated in collecting data, and drafted the manuscript. H.J. and Z.Y. performed the statistical analysis and participated in its design. Y.D.Q.、Z.Q.S. and X.L. participated in the acquisition and analysis of data. H.J.J. designed the study and carefully revised the manuscript. H.B.H.、J.P.L.、Q.W. and L.H.Z. examined the manuscript carefully. All authors read and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBosset JF, Collette L, Calais G et al (2006) Chemotherapy with preoperative radiotherapy in rectal cancer. N Engl J Med 355(11):1114\u0026ndash;1123. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1056/NEJMoa060829\u003c/span\u003e\u003cspan address=\"10.1056/NEJMoa060829\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCeelen W, Fierens K, Van Nieuwenhove Y, Pattyn P (2009) Preoperative chemoradiation versus radiation alone for stage II and III resectable rectal cancer: a systematic review and meta-analysis. Int J Cancer 124(12):2966\u0026ndash;2972. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ijc.24247\u003c/span\u003e\u003cspan address=\"10.1002/ijc.24247\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRodel C, Liersch T, Becker H et al (2012) Preoperative chemoradiotherapy and postoperative chemotherapy with fluorouracil and oxaliplatin versus fluorouracil alone in locally advanced rectal cancer: initial results of the German CAO/ARO/AIO-04 randomised phase 3 trial. Lancet Oncol 13(7):679\u0026ndash;687. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S1470-2045(12)70187-0\u003c/span\u003e\u003cspan address=\"10.1016/S1470-2045(12)70187-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan der Geest LG, Lam-Boer J, Koopman M, Verhoef C, Elferink MA, de Wilt JH (2015) Nationwide trends in incidence, treatment and survival of colorectal cancer patients with synchronous metastases. Clin Exp Metastasis 32(5):457\u0026ndash;465. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10585-015-9719-0\u003c/span\u003e\u003cspan address=\"10.1007/s10585-015-9719-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHur H, Ko YT, Min BS et al (2009) Comparative study of resection and radiofrequency ablation in the treatment of solitary colorectal liver metastases. Am J Surg 197(6):728\u0026ndash;736. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.amjsurg.2008.04.013\u003c/span\u003e\u003cspan address=\"10.1016/j.amjsurg.2008.04.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKanas GP, Taylor A, Primrose JN et al (2012) Survival after liver resection in metastatic colorectal cancer: review and meta-analysis of prognostic factors. Clin Epidemiol 4:283\u0026ndash;301. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2147/CLEP.S34285\u003c/span\u003e\u003cspan address=\"10.2147/CLEP.S34285\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO'Connor JPB (2017) Cancer heterogeneity and imaging. Semin Cell Dev Biol 64:48\u0026ndash;57. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.semcdb.2016.10.001\u003c/span\u003e\u003cspan address=\"10.1016/j.semcdb.2016.10.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu S, Zheng J, Li Y et al (2017) A Radiomics Nomogram for the Preoperative Prediction of Lymph Node Metastasis in Bladder Cancer. Clin Cancer Res 23(22):6904\u0026ndash;6911. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1158/1078-0432.CCR-17-1510\u003c/span\u003e\u003cspan address=\"10.1158/1078-0432.CCR-17-1510\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong Y, Feng Q, Yang W et al (2018) Preoperative prediction of sentinel lymph node metastasis in breast cancer based on radiomics of T2-weighted fat-suppression and diffusion-weighted MRI. Eur Radiol 28(2):582\u0026ndash;591. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-017-5005-7\u003c/span\u003e\u003cspan address=\"10.1007/s00330-017-5005-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoroller TP, Grossmann P, Hou Y et al (2015) CT-based radiomic signature predicts distant metastasis in lung adenocarcinoma. Radiother Oncol 114(3):345\u0026ndash;350. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.radonc.2015.02.015\u003c/span\u003e\u003cspan address=\"10.1016/j.radonc.2015.02.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu H, Zhang C, Wang L et al (2019) MRI radiomics analysis for predicting preoperative synchronous distant metastasis in patients with rectal cancer. Eur Radiol 29(8):4418\u0026ndash;4426. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-018-5802-7\u003c/span\u003e\u003cspan address=\"10.1007/s00330-018-5802-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang L, Dong D, Li H et al (2019) Development and validation of a magnetic resonance imaging-based model for the prediction of distant metastasis before initial treatment of nasopharyngeal carcinoma: A retrospective cohort study. EBioMedicine 40:327\u0026ndash;335. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ebiom.2019.01.013\u003c/span\u003e\u003cspan address=\"10.1016/j.ebiom.2019.01.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen A, Lu L, Pu X et al (2019) CT-Based Radiomics Model for Predicting Brain Metastasis in Category T1 Lung Adenocarcinoma. AJR Am J Roentgenol 213(1):134\u0026ndash;139. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2214/AJR.18.20591\u003c/span\u003e\u003cspan address=\"10.2214/AJR.18.20591\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang M, Cai Z, Zhang H et al (2019) Machine Learning-based Analysis of Rectal Cancer MRI Radiomics for Prediction of Metachronous Liver Metastasis. Acad Radiol 26(11):1495\u0026ndash;1504. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.acra.2018.12.019\u003c/span\u003e\u003cspan address=\"10.1016/j.acra.2018.12.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlaassen R, Larue R, Mearadji B et al (2018) Feasibility of CT radiomics to predict treatment response of individual liver metastases in esophagogastric cancer patients. PLoS ONE 13(11):e0207362. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0207362\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0207362\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGaitanidis A, Alevizakos M, Tsaroucha A, Tsalikidis C, Pitiakoudis M (2018) Predictive Nomograms for Synchronous Distant Metastasis in Rectal Cancer. J Gastrointest Surg 22(7):1268\u0026ndash;1276. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11605-018-3767-0\u003c/span\u003e\u003cspan address=\"10.1007/s11605-018-3767-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Y, Eresen A, Shangguan J et al (2019) Establishment of a new non-invasive imaging prediction model for liver metastasis in colon cancer. Am J Cancer Res 9(11):2482\u0026ndash;2492\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan L, Fang M, Tu W et al (2019) Radiomics Signature: A Biomarker for the Preoperative Distant Metastatic Prediction of Stage I Nonsmall Cell Lung Cancer. Acad Radiol 26(9):1253\u0026ndash;1261. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.acra.2018.11.004\u003c/span\u003e\u003cspan address=\"10.1016/j.acra.2018.11.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeng Y, Zhang Y, Dong D et al (2018) Novel radiomic signature as a prognostic biomarker for locally advanced rectal cancer. J Magn Reson Imaging. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jmri.25968\u003c/span\u003e\u003cspan address=\"10.1002/jmri.25968\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorvat N, Veeraraghavan H, Khan M et al (2018) MR Imaging of Rectal Cancer: Radiomics Analysis to Assess Treatment Response after Neoadjuvant Therapy. Radiology 287(3):833\u0026ndash;843. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1148/radiol.2018172300\u003c/span\u003e\u003cspan address=\"10.1148/radiol.2018172300\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNie K, Shi L, Chen Q et al (2016) Rectal Cancer: Assessment of Neoadjuvant Chemoradiation Outcome based on Radiomics of Multiparametric MRI. Clin Cancer Res 22(21):5256\u0026ndash;5264. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1158/1078-0432.CCR-15-2997\u003c/span\u003e\u003cspan address=\"10.1158/1078-0432.CCR-15-2997\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang YQ, Liang CH, He L et al (2016) Development and Validation of a Radiomics Nomogram for Preoperative Prediction of Lymph Node Metastasis in Colorectal Cancer. J Clin Oncol 34(18):2157\u0026ndash;2164. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1200/JCO.2015.65.9128\u003c/span\u003e\u003cspan address=\"10.1200/JCO.2015.65.9128\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaizlin ZV, Brown JA, So G et al (2010) Can CT replace MRI in preoperative assessment of the circumferential resection margin in rectal cancer? Dis Colon Rectum 53(3):308\u0026ndash;314. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/DCR.0b013e3181c5321e\u003c/span\u003e\u003cspan address=\"10.1007/DCR.0b013e3181c5321e\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang H, Guo W, Yu Z et al (2023) A Comprehensive Prediction Model Based on MRI Radiomics and Clinical Factors to Predict Tumor Response After Neoadjuvant Chemoradiotherapy in Rectal Cancer. Acad Radiol 30(Suppl 1):S185\u0026ndash;S98. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.acra.2023.04.032\u003c/span\u003e\u003cspan address=\"10.1016/j.acra.2023.04.032\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDinapoli N, Barbaro B, Gatta R et al (2018) Magnetic Resonance, Vendor-independent, Intensity Histogram Analysis Predicting Pathologic Complete Response After Radiochemotherapy of Rectal Cancer. Int J Radiat Oncol Biol Phys 102(4):765\u0026ndash;774. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ijrobp.2018.04.065\u003c/span\u003e\u003cspan address=\"10.1016/j.ijrobp.2018.04.065\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCusumano D, Dinapoli N, Boldrini L et al (2018) Fractal-based radiomic approach to predict complete pathological response after chemo-radiotherapy in rectal cancer. Radiol Med 123(4):286\u0026ndash;295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11547-017-0838-3\u003c/span\u003e\u003cspan address=\"10.1007/s11547-017-0838-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJhaveri KS, Hosseini-Nik H (2015) MRI of Rectal Cancer: An Overview and Update on Recent Advances. AJR Am J Roentgenol 205(1):W42\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2214/AJR.14.14201\u003c/span\u003e\u003cspan address=\"10.2214/AJR.14.14201\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNiekel MC, Bipat S, Stoker J (2010) Diagnostic imaging of colorectal liver metastases with CT, MR imaging, FDG PET, and/or FDG PET/CT: a meta-analysis of prospective studies including patients who have not previously undergone treatment. Radiology 257(3):674\u0026ndash;684. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1148/radiol.10100729\u003c/span\u003e\u003cspan address=\"10.1148/radiol.10100729\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee SJ, Zea R, Kim DH, Lubner MG, Deming DA, Pickhardt PJ (2018) CT texture features of liver parenchyma for predicting development of metastatic disease and overall survival in patients with colorectal cancer. Eur Radiol 28(4):1520\u0026ndash;1528. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-017-5111-6\u003c/span\u003e\u003cspan address=\"10.1007/s00330-017-5111-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKijima S, Sasaki T, Nagata K, Utano K, Lefor AT, Sugimoto H (2014) Preoperative evaluation of colorectal cancer using CT colonography, MRI, and PET/CT. World J Gastroenterol 20(45):16964\u0026ndash;16975. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3748/wjg.v20.i45.16964\u003c/span\u003e\u003cspan address=\"10.3748/wjg.v20.i45.16964\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFowler KJ, Linehan DC, Menias CO (2013) Colorectal liver metastases: state of the art imaging. Ann Surg Oncol 20(4):1185\u0026ndash;1193. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1245/s10434-012-2730-7\u003c/span\u003e\u003cspan address=\"10.1245/s10434-012-2730-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee KH, Lee JM, Park JH et al (2013) MR imaging in patients with suspected liver metastases: value of liver-specific contrast agent gadoxetic acid. Korean J Radiol 14(6):894\u0026ndash;904. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3348/kjr.2013.14.6.894\u003c/span\u003e\u003cspan address=\"10.3348/kjr.2013.14.6.894\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTsurusaki M, Sofue K, Murakami T (2016) Current evidence for the diagnostic value of gadoxetic acid-enhanced magnetic resonance imaging for liver metastasis. Hepatol Res 46(9):853\u0026ndash;861. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/hepr.12646\u003c/span\u003e\u003cspan address=\"10.1111/hepr.12646\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFloriani I, Torri V, Rulli E et al (2010) Performance of imaging modalities in diagnosis of liver metastases from colorectal cancer: a systematic review and meta-analysis. J Magn Reson Imaging 31(1):19\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jmri.22010\u003c/span\u003e\u003cspan address=\"10.1002/jmri.22010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJhaveri K, Cleary S, Audet P et al (2015) Consensus statements from a multidisciplinary expert panel on the utilization and application of a liver-specific MRI contrast agent (gadoxetic acid). AJR Am J Roentgenol 204(3):498\u0026ndash;509. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2214/AJR.13.12399\u003c/span\u003e\u003cspan address=\"10.2214/AJR.13.12399\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMerkle EM, Zech CJ, Bartolozzi C et al (2016) Consensus report from the 7th International Forum for Liver Magnetic Resonance Imaging. Eur Radiol 26(3):674\u0026ndash;682. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-015-3873-2\u003c/span\u003e\u003cspan address=\"10.1007/s00330-015-3873-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBugg WG, Andreou AK, Biswas D, Toms AP, Williams SM (2014) The prognostic significance of MRI-detected extramural venous invasion in rectal carcinoma. Clin Radiol 69(6):619\u0026ndash;623. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.crad.2014.01.010\u003c/span\u003e\u003cspan address=\"10.1016/j.crad.2014.01.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim YC, Kim JK, Kim MJ, Lee JH, Kim YB, Shin SJ (2016) Feasibility of mesorectal vascular invasion in predicting early distant metastasis in patients with stage T3 rectal cancer based on rectal MRI. Eur Radiol 26(2):297\u0026ndash;305. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-015-3837-6\u003c/span\u003e\u003cspan address=\"10.1007/s00330-015-3837-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSohn B, Lim JS, Kim H et al (2015) MRI-detected extramural vascular invasion is an independent prognostic factor for synchronous metastasis in patients with rectal cancer. Eur Radiol 25(5):1347\u0026ndash;1355. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-014-3527-9\u003c/span\u003e\u003cspan address=\"10.1007/s00330-014-3527-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChuang SC, Su YC, Lu CY et al (2011) Risk factors for the development of metachronous liver metastasis in colorectal cancer patients after curative resection. World J Surg 35(2):424\u0026ndash;429. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00268-010-0881-x\u003c/span\u003e\u003cspan address=\"10.1007/s00268-010-0881-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSivesgaard K, Larsen LP, Sorensen M et al (2018) Diagnostic accuracy of CE-CT, MRI and FDG PET/CT for detecting colorectal cancer liver metastases in patients considered eligible for hepatic resection and/or local ablation. Eur Radiol 28(11):4735\u0026ndash;4747. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-018-5469-0\u003c/span\u003e\u003cspan address=\"10.1007/s00330-018-5469-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"magnetic resonance imaging, rectal cancer, synchronous distant metastasis, radiomics","lastPublishedDoi":"10.21203/rs.3.rs-5041812/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5041812/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e \u003cp\u003eThe objective of this study was to develop and validate a new non-invasive artificial intelligence (AI) model based on preoperative magnetic resonance imaging (MRI) data to predict the presence of synchronous distant metastasis (SDM) in rectal cancer (RC).\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003e169 eligible RC patients were enrolled, and T2WI and DWI sequence images were collected. The radiomics features were extracted through the PyRadiomics package of Python language, and a total of 1688 radiomics features were extracted, including first-order features, shape features, texture features, and Baud signs. One clinical model and three comprehensive models of clinical imaging were constructed. Five indexes including receiver operating characteristic (ROC), area under curve (AUC), accuracy, sensitivity, specificity, and 95% confidence interval (CI) were selected to evaluate the model. The clinical model using four independent risk factors (CEA, age, CA199, and T stage). Combining the clinical factors and imaging characteristics of different sequences, we established three clinically-imaging models: the DWI\u0026thinsp;+\u0026thinsp;clinical model, the T2W\u0026thinsp;+\u0026thinsp;clinical model, and the nomogram (radiomics\u0026thinsp;+\u0026thinsp;clinical) model.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThis nomogram model performed the best in predicting rectal cancer SDM. In the training set, the AUC, accuracy, sensitivity, specificity and 95%CI of the nomogram model were 0.93, 0.85, 0.85, 0.86, 0.89\u0026ndash;0.96, respectively. In the test set, five indexes of the nomogram model were 0.94, 0.89, 0.88, 0.89, and 0.79\u0026thinsp;~\u0026thinsp;0.97, respectively. The correction plots were consistent between the predictions of the clinical radiomics model and the actual observed probabilities. Decision curve analysis showed that the nomogram model achieved the highest net benefit on the training set and the test set compared to the clinical model and the radiomics model.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eOur predictive model is valuable for guiding and managing patients with rectal cancer SDM, providing options for improving patient treatment decisions and guiding personalized treatment regimens.\u003c/p\u003e","manuscriptTitle":"Prediction of Preoperative Synchronous Distant Metastasis of Rectal Cancer Based on MRI Radiomics Model","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-22 08:57:34","doi":"10.21203/rs.3.rs-5041812/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"df6e31f6-ac88-4762-8e91-e8e8897a0256","owner":[],"postedDate":"October 22nd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-02-03T23:08:25+00:00","versionOfRecord":[],"versionCreatedAt":"2024-10-22 08:57:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5041812","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5041812","identity":"rs-5041812","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}