To investigate whether volumetric visceral adipose tissue (VAT) features extracted using radiomics and three-dimensional convolutional neural network (3D-CNN) approach are effective in differentiating Crohn’s disease (CD) and ulcerative colitis (UC).
This retrospective study enrolled 316 patients (mean age, 36.25 ± 13.58 [standard deviation]; 219 men) with confirmed diagnosis of CD and UC who underwent CT enterography between 2012 and 2021. Volumetric VAT was semi-automatically segmented on the arterial phase images. Radiomics analysis was performed using principal component analysis (PCA) and the least absolute shrinkage and selection operator (LASSO) logistic regression algorithm. We developed a 3D-CNN model using VAT imaging data from the training cohort. Clinical covariates including age, sex, modified body mass index, and disease duration that impact VAT were added to the machine learning model for adjustment. The model’s performance was evaluated on the testing cohort separating from the model’s development process by its discrimination and clinical utility.
Volumetric VAT radiomics analysis with LASSO had the highest AUC value of 0.717 (95% CI, 0.614–0.820), though difference of diagnostic performance among the 3D-CNN model (AUC = 0.693; 95% CI, 0.587–0.798) and radiomics analysis with PCA (AUC = 0.662; 95% CI, 0.548–0.776) and LASSO have not reached statistical significance (all p > 0.05). The radiomics score was higher in UC than in CD on the testing cohort (mean ± SD, UC 0.29 ± 1.05 versus CD −0.60 ± 1.25; p < 0.001). The LASSO model with adjustment of clinical covariates reached an AUC of 0.775 (95%CI, 0.683–0.868).
The developed volumetric VAT–based radiomics and 3D-CNN models provided comparable and effective performance for the characterization of CD from UC.
• High-output feature data extracted from volumetric visceral adipose tissue on CT enterography had an effective diagnostic performance for differentiating Crohn’s disease from ulcerative colitis.
• With adjustment of clinical covariates that cause difference in volumetric visceral adipose tissue, adjusted clinical machine learning model reached stronger performance when distinguishing Crohn’s disease patients from ulcerative colitis patients.
This is a preview of subscription content,to check access.
Access this article
Three-dimensional convolutional neural network
The area under the receiver operating characteristic curves
Inflammatory bowel disease
The least absolute shrinkage and selection operator
Principal component analysis
The receiver operating curve
Visceral adipose tissue
Uko V, Vortia E, Achkar JP et al (2014) Impact of abdominal visceral adipose tissue on disease outcome in pediatric Crohn’s disease. Inflamm Bowel Dis 20:2286–2291
Dobre M, Milanesi E, Mănuc TE et al (2018) Differential intestinal mucosa transcriptomic biomarkers for Crohn’s disease and ulcerative colitis. J Immunol Res. https://doi.org/10.1155/2018/9208274
Maconi G, Bolzoni E, Giussani A, Friedman AB, Duca P (2014) Accuracy and cost of diagnostic strategies for patients with suspected Crohn’s disease. J Crohns Colitis 8:1684–1692
Zulian A, Cancello R, Ruocco C et al (2013) Differences in visceral fat and fat bacterial colonization between ulcerative colitis and Crohn’s disease. An in vivo and in vitro study. PLoS One. https://doi.org/10.1371/journal.pone.0078495
Jahnsen J, Falch JA, Mowinckel P, Aadland E (2003) Body composition in patients with inflammatory bowel disease: a population-based study. Am J Gastroenterol 98:1556–1562
Zhang TH, Ding C, Xie TB et al (2017) Skeletal muscle depletion correlates with disease activity in ulcerative colitis and is reversed after colectomy. Clin Nutr 36:1586–1592
Cushing KC, Kordbacheh H, Gee MS, Kambadakone A, Ananthakrishnan AN (2018) Sarcopenia is a novel predictor of the need for rescue therapy in hospitalized ulcerative colitis patients. J Crohns Colitis 12:1036–1041
Erhayiem B, Dhingsa R, Hawkey CJ, Subramanian V (2011) Ratio of visceral to subcutaneous fat area is a biomarker of complicated Crohn’s disease. Clin Gastroenterol Hepatol 9:684–687
Bamba S, Inatomi O, Takahashi K et al (2021) Assessment of body composition from CT images at the level of the third lumbar vertebra in inflammatory bowel disease. Inflamm Bowel Dis 27:1435–1442
Büning C, von Kraft C, Hermsdorf M et al (2015) Visceral adipose tissue in patients with Crohn’s disease correlates with disease activity, inflammatory markers, and outcome. Inflamm Bowel Dis 21:2590–2597
Thiberge C, Charpentier C, Gillibert A et al (2018) Lower subcutaneous or visceral adiposity assessed by abdominal computed tomography could predict adverse outcome in patients with Crohn’s disease. J Crohns Colitis 12:1429–1437
Zulian A, Cancello R, Micheletto G et al (2012) Visceral adipocytes: old actors in obesity and new protagonists in Crohn’s disease? Gut 61:86–94
Rigiroli F, Hoye J, Lerebours R et al (2021) CT radiomic features of superior mesenteric artery involvement in pancreatic ductal adenocarcinoma: a pilot study. Radiology 301:610–622
Attanasio S, Forte SM, Restante G, Gabelloni M, Guglielmi G, Neri E (2020) Artificial intelligence, radiomics and other horizons in body composition assessment. Quant Imaging Med Surg 10:1650–1660
Kurowski JA, Barbur I, Gupta R et al (2018) Radiomic texture analysis shows differential expression within visceral adipose tissue regions on MRI reflecting severity of pediatric Crohn’s disease. Gastroenterology 6:S448–S448
Li X, Liang D, Meng J et al (2021) Development and validation of a novel computed-tomography enterography radiomic approach for characterization of intestinal fibrosis in Crohn’s disease. Gastroenterology 160:2303–2316
Shi J, Bao G, Hong J et al (2021) Deciphering CT texture features of human visceral fat to evaluate metabolic disorders and surgery-induced weight loss effects. EBioMedicine. https://doi.org/10.1016/j.ebiom.2021.103471
Truhn D, Schrading S, Haarburger C, Schneider H, Merhof D, Kuhl C (2019) Radiomic versus convolutional neural networks analysis for classification of contrast-enhancing lesions at multiparametric breast MRI. Radiology 290:290–297
Li H, Mo Y, Huang C et al (2021) An MSCT-based radiomics nomogram combined with clinical factors can identify Crohn’s disease and ulcerative colitis. Ann Transl Med 9:572
Bernstein CN, Eliakim A, Fedail S et al (2016) World Gastroenterology Organisation global guidelines inflammatory bowel disease: update August 2015. J Clin Gastroenterol 50:803–888
Satsangi J, Silverberg MS, Vermeire S, Colombel JF (2006) The Montreal classification of inflammatory bowel disease: controversies, consensus, and implications. Gut 55:749–753
Zhou Z, Xiong Z, Xie Q et al (2021) Computed tomography-based multiple body composition parameters predict outcomes in Crohn’s disease. Insights Imaging. https://doi.org/10.1186/s13244-021-01083-6
Zhou Z, Lu J, Morelli JN et al (2021) Utility of noncontrast MRI in the detection and risk grading of gastrointestinal stromal tumor: a comparison with contrast-enhanced CT. Quant Imaging Med Surg 11:2453–2464
Taylor SA, Avni F, Cronin CG et al (2017) The first joint ESGAR/ ESPR consensus statement on the technical performance of cross-sectional small bowel and colonic imaging. Eur Radiol 27:2570–2582
Lin L, Dekkers IA, Huang L et al (2021) Renal sinus fat volume in type 2 diabetes mellitus is associated with glycated hemoglobin and metabolic risk factors. J Diabetes Complicat. https://doi.org/10.1016/j.jdiacomp.2021.107973
Zwanenburg A, Vallieres M, Abdalah MA et al (2020) The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping. Radiology 295:328–338
van Griethuysen JJM, Fedorov A, Parmar C et al (2017) Computational radiomics system to decode the radiographic phenotype. Cancer Res 77:e104–e107
Bernatz S, Ackermann J, Mandel P et al (2020) Comparison of machine learning algorithms to predict clinically significant prostate cancer of the peripheral zone with multiparametric MRI using clinical assessment categories and radiomic features. Eur Radiol 30:6757–6769
Wang W, Cao K, Jin S, Zhu X, Ding J, Peng W (2020) Differentiation of renal cell carcinoma subtypes through MRI-based radiomics analysis. Eur Radiol 10:5738–5747
Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez J-C et al (2011) pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. https://doi.org/10.1186/1471-2105-12-77
Magudia K, Bridge CP, Bay CP et al (2021) Population-scale CT-based body composition analysis of a large outpatient population using deep learning to derive age-, sex-, and race-specific reference curves. Radiology 298:319–329
Engvall IL, Tengstrand B, Brismar K, Hafström I (2010) Infliximab therapy increases body fat mass in early rheumatoid arthritis independently of changes in disease activity and levels of leptin and adiponectin: a randomised study over 21 months. Arthritis Res Ther. https://doi.org/10.1186/ar3169
Goeman JJ, Solari A (2014) Multiple hypothesis testing in genomics. Stat Med 33:1946–1978
Van Der Sloot KW, Joshi AD, Bellavance DR et al (2017) Visceral adiposity, genetic susceptibility, and risk of complications among individuals with Crohn’s disease. Inflamm Bowel Dis 23:82–88
Li XH, Feng ST, Cao QH et al (2021) Degree of creeping fat assessed by computed tomography enterography is associated with intestinal fibrotic stricture in patients with Crohn’s disease: a potentially novel mesenteric creeping fat index. J Crohns Colitis 15:1161–1173
Grillot J, D’Engremont C, Parmentier AL et al (2020) Sarcopenia and visceral obesity assessed by computed tomography are associated with adverse outcomes in patients with Crohn’s disease. Clin Nutr 39:3024–3030
We thank Xinyueyuan Hao and Fangqin Tan for the segmentation of the visceral adipose tissue. The authors thank Yuhan Tang from the School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, China, for advice on statistical analysis.
This work is supported by grants from the National Natural Science Foundation of China (NSFC), Nos. 82071890 and 82071889.
The scientific guarantor of this publication is Yaqi Shen.
Conflict of interest
The authors of this manuscript declare that they have no conflict of interest.
Statistics and biometry
Two of the authors have significant statistical expertise.
Written informed consent was waived by the Institutional Review Board.
Institutional Review Board approval was obtained.
• observational/diagnostic study
• performed at one institution
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Zhou, Z., Xiong, Z., Cheng, R. et al. Volumetric visceral fat machine learning phenotype on CT for differential diagnosis of inflammatory bowel disease. Eur Radiol 33, 1862–1872 (2023). https://doi.org/10.1007/s00330-022-09171-x