MRI-based Radiomics analysis for differentiation degree of gastric cancer

MRI-based Radiomics analysis for differentiation degree of gastric cancer | 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 MRI-based Radiomics analysis for differentiation degree of gastric cancer Yilin Wang, Letian Yuan This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7493820/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 08 Dec, 2025 Read the published version in BMC Medical Imaging → Version 1 posted 16 You are reading this latest preprint version Abstract Background: Preoperative differentiation between poorly and highly differentiated gastric cancers is important for treatment decisions. Purpose: To investigate a radiomics model for preoperative differentiation between poorly and highly differentiated gastric cancers. Study type: Retrospective. Population: 239 patients with gastric cancer were included in the study. 167 patients were assigned to the training group and 72 patients comprised the testing group. Sequence: T2-weighted (T2WI), T1-weighted (T1WI), diffusion-weighted imaging(DWI)on a 3.0T MR scanner. Assessment: : A total of 2632 radiomics features were extracted from DWI and apparent diffusion coefficient (ADC) maps. Radiomics based on above features were built using four feature selection methods and three classifiers. All models were used to differentiate poorly and highly differentiated gastric cancers. Statistical tests: 1) An analysis of independent t test was performed for clinicopathological information. 2) Four feature selection methods (Least Absolute Shrinkage and Selection Operator [LASSO], Analysis Of Variance [ANOVA], Recursive Feature Elimination [RFE] and Kruskal Wallis[KW]) and three classifiers (Support Vector Machine [SVM], Linear Discriminant Analysis [LDA] and Logistic Regression [LR]) were used to construct twelve radiomics. 3) The performance of different radiomics model was assessed using area under the receiver-operating characteristic curve (AUC) and accuracy (Acc) values. Results: The model LASSO + SVM achieved the highest AUC of 0.854 with Acc of 0.793 in all radiomics model. The combination of ADC min and radiomics achieved the highest diagnostic efficiency with AUC of 0.919 and Acc of 0.823. Data Conclusion: The combined model of radiomics and ADC min was accurate for distinguishing poorly and highly differentiated gastric cancers. MRI radiomics ADCmin differentiation degree gastric cancer Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Gastric cancer is the fifth most common tumor with third mortality in the world. It was reported that the incidence rate and mortality of gastric cancer (GC) have gradually increased worldwide over the past decades. Every year, more than one million people worldwide have been diagnosed with gastric cancer. ( 1 – 4 ). Gastric cancer (GC) is a malignant tumor originated from gastrointestinal mucosal epithelium and gastric adenocarcinoma accounts for more than 90% of all GC ( 5 – 7 ). At present, the most effective treatment for gastric cancer is completely surgical resection. However, it is usually suitable for patients with early gastric cancer ( 8 – 9 ). For patients with poorly differentiated or advanced gastric cancer, even if the tumor is completely removed, most patients may encounter poor prognosis (e.g. local recurrence or distant metastasis) and the prognosis is associated with the differentiation degree (DD) and lymph node metastasis (LNM) ( 10 – 12 ). Therefore, accurate preoperative evaluation of DD of gastric cancer is very important for selecting the treatment strategy and predicting the prognosis. Several imaging examination techniques could diagnose gastric cancer, including computerized tomography (CT), magnetic resonance imaging (MRI) and gastroscope. Due to the advantage of high soft tissue resolution, MRI can better show the degree of gastric mucosal invasion and lymph node metastasis. Therefore, it has become an effective means for the diagnosis, staging and efficacy evaluation of gastric cancer ( 13 – 16 ). However, it is hard to determine the DD of gastric cancer, which can not rely on the subjective evaluation of radiologists. Therefore, it is of great clinical significance to find a noninvasive examination method that can effectively evaluate the DD of gastric cancer before operation. As a new methodology, radiomics can quantitatively extract the algorithmic features of imaging data, such as CT, MRI or PET images ( 17 – 19 ). Therefore, it can objectively reflect the heterogeneity of different lesions. Previous studies have proved that radiomics can effectively predict the histopathological grades and differentiation of several tumors such as soft tissue tumors, cholangiocarcinoma, glioma, pancreatic cancer and breast cancer ( 20 – 24 ). In addition, previous studies have proved that MRI based radiomics can be used for prediction of lymph node metastasis of gastric cancer ( 25 ). Therefore, the purpose of this research is to use the MRI-based radiomics to predict gastric cancer with different DD. Material and methods The flow chart of this study is shown in Fig. 1 . First, we retrospectively collected patient information and obtained MRI images. Secondly, tumor features (shape features, first-order features and texture features) are extracted from MRI images. For the independent prediction of the degree of differentiation of gastric cancer, the machine-based radiation grouping is divided into two steps, including feature selection and feature classification. Data cohort This retrospective study was approved by our Institutional Review Board,since DWI and ADC are part of the routine MRI protocol for preoperative diagnosis of gastric cancer patients, the right of informed consent was waived.Patients who underwent preoperative abdominal MRI from January 2018 to January 2022 and met the following inclusion criteria: 1) biopsy or operation after MRI; 2) Complete imaging and pathological data; 3) No neoadjuvant chemotherapy or radiotherapy before MRI; 4) Lesions more than 1 cm in diameter. Due to small sample size of moderately differentiated gastric cancer, we divided cases into poorly differentiated gastric cancer group (Group 1: 131 poorly differentiated gastric cancer cases) and highly differentiated gastric cancer group (Group 2: 48 moderately differentiated gastric cancer cases and 60 highly differentiated gastric cancer cases). The 239 study cases were randomly assigned to the training (n = 167,70%) and testing (n = 72,30%) datasets at a ratio of 7:3. MRI Imaging Sequence This prospective study was approved by the clinical research ethics committee of Shandong Provincial Hospital (ID: SDPH 3456792). All procedures conducted in studies involving human participants complied with the ethical standards of institutions and/or national research committees as well as the 1964 Helsinki Declaration and its subsequent amendments or similar ethical standards. All individual participants in the study provided informed consent. All preoperative abdominal MRI examinations were performed with a 3.0T MRI (Prisma, Siemens Medical Systems, Germany) equipped with 18 channel body phased array coils. Before examination, the patient received calm breathing exercise training and drank 800–1000 ml water. The scan ranged from the top of the transverse septum to the lowest edge of the liver (including the whole stomach). The scanning sequences included T1WI, T2WI, DWI and enhanced T1WI sequences. The scanning parameters are shown in Table 1 . VOI (3D volume of interest) Segmentation and measurements of ADC mean and ADC min value All MRI images were obtained by 3.0 T MRI with same scanning parameters. 3D volume of interest was mapped by ITK-SNAP software (version 3.6; www.itksnap . Org). With T2WI image and enhanced T1WI image as reference, two experienced radiologists jointly outlined the edge of the tumor lesion layer by layer on the axial DWI image (both of radiologists did not know the pathological information of all patients) and fused them into a VOI. VOI includes the inner edge of the lesion on each layer and avoids necrotic tissue and surrounding adipose tissue (As is shown in Fig. 2 .). Then, the tumor VOI layered surface on the DWI sequence is mapped to the same level of the corresponding ADC map and then the VOI of ADC map is obtained. When outlining VOI, two radiologists measured the ADC value of ROI at each layer and obtained the measurement of ADC mean value and ADC min value. The calculation formula of ADC value is as follows: Eq: ADC = (lnS0/S1) /(b1-b0) Feature extraction A total of 239 patients with gastric cancer (mean age 43 years, range 21–73 years) were analyzed. Seven feature groups are extracted by FAEPro V0.3.7, which including First-order statistics, Gray-level Co-occurrence Matrix (GLCX), Gray Level Run Length Matrix (GLRLM), Gray Level Size Zone Matrix (GLSZM), Gray Level Dependence Matrix (GLDM), Neighbouring Gray Tone Difference Matrix (NGTDM) and Shape-based feature. As a result, 2632 features were extracted from DWI and ADC image of each patient. Feature selection and classification First, we classified poorly differentiated gastric cancer as positive meanwhile classified highly differentiated gastric cancer as negative.We selected 167 cases as the training data set (91/76 = positive/negative)). And we selected the other 72 cases as the testing data set (40/32 = positive/negative). To Remove the unbalance of the training data set, we up-samples by repeating random cases to to make positive/negative samples balance. We applied the normalization on the feature matrix. For each feature vector, we calculated the mean value and the standard deviation. Each feature vector was subtracted by the mean value and was divided by the standard deviation. After normalization process, each vector has zero center and unit standard deviation. Since the dimension of feature space was high, we compared the similarity of each feature pair. If the PCC value of the feature pair was larger than 0.99, we removed one of them. After this process, the dimension of the feature space was reduced and each feature was independent to each other. Least Absolute Shrinkage and Selection Operator [LASSO], Analysis Of Variance [ANOVA], Recursive Feature Elimination [RFE] and Kruskal Wallis[KW] were selected as common methods to explore the significant features corresponding to the labels respectively( 26 – 28 ). F-value was calculated to evaluate the relationship between features and the label. We sorted features according to the corresponding F-value and selected top 99 features according to validation performance. Support Vector Machine [SVM], Linear Discriminant Analysis [LDA] and Logistic Regression [LR] were used as feature classifiers.LR is a linear classifier that combines all the features. LDA is an linear classifier by fitting class conditional densities to the data and using Bayes’rule. SVM is an effective and robust classifier to build the model( 29 – 31 ).The kernel function has the ability to map the features into a higher dimension to search the hyper-plane for separating the cases with different labels.L1 norm is added in the final lost function and the weights was constrained, which make the features sparse. To determine the hyper-parameter (e.g. the number of features) of model, we applied cross validation with 5-fold on the training data set. The hyper-parameters were set according to the model performance on the validation data set. The performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis. The area under the ROC curve (AUC) was calculated for quantification. The accuracy (Acc), sensitivity and specificity were also calculated at a cutoff value that maximized the value of the Yorden index. We also estimated the 95% confidence interval by bootstrape with 1000 samples.The algorithm with the highest AUC and accuracy was identified as best radiomics model. All above processes were implemented with FeAture Explorer Pro (FAEPro, V 0.3.7) on Python (3.7.6)。 Statistical analysis All statistical analyses were performed using IBM SPSS statistical data (v.23.0, Armonk, NY). Independent t-test was used to verify whether there were significant differences in age, gender, height, weight, tumor volume, location, Lauren classification, pathological classification, MRI reporting T-stage, MRI reporting N-stage, ADC mean value and ADC min value between patients with poorly differentiated gastric cancer and patients with highly differentiated gastric cancer. The confidence interval P < 0.05 was considered statistically significant. Result Clinical Characteristics of the Patients The clinicopathological features of all patients are shown in Table 2. There was no significant difference in age, gender, height, weight, tumor volume, location, Lauren classification, pathological classification, MRI reporting T-stage, MRI reporting N-stage and ADC mean between the two groups ( P > 0.05). The ADC min in poorly differentiated gastric cancer group was significantly lower than that in highly differentiated gastric cancer group (Table 2) ( P < 0.05). Diagnostic efficiency 2632 features are extracted from DWI and ADC images. We apply four feature selection methods respectively, sort each feature according to the corresponding F value and select the first 99 features according to the verification performance. For the the above four feature selection methods, we applied three feature classification methods to cross combine and constructed 12 radiomics models. The results are shown in Table 3. Among all of the models, the combination that feature selection method Lasso and feature classification method SVM has highest diagnostic efficiency with AUC value = 0.854, Acc value = 0.793, sensitivity = 0.765 and specificity = 0.772. Best feature selection The corresponding DD radiomics with the best AUC (feature selection method Lasso and feature classification method SVM) included 10 features of 4 ADC features and 6 DWI features. These 10 features include 8 texture features and 2 first-order features (Fig. 3B and Fig. 3C). In terms of differential diagnosis of poorly and highly differentiated gastric cancer, the combination of ADC min and radiomics has highest diagnostic efficiency with AUC value = 0.919, Acc value = 0.823, sensitivity = 0.853 and specificity = 0.834 (Fig. 4, Table 4). Discussion Gastric cancer is a malignant tumor originated from gastrointestinal mucosal epithelium. At present, the new annual cases of gastric cancer in China account for more than 40% of the world's new cases ( 32 ). Previous studies had proven that the lower of DD for gastric cancer, the worse of prognosis; therefore the differential diagnosis for DD of gastric cancer has important clinical significance ( 33 – 35 ). In this research, we selected four feature selection methods and three classification algorithms to construct radiomics features for DD of gastric cancer. The results show that combination of feature selection method Lasso and classification method SVM has best prediction performance with Acc = 0.85 and AUC = 0.90 (95% CI: 0.75–1.00). The results showed that the radiomics based on DWI and ADC images had great diagnostic performance in distinguishing the DD of gastric cancer. This also shows that it is of great significance to guide clinicians to formulate best treatment strategy for gastric cancer and improve its prognosis. At present, there are still few studies using the morphological features of MRI to distinguish gastric cancer with different DD. The possible reason is that relying solely on the subjective judgment of radiologists can not effectively distinguish between poorly and highly-grade gastric cancer. In contrast, radiomics can more quantitatively reflect information extracted from images and apply high-dimensional data to clinical decision-making ( 17 , 36 ). The heterogeneity of signal intensity is the main criterion for the diagnosis of malignant tumors. There are obvious heterogeneity in DWI and ADC images of tumors with different DD. The possible reason is that the proliferation of poorly differentiated gastric cancer cells is often strong and they tend to have a higher nucleocytoplasmic ratio. These mean that it will limit the free diffusion of water molecules by reducing the intracellular and extracellular space, which can also be reflected by DWI and ADC diagrams ( 13 , 37 , 38 , 39 ). Radiomics features include first-order features, texture features and shape features. Texture feature analysis can quantitatively reflect the internal texture and tissue distribution of tumor, which are difficult to be simply perceived by human vision. The first-order statistical features are used to reflect the distribution of gray intensity in tumors. Shape features reflect the shape, size and regularity of tumors ( 19 , 40 ). In this research, when comparing the performance of different feature selection methods and classification methods, we found that the radiomics model (Lasso as the feature selection method and SVM as the feature classification method) obtained the highest AUC value. Lasso is a better feature selecting method than other methods. The reason is that by designing a penalty function, the regression coefficients are compressed, so that some coefficients are equal to zero and then an ideal model is obtained. Therefore, Lasso can obtain the good characteristics of ridge regression and subset screening and realize the modeling and dimensionality reduction of binary classification data. In feature classification, SVM can provide better classification performance because it can use the existing information to achieve best results and has better ability in processing high-dimensional data ( 26 , 27 ). Based on the radiomics model of Lasso-SVM, by reducing the number of redundant features, an optimal feature subset with highest AUC value and Acc value is confirmed. This optimal subset includes 10 features. Among them, texture features account for 8/10, which include entropy, heterogeneity, covariance and energy. Entropy can reflect the complexity of texture in the image, non-uniformity reflects the non-uniformity of gray distribution and the irregularity of texture, covariance reflects the clarity and irregularity of texture and energy reflects the complexity of texture in the image ( 19 , 41 , 42 , 43 ). Due to excessive cell proliferation, compared with moderately and highly differentiated gastric cancer, there are more areas of cystic degeneration, necrosis and bleeding in poorly differentiated gastric cancer, which indicates that there is heterogeneity between poorly and highly differentiated gastric cancer ( 5 , 6 , 9 , 12 ). The analysis of entropy and non-uniformity shows that with the decline of gastric cancer differentiation, entropy and non-uniformity continue to rise, which indicates that entropy and non-uniformity can be used as important texture features reflecting tumor heterogeneity ( 23 , 44 ). In addition, with the decrease of the DD of gastric cancer, the higher the energy and covariance, which indicates that it can effectively reflect the heterogeneity of lesions with different DD ( 22 , 41 ). Therefore, the results show that texture features can play a more important role in distinguishing the heterogeneity of different differentiated gastric cancer. As a functional imaging index, ADC value mainly reflects the diffusion of water molecules inside and outside cells, which can provide a more reliable basis for clinical judgment of benign and malignant tumors and the degree of differentiation( 13 , 39 ). As is shown in the previous research results, the difference between poorly differentiated gastric cancer and moderately and highly differentiated gastric cancer is not statistically significant in the average ADC value. The possible reason is that the ADC mean value may include the small cystic and necrotic components in the tumor tissue, which will increase the ADC mean value and affect the results ( 45 ). For the ADC min , there was significant difference between the two groups. The possible reason is that compared with ADC mean , ADC min can better reflect the most heterogeneous components in different tumor tissues, such as the solid components of tumors ( 46 , 47 ). And this indicates that ADC min has the ability to distinguish tumor heterogeneity to a certain extent, which is consistent with the previous research results ( 48 ). In this research, the model combining ADC min and radiomics found that it has higher AUC and ACC values than the model of ADC min or radiomics alone. This indicates that the model combining ADC min and radiomics can better reflect the heterogeneity of tumors. This is of great significance for clinicians to make treatment plans for patients with gastric cancer in the future. Limitations However, there are still some limitations in this research. The first obvious limitation is that the number of sample sets is small and unbalanced. Although the combination of different feature selection methods and classification methods has preliminarily verified the improvement of diagnostic performance, more original data sets with balanced sample size between two groups are still needed to further reassess the proposed method. The second limitation is that all data sets in this study are from one clinical center. In the future, more samples from multiple centers are needed to distinguish between poorly, moderately and highly differentiated gastric cancer. Conclusion In conclusion, the combined model of radiomics and ADC min is helpful to distinguish gastric cancer with different DD, which is of great significance for the formulation of treatment plan for patients with gastric cancer. Abbreviations Magnetic resonance imaging, MRI; Least Absolute Shrinkage and Selection Operator, LASSO; Analysis Of Variance, ANOVA; Recursive Feature Elimination, RFE; Kruskal Wallis, KW; Support Vector Machine, SVM; Linear Discriminant Analysis, LDA; Logistic Regression, LR. Declarations Ethics approval and consent to participate The studies involving human participants were reviewed and approved by the clinical research ethics committee of Shandong Provincial Hospital (Approved number: SDPH 3456792). Written informed consent was obtained from all individual participants included in this study. All procedures complied with the Declaration of Helsinki. Consent for publication Not applicable. Availability of data and materials Through communication with the research participants, participants of this study did not agree for their data to be shared publicly, so supporting data is not available. Competing Interests The authors declare no competing financial interest. Funding This experiment did not receive any financial assistance. Authors' contributions Letian Yuan is responsible for data collection and post-processing; Yilin Wang is responsible for writing articles Acknowledgements ： This study thanks for Professor Lin from Shandong Provincial Hospital for his technical support. References Smyth EC, Nilsson M, Grabsch HI, et al. Gastric cancer. Lancet 2020; 396(10251): 635-648. Matsuoka T and Yashiro M. Biomarkers of gastric cancer: Current topics and future perspective. World J Gastroenterol 2018; 24(26): 2818-2832. Thrift AP and El-Serag HB. Burden of Gastric Cancer. Clin Gastroenterol Hepatol 2020; 18(3): 534-542. Gullo I, Grillo F, Mastracci L, et al. Precancerous lesions of the stomach, gastric cancer and hereditary gastric cancer syndromes. Pathological 2020; 112(3): 166-185. Selcukbiricik F, Buyukunal E, Tural D, et al. Clinicopathological features and outcomes of patients with gastric cancer: a single-center experience. World J Gastroenterol 2013; 19(14): 2154-2161. Machlowska J, Baj J, Sitarz M, et al. Gastric Cancer: Epidemiology, Risk Factors, Classification, Genomic Characteristics and Treatment Strategies. Int J Mol Sci 2020; 21(11): 4012-4022. Molina-Castro S, Pereira-Marques J, Figueiredo C, et al. Gastric cancer: Basic aspects. Helicobacter 2017; 22 Suppl 1:10.1111/hel.12412. Venerito M, Vasapolli R, Rokkas T, et al. Gastric cancer: epidemiology, prevention and therapy. Helicobacter 2018; 23 Suppl 1:e12518. Agnes A, Biondi A, Laurino A, et al. Global updates in the treatment of gastric cancer: a systematic review. Part 1: staging, classification and surgical treatment. Updates Surg 2020; 72(2): 341-353. Song Z, Wu Y, Yang J, et al. Progress in the treatment of advanced gastric cancer. Tumour Biol 2017; 39(7):1010428317714626. Digklia A and Wagner AD. Advanced gastric cancer: Current treatment landscape and future perspectives. World J Gastroenterol 2016; 22(8): 2403-2414. Li F, Sun Y, Huang J, et al. CD4/CD8 + T cells, DC subsets, Foxp3, and IDO expression are predictive indictors of gastric cancer prognosis. Cancer Med 2019; 8(17): 7330-7344. Zhang Y and Yu J. The role of MRI in the diagnosis and treatment of gastric cancer. Diagn Interv Radiol 2020; 26(3): 176-182. Borggreve AS, Goense L, Brenkman HJF, et al. Imaging strategies in the management of gastric cancer: current role and future potential of MRI. Br J Radiol 2019; 92(1097):20181044. Arocena MG, Barturen A, Bujanda L, et al. MRI and endoscopic ultrasonography in the staging of gastric cancer. Rev Esp Enferm Dig 2006; 98(8):582-590. Ma L, Xu X, Zhang M, et al. Dynamic contrast-enhanced MRI of gastric cancer: Correlations of the pharmacokinetic parameters with histological type, Lauren classification, and angiogenesis. Magn Reson Imaging 2017; 37:27-32. Litvin AA, Burkin DA, Kropinov AA, et al. Radiomics and Digital Image Texture Analysis in Oncology (Review). Sovrem Tekhnologii Med 2021; 13(2): 97-104. Tomaszewski MR and Gillies RJ. The Biological Meaning of Radiomic Features. Radiology 2021; 298(3): 505-516. Mayerhoefer ME, Materka A, Langs G, et al. Introduction to Radiomics. J Nucl Med 2020; 61(4): 488-495. Tang Y, Yang CM, Su S, et al. Machine learning-based Radiomics analysis for differentiation degree and lymphatic node metastasis of extrahepatic cholangiocarcinoma. BMC Cancer 2021; 21(1): 1268-1280. Malik N, Geraghty B, Dasgupta A, et al. MRI radiomics to differentiate between low grade glioma and glioblastoma peritumoral region. J Neurooncol 2021; 155(2): 181-191. Zhang Q, Peng Y, Liu W, et al. Radiomics Based on Multimodal MRI for the Differential Diagnosis of Benign and Malignant Breast Lesions. J Magn Reson Imaging 2020; 52(2): 596-607. Wang H, Zhang J, Bao S, et al. Preoperative MRI-Based Radiomic Machine-Learning Nomogram May Accurately Distinguish Between Benign and Malignant Soft-Tissue Lesions: A Two-Center Study. J Magn Reson Imaging 2020; 52(3): 873-882. Shi YJ, Zhu HT, Liu YL, et al. Radiomics Analysis Based on Diffusion Kurtosis Imaging and T2 Weighted Imaging for Differentiation of Pancreatic Neuroendocrine Tumors From Solid Pseudopapillary Tumors. Front Oncol 2020; 10:1624. Chen W, Wang S, Dong D, et al. Evaluation of Lymph Node Metastasis in Advanced Gastric Cancer Using Magnetic Resonance Imaging-Based Radiomics. Front Oncol 2019; 9: 1265. Tosin MC, Majolo M, Chedid R, et al. sEMG feature selection and classification using SVM-RFE. Annu Int Conf IEEE Eng Med Biol Soc 2017; 2017:390-393. Zhang S, Zhu F, Yu Q, et al. Identifying DNA-binding proteins based on multi-features and LASSO feature selection. Biopolymers 2021; 112(2): e23419. Schober P and Vetter TR. Analysis of Variance in Medical Research. Anesth Analg 2020; 131(2): 508-509. Niijima S and Okuno Y. Laplacian linear discriminant analysis approach to unsupervised feature selection. IEEE/ACM Trans Comput Biol Bioinform 2009; 6(4): 605-614. Adeli E, Li X, Kwon D, et al. Logistic Regression Confined by Cardinality-Constrained Sample and Feature Selection. IEEE Trans Pattern Anal Mach Intell 2020; 42(7):1713-1728. Tosin MC, Majolo M, Chedid R, et al. sEMG feature selection and classification using SVM-RFE. Annu Int Conf IEEE Eng Med Biol Soc 2017; 2017: 390-393. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021; 71(3): 209-249. Feng F, Liu J, Wang F, et al. Prognostic value of differentiation status in gastric cancer. BMC Cancer 2018; 18(1):865. Machlowska J, Maciejewski R, Sitarz R et al. The Pattern of Signatures in Gastric Cancer Prognosis. Int J Mol Sci 2018; 19(6):1658. Wu CW, Chi CW, Lin WC et al. Gastric cancer: prognostic and diagnostic advances. Expert Rev Mol Med 2002; 4(6):1-12. Piñeiro-Fiel M, Moscoso A, Pubul V, et al. A Systematic Review of PET Textural Analysis and Radiomics in Cancer. Diagnostics (Basel) 2021; 11(2):380 Giganti F, Tang L, Baba H et al. Gastric cancer and imaging biomarkers: Part 1 - a critical review of DWI-MRI and CE-MDCT findings. Eur Radiol 2019; 29(4):1743-1753. Borggreve AS, Goense L, Brenkman HJF, et al. Imaging strategies in the management of gastric cancer: current role and future potential of MRI. Br J Radiol 2019; 92(1097): 20181044. Joo I, Lee JM, Kim JH, et al. Prospective comparison of 3T MRI with diffusion-weighted imaging and MDCT for the preoperative TNM staging of gastric cancer. J Magn Reson Imaging 2015; 41(3):814-821. Yu Y, Tan Y, Xie C, et al. Development and Validation of a Preoperative Magnetic Resonance Imaging Radiomics-Based Signature to Predict Axillary Lymph Node Metastasis and Disease-Free Survival in Patients With Early-Stage Breast Cancer. JAMA network Open 2020; 3(12):e2028086. Wang H, Nie P, Wang Y, et al. Radiomics nomogram for differentiating between benign and malignant soft-tissue masses of the extremities. J Magn Reson Imaging 2020; 51(1):155-163. Wang H, Chen H, Duan S, et al. Radiomics and Machine Learning With Multiparametric Preoperative MRI May Accurately Predict the Histopathological Grades of Soft Tissue Sarcomas. J Magn Reson Imaging 2020; 51(3):791-797. Yip SSF and Aerts HJW. Applications and limitations of radiomics. Phys Med Biol 2016; 61(13): R150-R166. Liu S, Zheng H, Zhang Y, et al. Whole-volume apparent diffusion coefficient-based entropy parameters for assessment of gastric cancer aggressiveness. J Magn Reson Imaging 2018; 47(1): 168-175. Connolly M and Srinivasan A. Diffusion-Weighted Imaging in Head and Neck Cancer: Technique, Limitations and Applications. Magn Reson Imaging Clin N Am 2018; 26(1):121-133. Zhang Y, Chen J, Liu S, et al. Assessment of histological differentiation in gastric cancers using whole-volume histogram analysis of apparent diffusion coefficient maps. J Magn Reson Imaging (2016); 45:440–449. Sah BR, Owczarczyk K, Siddique M, et al. Radiomics in esophageal and gastric cancer. Abdominal Radiol (2018); 44: 2048–2058. Liu S, Guan W, Wang H, et al. Apparent diffusion coefficient value of gastric cancer by diffusion-weighted imaging: correlations with the histological differentiation and Lauren classification. Eur J Radiol 2014; 83(12): 2122-2128. Tables Table 2 is available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files Table2.docx Cite Share Download PDF Status: Published Journal Publication published 08 Dec, 2025 Read the published version in BMC Medical Imaging → Version 1 posted Editorial decision: Revision requested 30 Sep, 2025 Reviews received at journal 29 Sep, 2025 Reviews received at journal 29 Sep, 2025 Reviewers agreed at journal 28 Sep, 2025 Reviews received at journal 27 Sep, 2025 Reviewers agreed at journal 26 Sep, 2025 Reviewers agreed at journal 25 Sep, 2025 Reviewers agreed at journal 25 Sep, 2025 Reviewers agreed at journal 24 Sep, 2025 Reviewers agreed at journal 23 Sep, 2025 Reviewers agreed at journal 20 Sep, 2025 Reviewers invited by journal 15 Sep, 2025 Editor assigned by journal 10 Sep, 2025 Editor invited by journal 10 Sep, 2025 Submission checks completed at journal 09 Sep, 2025 First submitted to journal 09 Sep, 2025 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-7493820","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":518140257,"identity":"cff275ab-f7f8-4d70-b5ab-e6e9acd09f38","order_by":0,"name":"Yilin Wang","email":"","orcid":"","institution":"Capital Medical University Affiliated Chest Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yilin","middleName":"","lastName":"Wang","suffix":""},{"id":518140259,"identity":"484b80c5-de2c-4a62-9b76-c93186fac4bc","order_by":1,"name":"Letian Yuan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIie3PsQrCMBCA4ZRCXIKuVxR9hQMnwd3XSBA6KThWEEyppIOIq76Fo5tKoC5x1619BHFxcNBdMXVzyDffz90R4jh/iFYSnQcPmO6OcZzzaGJPqiwLsSO7njQ6wdxk9qQJA4SxDD157qmgmPklDmN7xMtW+97KU5GQlNTSObf8InmxNpr69VidxbZBwJw2ti37NlDNaOPwSgwlCENLAkLWH1QDA6FGQvllkj6BQIUIr4SUS1hGEUyXIzskwE3GrL+00uUthwg4VtLieo8mzVq6+J68Yb+NO47jOB89AZE4TdLP85muAAAAAElFTkSuQmCC","orcid":"","institution":"Shandong Provincial Hospital","correspondingAuthor":true,"prefix":"","firstName":"Letian","middleName":"","lastName":"Yuan","suffix":""}],"badges":[],"createdAt":"2025-08-30 08:08:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7493820/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7493820/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12880-025-02108-y","type":"published","date":"2025-12-08T15:58:45+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":92063509,"identity":"e16c5851-4079-45f2-a7c9-c11d2bb449ee","added_by":"auto","created_at":"2025-09-24 08:38:32","extension":"doc","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1084600,"visible":true,"origin":"","legend":"","description":"","filename":"MRIbasedradiomicstodistinguishgastriccancerwithdifferentdegreesofdifferentiation.doc","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/8786106fe86b3d0a2f0937b4.doc"},{"id":92063529,"identity":"7acb57ca-2733-47fe-9616-2b99f0264da9","added_by":"auto","created_at":"2025-09-24 08:38:33","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3103,"visible":true,"origin":"","legend":"","description":"","filename":"c45601dabe89448fbb871e6713a80f53.json","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/5751309b5e9e63fc585ecf3a.json"},{"id":92063480,"identity":"161409fc-443c-42d5-86b9-9effee0efacb","added_by":"auto","created_at":"2025-09-24 08:38:27","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":98784,"visible":true,"origin":"","legend":"","description":"","filename":"c45601dabe89448fbb871e6713a80f531enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/20a58a0dddce370daaf0f9c1.xml"},{"id":92063479,"identity":"f03a1784-2e9e-4079-b46a-3c7879f105eb","added_by":"auto","created_at":"2025-09-24 08:38:27","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":230597,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/625cacd27b221f89d79d5f52.jpeg"},{"id":92063511,"identity":"b70f5ee1-14d8-4902-a354-73010fd7001d","added_by":"auto","created_at":"2025-09-24 08:38:32","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":905508,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/8f2cab0fd7c523befca5a0eb.jpeg"},{"id":92063417,"identity":"b2f04b1f-c909-4622-8ccc-78e9b1f70008","added_by":"auto","created_at":"2025-09-24 08:38:18","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":459517,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/ada2f08bb87ae6175f51d362.jpeg"},{"id":92063469,"identity":"a2164b4f-34f0-415e-b105-53fd0e1eb780","added_by":"auto","created_at":"2025-09-24 08:38:26","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":247814,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/f73117ab2a58c39fe6fcf1fb.jpeg"},{"id":92063475,"identity":"b75dc812-b572-4d3e-9ff2-efff493453ec","added_by":"auto","created_at":"2025-09-24 08:38:26","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":47943,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/35d3c2e87ff1a7192f50c4fd.png"},{"id":92063543,"identity":"f674042b-e8c9-4712-9eeb-1d1c1275b3b3","added_by":"auto","created_at":"2025-09-24 08:38:35","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":150909,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/d04c44f530b865ed38921e9b.png"},{"id":92063552,"identity":"b8ce94ad-86e4-4634-81e1-06e2b603ca41","added_by":"auto","created_at":"2025-09-24 08:38:38","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":92091,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/733fb221cbc80daffa64cce1.png"},{"id":92063445,"identity":"71063a91-bd6b-4702-aca1-72a054a8373a","added_by":"auto","created_at":"2025-09-24 08:38:23","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40728,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/285e2bd7a4a8a90987a38e13.png"},{"id":92063449,"identity":"cbe7f2a4-58d8-4e13-a80d-8d6e566e421f","added_by":"auto","created_at":"2025-09-24 08:38:24","extension":"xml","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95717,"visible":true,"origin":"","legend":"","description":"","filename":"c45601dabe89448fbb871e6713a80f531structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/ee292220114c0962605d7fb0.xml"},{"id":92063456,"identity":"41826b59-0385-4980-8120-5a8127bbc317","added_by":"auto","created_at":"2025-09-24 08:38:24","extension":"html","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106835,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/4be3b5fdeb5a59b9326cf457.html"},{"id":92063433,"identity":"8b7de4dd-9eca-4f9f-8cc2-374b77b50e54","added_by":"auto","created_at":"2025-09-24 08:38:21","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":161754,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of radiomics implementation in this study.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/c80d1e898ca2aaa92074d947.png"},{"id":92063454,"identity":"b856f771-5265-4872-b1f5-88436a1bce70","added_by":"auto","created_at":"2025-09-24 08:38:24","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":512397,"visible":true,"origin":"","legend":"\u003cp\u003eROIs were placed on each section of the tumor, avoiding necrotic tissue and surrounding adipose tissue on DWI (A-C) and ADC (D-F) respectively.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/5de5c316802eb01925f2af4e.png"},{"id":92063538,"identity":"3bbcfebf-123a-438b-b09e-ce586bdcf54d","added_by":"auto","created_at":"2025-09-24 08:38:34","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":317651,"visible":true,"origin":"","legend":"\u003cp\u003eA: The best prognostic algorithms: feature selection method LASSO + classifier SVM. The AUC of validation cohort = 0.820, AUC of testing cohort = 0.855, AUC of training cohort = 0.832.\u003c/p\u003e\n\u003cp\u003eB: Results of the optimal feature subset selection for differentiating poorly differentiated gastric cancer from highly differentiated gastric cancer with DWI and ADC images. The red arrow on the bottom left indicates the 10 top-ranked features achieving the highest AUC value (0.854) are determined as the optimal subset.\u003c/p\u003e\n\u003cp\u003eC: Components and ranking of optimal feature subset.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/0ec3723e5bd7fad67acbe608.png"},{"id":92063453,"identity":"68990578-617f-42e6-a2e8-571758b3a1a6","added_by":"auto","created_at":"2025-09-24 08:38:24","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":152679,"visible":true,"origin":"","legend":"\u003cp\u003eROC curve of different models for distinguishing poorly and highly differentiated gastric cancers. A: Radiomics+ADC\u003csub\u003emin\u003c/sub\u003e; B: Radiomics ; C:ADC\u003csub\u003emin.\u003c/sub\u003e\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/0fe69e5f6f793f1819ecd4bd.png"},{"id":98243963,"identity":"b3bc455a-f301-4248-bbdc-6fae0ef4d94c","added_by":"auto","created_at":"2025-12-15 16:11:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1818741,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/9ca7ce04-1d3c-4fb6-babd-ded9dddcc6bd.pdf"},{"id":92063483,"identity":"c1523c14-523a-4f57-a41f-dca815256a6f","added_by":"auto","created_at":"2025-09-24 08:38:27","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":16741,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.docx","url":"https://assets-eu.researchsquare.com/files/rs-7493820/v1/a2296fd9246afc5eb869ff38.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"MRI-based Radiomics analysis for differentiation degree of gastric cancer","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGastric cancer is the fifth most common tumor with third mortality in the world. It was reported that the incidence rate and mortality of gastric cancer (GC) have gradually increased worldwide over the past decades. Every year, more than one million people worldwide have been diagnosed with gastric cancer. (\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). Gastric cancer (GC) is a malignant tumor originated from gastrointestinal mucosal epithelium and gastric adenocarcinoma accounts for more than 90% of all GC (\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). At present, the most effective treatment for gastric cancer is completely surgical resection. However, it is usually suitable for patients with early gastric cancer (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). For patients with poorly differentiated or advanced gastric cancer, even if the tumor is completely removed, most patients may encounter poor prognosis (e.g. local recurrence or distant metastasis) and the prognosis is associated with the differentiation degree (DD) and lymph node metastasis (LNM) (\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Therefore, accurate preoperative evaluation of DD of gastric cancer is very important for selecting the treatment strategy and predicting the prognosis.\u003c/p\u003e\u003cp\u003eSeveral imaging examination techniques could diagnose gastric cancer, including computerized tomography (CT), magnetic resonance imaging (MRI) and gastroscope. Due to the advantage of high soft tissue resolution, MRI can better show the degree of gastric mucosal invasion and lymph node metastasis. Therefore, it has become an effective means for the diagnosis, staging and efficacy evaluation of gastric cancer (\u003cspan additionalcitationids=\"CR14 CR15\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). However, it is hard to determine the DD of gastric cancer, which can not rely on the subjective evaluation of radiologists. Therefore, it is of great clinical significance to find a noninvasive examination method that can effectively evaluate the DD of gastric cancer before operation.\u003c/p\u003e\u003cp\u003eAs a new methodology, radiomics can quantitatively extract the algorithmic features of imaging data, such as CT, MRI or PET images (\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e). Therefore, it can objectively reflect the heterogeneity of different lesions. Previous studies have proved that radiomics can effectively predict the histopathological grades and differentiation of several tumors such as soft tissue tumors, cholangiocarcinoma, glioma, pancreatic cancer and breast cancer (\u003cspan additionalcitationids=\"CR21 CR22 CR23\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). In addition, previous studies have proved that MRI based radiomics can be used for prediction of lymph node metastasis of gastric cancer (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). Therefore, the purpose of this research is to use the MRI-based radiomics to\u003c/p\u003e\u003cp\u003epredict gastric cancer with different DD.\u003c/p\u003e"},{"header":"Material and methods","content":"\u003cp\u003eThe flow chart of this study is shown in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. First, we retrospectively collected patient information and obtained MRI images. Secondly, tumor features (shape features, first-order features and texture features) are extracted from MRI images. For the independent prediction of the degree of differentiation of gastric cancer, the machine-based radiation grouping is divided into two steps, including feature selection and feature classification.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003eData cohort\u003c/h2\u003e\n \u003cp\u003eThis retrospective study was approved by our Institutional Review Board,since DWI and ADC are part of the routine MRI protocol for preoperative diagnosis of gastric cancer patients, the right of informed consent was waived.Patients who underwent preoperative abdominal MRI from January 2018 to January 2022 and met the following inclusion criteria: 1) biopsy or operation after MRI; 2) Complete imaging and pathological data; 3) No neoadjuvant chemotherapy or radiotherapy before MRI; 4) Lesions more than 1 cm in diameter. Due to small sample size of moderately differentiated gastric cancer, we divided cases into poorly differentiated gastric cancer group (Group 1: 131 poorly differentiated gastric cancer cases) and highly differentiated gastric cancer group (Group 2: 48 moderately differentiated gastric cancer cases and 60 highly differentiated gastric cancer cases). The 239 study cases were randomly assigned to the training (n\u0026thinsp;=\u0026thinsp;167,70%) and testing (n\u0026thinsp;=\u0026thinsp;72,30%) datasets at a ratio of 7:3.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eMRI Imaging Sequence\u003c/h3\u003e\n\u003cp\u003eThis prospective study was approved by the clinical research ethics committee of Shandong Provincial Hospital (ID: SDPH 3456792). All procedures conducted in studies involving human participants complied with the ethical standards of institutions and/or national research committees as well as the 1964 Helsinki Declaration and its subsequent amendments or similar ethical standards. All individual participants in the study provided informed consent.\u003c/p\u003e\n\u003cp\u003eAll preoperative abdominal MRI examinations were performed with a 3.0T MRI (Prisma, Siemens Medical Systems, Germany) equipped with 18 channel body phased array coils. Before examination, the patient received calm breathing exercise training and drank 800\u0026ndash;1000 ml water. The scan ranged from the top of the transverse septum to the lowest edge of the liver (including the whole stomach). The scanning sequences included T1WI, T2WI, DWI and enhanced T1WI sequences. The scanning parameters are shown in Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\n\u003ch3\u003e\u003cem\u003eVOI (3D volume of interest) Segmentation and measurements of ADC\u003csub\u003emean\u0026nbsp;\u003c/sub\u003eand ADC\u003csub\u003emin\u003c/sub\u003e value\u003c/em\u003e\u003c/h3\u003e\n\u003cp\u003eAll MRI images were obtained by 3.0 T MRI with same scanning parameters. 3D volume of interest was mapped by ITK-SNAP software (version 3.6; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.itksnap\u003c/span\u003e\u003c/span\u003e. Org). With T2WI image and enhanced T1WI image as reference, two experienced radiologists jointly outlined the edge of the tumor lesion layer by layer on the axial DWI image (both of radiologists did not know the pathological information of all patients) and fused them into a VOI. VOI includes the inner edge of the lesion on each layer and avoids necrotic tissue and surrounding adipose tissue (As is shown in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e.). Then, the tumor VOI layered surface on the DWI sequence is mapped to the same level of the corresponding ADC map and then the VOI of ADC map is obtained. When outlining VOI, two radiologists measured the ADC value of ROI at each layer and obtained the measurement of ADC\u003csub\u003emean\u003c/sub\u003e value and ADC\u003csub\u003emin\u003c/sub\u003e value.\u003c/p\u003e\n\u003cp\u003eThe calculation formula of ADC value is as follows:\u003c/p\u003e\n\u003cp\u003eEq: ADC = (lnS0/S1) /(b1-b0)\u003c/p\u003e\n\u003ch3\u003eFeature extraction\u003c/h3\u003e\n\u003cp\u003eA total of 239 patients with gastric cancer (mean age 43 years, range 21\u0026ndash;73 years) were analyzed. Seven feature groups are extracted by FAEPro V0.3.7, which including First-order statistics, Gray-level Co-occurrence Matrix (GLCX), Gray Level Run Length Matrix (GLRLM), Gray Level Size Zone Matrix (GLSZM), Gray Level Dependence Matrix (GLDM), Neighbouring Gray Tone Difference Matrix (NGTDM) and Shape-based feature. As a result, 2632 features were extracted from DWI and ADC image of each patient.\u003c/p\u003e\n\u003ch3\u003eFeature selection and classification\u003c/h3\u003e\n\u003cp\u003eFirst, we classified poorly differentiated gastric cancer as positive meanwhile classified highly differentiated gastric cancer as negative.We selected 167 cases as the training data set (91/76\u0026thinsp;=\u0026thinsp;positive/negative)). And we selected the other 72 cases as the testing data set (40/32\u0026thinsp;=\u0026thinsp;positive/negative).\u003c/p\u003e\n\u003cp\u003eTo Remove the unbalance of the training data set, we up-samples by repeating random cases to to make positive/negative samples balance. We applied the normalization on the feature matrix. For each feature vector, we calculated the mean value and the standard deviation. Each feature vector was subtracted by the mean value and was divided by the standard deviation. After normalization process, each vector has zero center and unit standard deviation. Since the dimension of feature space was high, we compared the similarity of each feature pair. If the PCC value of the feature pair was larger than 0.99, we removed one of them. After this process, the dimension of the feature space was reduced and each feature was independent to each other. Least Absolute Shrinkage and Selection Operator [LASSO], Analysis Of Variance [ANOVA], Recursive Feature Elimination [RFE] and Kruskal Wallis[KW] were selected as common methods to explore the significant features corresponding to the labels respectively(\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e). F-value was calculated to evaluate the relationship between features and the label. We sorted features according to the corresponding F-value and selected top 99 features according to validation performance.\u003c/p\u003e\n\u003cp\u003eSupport Vector Machine [SVM], Linear Discriminant Analysis [LDA] and Logistic Regression [LR] were used as feature classifiers.LR is a linear classifier that combines all the features. LDA is an linear classifier by fitting class conditional densities to the data and using Bayes\u0026rsquo;rule. SVM is an effective and robust classifier to build the model(\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e).The kernel function has the ability to map the features into a higher dimension to search the hyper-plane for separating the cases with different labels.L1 norm is added in the final lost function and the weights was constrained, which make the features sparse. To determine the hyper-parameter (e.g. the number of features) of model, we applied cross validation with 5-fold on the training data set. The hyper-parameters were set according to the model performance on the validation data set.\u003c/p\u003e\n\u003cp\u003eThe performance of the model was evaluated using receiver operating characteristic (ROC) curve analysis. The area under the ROC curve (AUC) was calculated for quantification. The accuracy (Acc), sensitivity and specificity were also calculated at a cutoff value that maximized the value of the Yorden index. We also estimated the 95% confidence interval by bootstrape with 1000 samples.The algorithm with the highest AUC and accuracy was identified as best radiomics model. All above processes were implemented with FeAture Explorer Pro (FAEPro, V 0.3.7) on Python (3.7.6)。\u003c/p\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003eStatistical analysis\u003c/h2\u003e\n \u003cp\u003eAll statistical analyses were performed using IBM SPSS statistical data (v.23.0, Armonk, NY). Independent t-test was used to verify whether there were significant differences in age, gender, height, weight, tumor volume, location, Lauren classification, pathological classification, MRI reporting T-stage, MRI reporting N-stage, ADC\u003csub\u003emean\u003c/sub\u003e value and ADC\u003csub\u003emin\u003c/sub\u003e value between patients with poorly differentiated gastric cancer and patients with highly differentiated gastric cancer. The confidence interval \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Result","content":"\u003cdiv id=\"Sec10\"\u003e\n \u003ch2\u003eClinical Characteristics of the Patients\u003c/h2\u003e\n \u003cp\u003eThe clinicopathological features of all patients are shown in Table\u0026nbsp;2. There was no significant difference in age, gender, height, weight, tumor volume, location, Lauren classification, pathological classification, MRI reporting T-stage, MRI reporting N-stage and ADC\u003csub\u003emean\u003c/sub\u003e between the two groups (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). The ADC\u003csub\u003emin\u003c/sub\u003e in poorly differentiated gastric cancer group was significantly lower than that in highly differentiated gastric cancer group (Table 2) (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\"\u003e\n \u003ch2\u003eDiagnostic efficiency\u003c/h2\u003e\n \u003cp\u003e2632 features are extracted from DWI and ADC images. We apply four feature selection methods respectively, sort each feature according to the corresponding F value and select the first 99 features according to the verification performance. For the the above four feature selection methods, we applied three feature classification methods to cross combine and constructed 12 radiomics models. The results are shown in Table 3. Among all of the models, the combination that feature selection method Lasso and feature classification method SVM has highest diagnostic efficiency with AUC value\u0026thinsp;=\u0026thinsp;0.854, Acc value\u0026thinsp;=\u0026thinsp;0.793, sensitivity\u0026thinsp;=\u0026thinsp;0.765 and specificity\u0026thinsp;=\u0026thinsp;0.772.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\"\u003e\n \u003ch2\u003eBest feature selection\u003c/h2\u003e\n \u003cp\u003eThe corresponding DD radiomics with the best AUC (feature selection method Lasso and feature classification method SVM) included 10 features of 4 ADC features and 6 DWI features. These 10 features include 8 texture features and 2 first-order features (Fig.\u0026nbsp;3B and Fig.\u0026nbsp;3C).\u003c/p\u003e\n \u003cp\u003eIn terms of differential diagnosis of poorly and highly differentiated gastric cancer, the combination of ADC\u003csub\u003emin\u003c/sub\u003e and radiomics has highest diagnostic efficiency with AUC value\u0026thinsp;=\u0026thinsp;0.919, Acc value\u0026thinsp;=\u0026thinsp;0.823, sensitivity\u0026thinsp;=\u0026thinsp;0.853 and specificity\u0026thinsp;=\u0026thinsp;0.834 (Fig. 4, Table 4).\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eGastric cancer is a malignant tumor originated from gastrointestinal mucosal epithelium. At present, the new annual cases of gastric cancer in China account for more than 40% of the world's new cases (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). Previous studies had proven that the lower of DD for gastric cancer, the worse of prognosis; therefore the differential diagnosis for DD of gastric cancer has important clinical significance (\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e). In this research, we selected four feature selection methods and three classification algorithms to construct radiomics features for DD of gastric cancer. The results show that combination of feature selection method Lasso and classification method SVM has best prediction performance with Acc\u0026thinsp;=\u0026thinsp;0.85 and AUC\u0026thinsp;=\u0026thinsp;0.90 (95% CI: 0.75\u0026ndash;1.00). The results showed that the radiomics based on DWI and ADC images had great diagnostic performance in distinguishing the DD of gastric cancer. This also shows that it is of great significance to guide clinicians to formulate best treatment strategy for gastric cancer and improve its prognosis.\u003c/p\u003e\u003cp\u003eAt present, there are still few studies using the morphological features of MRI to distinguish gastric cancer with different DD. The possible reason is that relying solely on the subjective judgment of radiologists can not effectively distinguish between poorly and highly-grade gastric cancer. In contrast, radiomics can more quantitatively reflect information extracted from images and apply high-dimensional data to clinical decision-making (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe heterogeneity of signal intensity is the main criterion for the diagnosis of malignant tumors. There are obvious heterogeneity in DWI and ADC images of tumors with different DD. The possible reason is that the proliferation of poorly differentiated gastric cancer cells is often strong and they tend to have a higher nucleocytoplasmic ratio. These mean that it will limit the free diffusion of water molecules by reducing the intracellular and extracellular space, which can also be reflected by DWI and ADC diagrams (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRadiomics features include first-order features, texture features and shape features. Texture feature analysis can quantitatively reflect the internal texture and tissue distribution of tumor, which are difficult to be simply perceived by human vision. The first-order statistical features are used to reflect the distribution of gray intensity in tumors. Shape features reflect the shape, size and regularity of tumors (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn this research, when comparing the performance of different feature selection methods and classification methods, we found that the radiomics model (Lasso as the feature selection method and SVM as the feature classification method) obtained the highest AUC value. Lasso is a better feature selecting method than other methods. The reason is that by designing a penalty function, the regression coefficients are compressed, so that some coefficients are equal to zero and then an ideal model is obtained. Therefore, Lasso can obtain the good characteristics of ridge regression and subset screening and realize the modeling and dimensionality reduction of binary classification data. In feature classification, SVM can provide better classification performance because it can use the existing information to achieve best results and has better ability in processing high-dimensional data (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBased on the radiomics model of Lasso-SVM, by reducing the number of redundant features, an optimal feature subset with highest AUC value and Acc value is confirmed. This optimal subset includes 10 features. Among them, texture features account for 8/10, which include entropy, heterogeneity, covariance and energy. Entropy can reflect the complexity of texture in the image, non-uniformity reflects the non-uniformity of gray distribution and the irregularity of texture, covariance reflects the clarity and irregularity of texture and energy reflects the complexity of texture in the image (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). Due to excessive cell proliferation, compared with moderately and highly differentiated gastric cancer, there are more areas of cystic degeneration, necrosis and bleeding in poorly differentiated gastric cancer, which indicates that there is heterogeneity between poorly and highly differentiated gastric cancer (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). The analysis of entropy and non-uniformity shows that with the decline of gastric cancer differentiation, entropy and non-uniformity continue to rise, which indicates that entropy and non-uniformity can be used as important texture features reflecting tumor heterogeneity (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e). In addition, with the decrease of the DD of gastric cancer, the higher the energy and covariance, which indicates that it can effectively reflect the heterogeneity of lesions with different DD (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e). Therefore, the results show that texture features can play a more important role in distinguishing the heterogeneity of different differentiated gastric cancer.\u003c/p\u003e\u003cp\u003eAs a functional imaging index, ADC value mainly reflects the diffusion of water molecules inside and outside cells, which can provide a more reliable basis for clinical judgment of benign and malignant tumors and the degree of differentiation(\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). As is shown in the previous research results, the difference between poorly differentiated gastric cancer and moderately and highly differentiated gastric cancer is not statistically significant in the average ADC value. The possible reason is that the ADC\u003csub\u003emean\u003c/sub\u003e value may include the small cystic and necrotic components in the tumor tissue, which will increase the ADC\u003csub\u003emean\u003c/sub\u003e value and affect the results (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e). For the ADC\u003csub\u003emin\u003c/sub\u003e, there was significant difference between the two groups. The possible reason is that compared with ADC\u003csub\u003emean\u003c/sub\u003e, ADC\u003csub\u003emin\u003c/sub\u003e can better reflect the most heterogeneous components in different tumor tissues, such as the solid components of tumors (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e). And this indicates that ADC\u003csub\u003emin\u003c/sub\u003e has the ability to distinguish tumor heterogeneity to a certain extent, which is consistent with the previous research results (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e). In this research, the model combining ADC\u003csub\u003emin\u003c/sub\u003e and radiomics found that it has higher AUC and ACC values than the model of ADC\u003csub\u003emin\u003c/sub\u003e or radiomics alone. This indicates that the model combining ADC\u003csub\u003emin\u003c/sub\u003e and radiomics can better reflect the heterogeneity of tumors. This is of great significance for clinicians to make treatment plans for patients with gastric cancer in the future.\u003c/p\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eLimitations\u003c/h2\u003e\u003cp\u003eHowever, there are still some limitations in this research. The first obvious limitation is that the number of sample sets is small and unbalanced. Although the combination of different feature selection methods and classification methods has preliminarily verified the improvement of diagnostic performance, more original data sets with balanced sample size between two groups are still needed to further reassess the proposed method. The second limitation is that all data sets in this study are from one clinical center. In the future, more samples from multiple centers are needed to distinguish between poorly, moderately and highly differentiated gastric cancer.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn conclusion, the combined model of radiomics and ADC\u003csub\u003emin\u003c/sub\u003e is helpful to distinguish gastric cancer with different DD, which is of great significance for the formulation of treatment plan for patients with gastric cancer.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eMagnetic resonance imaging, MRI; Least Absolute Shrinkage and Selection Operator, LASSO; Analysis Of Variance, ANOVA; Recursive Feature Elimination, RFE; Kruskal Wallis, KW; Support Vector Machine, SVM; Linear Discriminant Analysis, LDA; Logistic Regression, LR.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe studies involving human participants were reviewed and approved by\u0026nbsp;the clinical research ethics committee of Shandong Provincial Hospital (Approved number: SDPH 3456792).\u0026nbsp;Written informed consent was obtained from all individual participants included in this study. All procedures complied with the Declaration of Helsinki.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThrough communication with the research participants, participants of this study did not agree for their data to be shared publicly, so supporting data is not available.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing financial interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis experiment did not receive any financial assistance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLetian Yuan is responsible for data collection and post-processing; Yilin Wang is\u003c/p\u003e\n\u003cp\u003eresponsible for writing articles\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003cstrong\u003e：\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study thanks for Professor Lin from Shandong Provincial Hospital for his technical support.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSmyth EC, Nilsson M, Grabsch HI, et al. Gastric cancer. Lancet 2020; 396(10251): 635-648. \u003c/li\u003e\n\u003cli\u003eMatsuoka T and Yashiro M. Biomarkers of gastric cancer: Current topics and future perspective. World J Gastroenterol 2018; 24(26): 2818-2832.\u003c/li\u003e\n\u003cli\u003eThrift AP and El-Serag HB. Burden of Gastric Cancer. Clin Gastroenterol Hepatol 2020; 18(3): 534-542.\u003c/li\u003e\n\u003cli\u003eGullo I, Grillo F, Mastracci L, et al. Precancerous lesions of the stomach, gastric cancer and hereditary gastric cancer syndromes. Pathological 2020; 112(3): 166-185.\u003c/li\u003e\n\u003cli\u003eSelcukbiricik F, Buyukunal E, Tural D, et al. Clinicopathological features and outcomes of patients with gastric cancer: a single-center experience. World J Gastroenterol 2013; 19(14): 2154-2161. \u003c/li\u003e\n\u003cli\u003eMachlowska J, Baj J, Sitarz M, et al. Gastric Cancer: Epidemiology, Risk Factors, Classification, Genomic Characteristics and Treatment Strategies. Int J Mol Sci 2020; 21(11): 4012-4022. \u003c/li\u003e\n\u003cli\u003eMolina-Castro S, Pereira-Marques J, Figueiredo C, et al. Gastric cancer: Basic aspects. Helicobacter 2017; 22 Suppl 1:10.1111/hel.12412.\u003c/li\u003e\n\u003cli\u003eVenerito M, Vasapolli R, Rokkas T, et al. Gastric cancer: epidemiology, prevention and therapy. Helicobacter 2018; 23 Suppl 1:e12518.\u003c/li\u003e\n\u003cli\u003eAgnes A, Biondi A, Laurino A, et al. Global updates in the treatment of gastric cancer: a systematic review. Part 1: staging, classification and surgical treatment. Updates Surg 2020; 72(2): 341-353.\u003c/li\u003e\n\u003cli\u003eSong Z, Wu Y, Yang J, et al. Progress in the treatment of advanced gastric cancer. Tumour Biol 2017; 39(7):1010428317714626.\u003c/li\u003e\n\u003cli\u003eDigklia A and Wagner AD. Advanced gastric cancer: Current treatment landscape and future perspectives. World J Gastroenterol 2016; 22(8): 2403-2414.\u003c/li\u003e\n\u003cli\u003eLi F, Sun Y, Huang J, et al. CD4/CD8 + T cells, DC subsets, Foxp3, and IDO expression are predictive indictors of gastric cancer prognosis. Cancer Med 2019; 8(17): 7330-7344.\u003c/li\u003e\n\u003cli\u003eZhang Y and Yu J. The role of MRI in the diagnosis and treatment of gastric cancer. Diagn Interv Radiol 2020; 26(3): 176-182. \u003c/li\u003e\n\u003cli\u003eBorggreve AS, Goense L, Brenkman HJF, et al. Imaging strategies in the management of gastric cancer: current role and future potential of MRI. Br J Radiol\u003cem\u003e \u003c/em\u003e2019; 92(1097):20181044.\u003c/li\u003e\n\u003cli\u003eArocena MG, Barturen A, Bujanda L, et al. MRI and endoscopic ultrasonography in the staging of gastric cancer. Rev Esp Enferm Dig 2006; 98(8):582-590.\u003c/li\u003e\n\u003cli\u003eMa L, Xu X, Zhang M, et al. Dynamic contrast-enhanced MRI of gastric cancer: Correlations of the pharmacokinetic parameters with histological type, Lauren classification, and angiogenesis. Magn Reson Imaging 2017; 37:27-32.\u003c/li\u003e\n\u003cli\u003eLitvin AA, Burkin DA, Kropinov AA, et al. Radiomics and Digital Image Texture Analysis in Oncology (Review). Sovrem Tekhnologii Med 2021; 13(2): 97-104.\u003c/li\u003e\n\u003cli\u003eTomaszewski MR and Gillies RJ. The Biological Meaning of Radiomic Features. Radiology 2021; 298(3): 505-516. \u003c/li\u003e\n\u003cli\u003eMayerhoefer ME, Materka A, Langs G, et al. Introduction to Radiomics. J Nucl Med\u003cem\u003e \u003c/em\u003e2020; 61(4): 488-495.\u003c/li\u003e\n\u003cli\u003eTang Y, Yang CM, Su S, et al. Machine learning-based Radiomics analysis for differentiation degree and lymphatic node metastasis of extrahepatic cholangiocarcinoma. BMC Cancer\u003cem\u003e \u003c/em\u003e2021; 21(1): 1268-1280.\u003c/li\u003e\n\u003cli\u003eMalik N, Geraghty B, Dasgupta A, et al. MRI radiomics to differentiate between low grade glioma and glioblastoma peritumoral region. J Neurooncol 2021; 155(2): 181-191. \u003c/li\u003e\n\u003cli\u003eZhang Q, Peng Y, Liu W, et al. Radiomics Based on Multimodal MRI for the Differential Diagnosis of Benign and Malignant Breast Lesions. J Magn Reson Imaging\u003cem\u003e \u003c/em\u003e2020; 52(2): 596-607. \u003c/li\u003e\n\u003cli\u003eWang H, Zhang J, Bao S, et al. Preoperative MRI-Based Radiomic Machine-Learning Nomogram May Accurately Distinguish Between Benign and Malignant Soft-Tissue Lesions: A Two-Center Study. J Magn Reson Imaging 2020; 52(3): 873-882.\u003c/li\u003e\n\u003cli\u003eShi YJ, Zhu HT, Liu YL, et al. Radiomics Analysis Based on Diffusion Kurtosis Imaging and T2 Weighted Imaging for Differentiation of Pancreatic Neuroendocrine Tumors From Solid Pseudopapillary Tumors. Front Oncol 2020; 10:1624. \u003c/li\u003e\n\u003cli\u003eChen W, Wang S, Dong D, et al. Evaluation of Lymph Node Metastasis in Advanced Gastric Cancer Using Magnetic Resonance Imaging-Based Radiomics. Front Oncol 2019; 9: 1265. \u003c/li\u003e\n\u003cli\u003eTosin MC, Majolo M, Chedid R, et al. sEMG feature selection and classification using SVM-RFE. Annu Int Conf IEEE Eng Med Biol Soc 2017; 2017:390-393. \u003c/li\u003e\n\u003cli\u003eZhang S, Zhu F, Yu Q, et al. Identifying DNA-binding proteins based on multi-features and LASSO feature selection. Biopolymers\u003cem\u003e \u003c/em\u003e2021; 112(2): e23419.\u003c/li\u003e\n\u003cli\u003eSchober P and Vetter TR. Analysis of Variance in Medical Research. Anesth Analg\u003cem\u003e \u003c/em\u003e2020; 131(2): 508-509.\u003c/li\u003e\n\u003cli\u003eNiijima S and Okuno Y. Laplacian linear discriminant analysis approach to unsupervised feature selection. IEEE/ACM Trans Comput Biol Bioinform 2009; 6(4): 605-614. \u003c/li\u003e\n\u003cli\u003eAdeli E, Li X, Kwon D, et al. Logistic Regression Confined by Cardinality-Constrained Sample and Feature Selection. IEEE Trans Pattern Anal Mach Intell\u003cem\u003e \u003c/em\u003e2020; 42(7):1713-1728. \u003c/li\u003e\n\u003cli\u003eTosin MC, Majolo M, Chedid R, et al. sEMG feature selection and classification using SVM-RFE. Annu Int Conf IEEE Eng Med Biol Soc 2017; 2017: 390-393. \u003c/li\u003e\n\u003cli\u003eSung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021; 71(3): 209-249.\u003c/li\u003e\n\u003cli\u003eFeng F, Liu J, Wang F, et al. Prognostic value of differentiation status in gastric cancer. BMC Cancer\u003cem\u003e \u003c/em\u003e2018; 18(1):865.\u003c/li\u003e\n\u003cli\u003eMachlowska J, Maciejewski R, Sitarz R et al. The Pattern of Signatures in Gastric Cancer Prognosis. Int J Mol Sci 2018; 19(6):1658.\u003c/li\u003e\n\u003cli\u003eWu CW, Chi CW, Lin WC et al. Gastric cancer: prognostic and diagnostic advances. Expert Rev Mol Med 2002; 4(6):1-12.\u003c/li\u003e\n\u003cli\u003ePi\u0026ntilde;eiro-Fiel M, Moscoso A, Pubul V, et al. A Systematic Review of PET Textural Analysis and Radiomics in Cancer. Diagnostics (Basel) 2021; 11(2):380\u003c/li\u003e\n\u003cli\u003eGiganti F, Tang L, Baba H et al. Gastric cancer and imaging biomarkers: Part 1 - a critical review of DWI-MRI and CE-MDCT findings. Eur Radiol 2019; 29(4):1743-1753.\u003c/li\u003e\n\u003cli\u003eBorggreve AS, Goense L, Brenkman HJF, et al. Imaging strategies in the management of gastric cancer: current role and future potential of MRI. Br J Radiol 2019; 92(1097): 20181044. \u003c/li\u003e\n\u003cli\u003eJoo I, Lee JM, Kim JH, et al. Prospective comparison of 3T MRI with diffusion-weighted imaging and MDCT for the preoperative TNM staging of gastric cancer. \u003cem\u003eJ Magn Reson Imaging \u003c/em\u003e2015; 41(3):814-821.\u003c/li\u003e\n\u003cli\u003eYu Y, Tan Y, Xie C, et al. Development and Validation of a Preoperative Magnetic Resonance Imaging Radiomics-Based Signature to Predict Axillary Lymph Node Metastasis and Disease-Free Survival in Patients With Early-Stage Breast Cancer. JAMA network Open 2020; 3(12):e2028086.\u003c/li\u003e\n\u003cli\u003eWang H, Nie P, Wang Y, et al. Radiomics nomogram for differentiating between benign and malignant soft-tissue masses of the extremities. J Magn Reson Imaging 2020; 51(1):155-163.\u003c/li\u003e\n\u003cli\u003eWang H, Chen H, Duan S, et al. Radiomics and Machine Learning With Multiparametric Preoperative MRI May Accurately Predict the Histopathological Grades of Soft Tissue Sarcomas. J Magn Reson Imaging 2020; 51(3):791-797.\u003c/li\u003e\n\u003cli\u003eYip SSF and Aerts HJW. Applications and limitations of radiomics. Phys Med Biol 2016; 61(13): R150-R166.\u003c/li\u003e\n\u003cli\u003eLiu S, Zheng H, Zhang Y, et al. Whole-volume apparent diffusion coefficient-based entropy parameters for assessment of gastric cancer aggressiveness. J Magn Reson Imaging 2018; 47(1): 168-175.\u003c/li\u003e\n\u003cli\u003eConnolly M and Srinivasan A. Diffusion-Weighted Imaging in Head and Neck Cancer: Technique, Limitations and Applications. Magn Reson Imaging Clin N Am 2018; 26(1):121-133. \u003c/li\u003e\n\u003cli\u003eZhang Y, Chen J, Liu S, et al. Assessment of histological differentiation in gastric cancers using whole-volume histogram analysis of apparent diffusion coefficient maps. J Magn Reson Imaging (2016); 45:440\u0026ndash;449. \u003c/li\u003e\n\u003cli\u003eSah BR, Owczarczyk K, Siddique M, et al. Radiomics in esophageal and gastric cancer. Abdominal Radiol (2018); 44: 2048\u0026ndash;2058. \u003c/li\u003e\n\u003cli\u003eLiu S, Guan W, Wang H, et al. Apparent diffusion coefficient value of gastric cancer by diffusion-weighted imaging: correlations with the histological differentiation and Lauren classification. Eur J Radiol 2014; 83(12): 2122-2128.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003eTable 2 is available in the Supplementary Files section.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/p\u003e\n"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmim","sideBox":"Learn more about [BMC Medical Imaging](http://bmcmedimaging.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bmim/default.aspx","title":"BMC Medical Imaging","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"MRI, radiomics, ADCmin, differentiation degree, gastric cancer","lastPublishedDoi":"10.21203/rs.3.rs-7493820/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7493820/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003ePreoperative differentiation between poorly and highly differentiated gastric cancers is important for treatment decisions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePurpose:\u003c/strong\u003eTo investigate a radiomics model for preoperative differentiation between\u0026nbsp; poorly and highly differentiated gastric cancers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy type:\u003c/strong\u003eRetrospective.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePopulation: \u003c/strong\u003e239 patients with gastric cancer were included in the study. 167 patients were assigned to the training group and 72 patients comprised the testing group.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSequence: \u003c/strong\u003eT2-weighted (T2WI), T1-weighted (T1WI), diffusion-weighted imaging(DWI)on a 3.0T MR scanner.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment: \u003c/strong\u003e: A total of 2632 radiomics features were extracted from\u003cstrong\u003e \u003c/strong\u003eDWI and apparent diffusion coefficient (ADC) maps. Radiomics based on above features were built using four feature selection methods and three classifiers. All models were used to differentiate poorly and highly differentiated gastric cancers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical tests:\u003c/strong\u003e1) An analysis of independent t test was performed for clinicopathological information. 2) Four feature selection methods (Least Absolute Shrinkage and Selection Operator [LASSO], Analysis Of Variance [ANOVA], Recursive Feature Elimination [RFE] and Kruskal Wallis[KW]) and three classifiers (Support Vector Machine [SVM], Linear Discriminant Analysis [LDA] and Logistic Regression [LR]) were used to construct twelve radiomics. 3) The performance of different radiomics model was assessed using area under the receiver-operating characteristic curve (AUC) and accuracy (Acc) values.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003eThe model LASSO + SVM achieved the highest AUC of 0.854 with Acc of 0.793 in all radiomics model. The combination of ADC\u003csub\u003emin\u003c/sub\u003e and radiomics achieved the highest diagnostic efficiency with AUC of 0.919 and Acc of 0.823.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Conclusion:\u003c/strong\u003eThe combined model of radiomics and ADC\u003csub\u003emin\u003c/sub\u003e was accurate for distinguishing poorly and highly differentiated gastric cancers.\u003c/p\u003e","manuscriptTitle":"MRI-based Radiomics analysis for differentiation degree of gastric cancer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-24 05:21:02","doi":"10.21203/rs.3.rs-7493820/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-09-30T06:09:45+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-29T13:12:43+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-29T12:41:50+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"290999487255470844553929288690960068715","date":"2025-09-28T21:43:00+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-27T14:54:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"233106125082942890321850014951093996230","date":"2025-09-26T07:11:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"164295004233091735657480037206284441501","date":"2025-09-26T00:37:31+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"247070730019798889954881392454797888922","date":"2025-09-25T19:17:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"99144880466166711063344600182384316124","date":"2025-09-24T07:14:51+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"301527202498204111479264472763694759715","date":"2025-09-23T19:41:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"267295217173559285739908957151835461139","date":"2025-09-20T21:55:55+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-15T20:50:13+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-10T18:18:42+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-09-10T04:33:25+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-09T10:31:36+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Imaging","date":"2025-09-09T10:28:13+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-imaging","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bmim","sideBox":"Learn more about [BMC Medical Imaging](http://bmcmedimaging.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bmim/default.aspx","title":"BMC Medical Imaging","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"aef26409-8bde-4670-9fd1-b91cfa012cf0","owner":[],"postedDate":"September 24th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-15T16:04:20+00:00","versionOfRecord":{"articleIdentity":"rs-7493820","link":"https://doi.org/10.1186/s12880-025-02108-y","journal":{"identity":"bmc-medical-imaging","isVorOnly":false,"title":"BMC Medical Imaging"},"publishedOn":"2025-12-08 15:58:45","publishedOnDateReadable":"December 8th, 2025"},"versionCreatedAt":"2025-09-24 05:21:02","video":"","vorDoi":"10.1186/s12880-025-02108-y","vorDoiUrl":"https://doi.org/10.1186/s12880-025-02108-y","workflowStages":[]},"version":"v1","identity":"rs-7493820","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7493820","identity":"rs-7493820","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}