A Novel Clinical Radiomics Nomogram to Identify Crohn's Disease from Intestinal Tuberculosis

A novel multidisciplinary machine learning approach based on clinical, imaging, colonoscopy, and pathology features for distinguishing intestinal tuberculosis from Crohn's disease

OBJECTIVES

Differentiating intestinal tuberculosis (ITB) from Crohn's disease (CD) remains a diagnostic dilemma. Misdiagnosis carries potential grave implications. We aim to establish a multidisciplinary-based model using machine learning approach for distinguishing ITB from CD.

METHODS

Eighty-two patients including 25 patients with ITB and 57 patients with CD were retrospectively recruited (54 in training cohort and 28 in testing cohort). The region of interest (ROI) for the lesion was delineated on magnetic resonance enterography (MRE) and colonoscopy images. Radiomic features were extracted by least absolute shrinkage and selection operator regression. Pathological feature was extracted automatically by deep-learning method. Clinical features were filtered by logistic regression analysis. Diagnostic performance was evaluated by receiver operating characteristic (ROC) curve and decision curve analysis (DCA). Delong's test was applied to compare the efficiency between the multidisciplinary-based model and the other four single-disciplinary-based models.

RESULTS

The radiomics model based on MRE features yielded an AUC of 0.87 (95% confidence interval [CI] 0.68-0.96) on the test data set, which was similar to the clinical model (AUC, 0.90 [95% CI 0.71-0.98]) and higher than the colonoscopy radiomics model (AUC, 0.68 [95% CI 0.48-0.84]) and pathology deep-learning model (AUC, 0.70 [95% CI 0.49-0.85]). Multidisciplinary model, integrating 3 clinical, 21 MRE radiomic, 5 colonoscopy radiomic, and 4 pathology deep-learning features, could significantly improve the diagnostic performance (AUC of 0.94, 95% CI 0.78-1.00) on the bases of single-disciplinary-based models. DCA confirmed the clinical utility.

CONCLUSIONS

Multidisciplinary-based model integrating clinical, MRE, colonoscopy, and pathology features was useful in distinguishing ITB from CD.

AI-luminating Artificial Intelligence in Inflammatory Bowel Diseases: A Narrative Review on the Role of AI in Endoscopy, Histology, and Imaging for IBD

Endoscopy, histology, and cross-sectional imaging serve as fundamental pillars in the detection, monitoring, and prognostication of inflammatory bowel disease (IBD). However, interpretation of these studies often relies on subjective human judgment, which can lead to delays, intra- and interobserver variability, and potential diagnostic discrepancies. With the rising incidence of IBD globally coupled with the exponential digitization of these data, there is a growing demand for innovative approaches to streamline diagnosis and elevate clinical decision-making. In this context, artificial intelligence (AI) technologies emerge as a timely solution to address the evolving challenges in IBD. Early studies using deep learning and radiomics approaches for endoscopy, histology, and imaging in IBD have demonstrated promising results for using AI to detect, diagnose, characterize, phenotype, and prognosticate IBD. Nonetheless, the available literature has inherent limitations and knowledge gaps that need to be addressed before AI can transition into a mainstream clinical tool for IBD. To better understand the potential value of integrating AI in IBD, we review the available literature to summarize our current understanding and identify gaps in knowledge to inform future investigations.

Artificial intelligence for discrimination of Crohn's disease and gastrointestinal tuberculosis: A systematic review

BACKGROUND AND AIM

Discrimination of gastrointestinal tuberculosis (GITB) and Crohn's disease (CD) is difficult. Use of artificial intelligence (AI)-based technologies may help in discriminating these two entities.

METHODS

We conducted a systematic review on the use of AI for discrimination of GITB and CD. Electronic databases (PubMed and Embase) were searched on June 6, 2022, to identify relevant studies. We included any study reporting the use of clinical, endoscopic, and radiological information (textual or images) to discriminate GITB and CD using any AI technique. Quality of studies was assessed with MI-CLAIM checklist.

RESULTS

Out of 27 identified results, a total of 9 studies were included. All studies used retrospective databases. There were five studies of only endoscopy-based AI, one of radiology-based AI, and three of multiparameter-based AI. The AI models performed fairly well with high accuracy ranging from 69.6-100%. Text-based convolutional neural network was used in three studies and Classification and regression tree analysis used in two studies. Interestingly, irrespective of the AI method used, the performance of discriminating GITB and CD did not match in discriminating from other diseases (in studies where a third disease was also considered).

CONCLUSION

The use of AI in differentiating GITB and CD seem to have acceptable accuracy but there were no direct comparisons with traditional multiparameter models. The use of multiple parameter-based AI models have the potential for further exploration in search of an ideal tool and improve on the accuracy of traditional models.

Machine Learning Diagnosis of Small-Bowel Crohn Disease Using T2-Weighted MRI Radiomic and Clinical Data

BACKGROUND. Radiologists have variable diagnostic performance and considerable interreader variability when interpreting MR enterography (MRE) examinations for suspected Crohn disease (CD). OBJECTIVE. The purposes of this study were to develop a machine learning method for predicting ileal CD by use of radiomic features of ileal wall and mesenteric fat from noncontrast T2-weighted MRI and to compare the performance of the method with that of expert radiologists. METHODS. This single-institution study included retrospectively identified patients who underwent MRE for suspected ileal CD from January 1, 2020, to January 31, 2021, and prospectively enrolled participants (patients with newly diagnosed ileal CD or healthy control participants) from December 2018 to October 2021. Using axial T2-weighted SSFSE images, a radiologist selected two slices showing greatest terminal ileal wall thickening. Four ROIs were segmented, and radiomic features were extracted from each ROI. After feature selection, support-vector machine models were trained to classify the presence of ileal CD. Three fellowship-trained pediatric abdominal radiologists independently classified the presence of ileal CD on SSFSE images. The reference standard was clinical diagnosis of ileal CD based on endoscopy and biopsy results. Radiomic-only, clinical-only, and radiomic-clinical ensemble models were trained and evaluated by nested cross-validation. RESULTS. The study included 135 participants (67 female, 68 male; mean age, 15.2 ± 3.2 years); 70 were diagnosed with ileal CD. The three radiologists had accuracies of 83.7% (113/135), 88.1% (119/135), and 86.7% (117/135) for diagnosing CD; consensus accuracy was 88.1%. Interradiologist agreement was substantial (κ = 0.78). The best-performing ROI was bowel core (AUC, 0.95; accuracy, 89.6%); other ROIs had worse performance (whole-bowel AUC, 0.86; fat-core AUC, 0.70; whole-fat AUC, 0.73). For the clinical-only model, AUC was 0.85 and accuracy was 80.0%. The ensemble model combining bowel-core radiomic and clinical models had AUC of 0.98 and accuracy of 93.5%. The bowel-core radiomic-only model had significantly greater accuracy than radiologist 1 (p = .009) and radiologist 2 (p = .02) but not radiologist 3 (p > .99) or the radiologists in consensus (p = .05). The ensemble model had greater accuracy than the radiologists in consensus (p = .02). CONCLUSION. A radiomic machine learning model predicted CD diagnosis with better performance than two of three expert radiologists. Model performance improved when radiomic data were ensembled with clinical data. CLINICAL IMPACT. Deployment of a radiomic-based model including T2-weighted MRI data could decrease interradiologist variability and increase diagnostic accuracy for pediatric CD.

Automatic Segmentation and Radiomics for Identification and Activity Assessment of CTE Lesions in Crohn's Disease

BACKGROUND

The purpose of this article is to develop a deep learning automatic segmentation model for the segmentation of Crohn's disease (CD) lesions in computed tomography enterography (CTE) images. Additionally, the radiomics features extracted from the segmented CD lesions will be analyzed and multiple machine learning classifiers will be built to distinguish CD activity.

METHODS

This was a retrospective study with 2 sets of CTE image data. Segmentation datasets were used to establish nnU-Net neural network's automatic segmentation model. The classification dataset was processed using the automatic segmentation model to obtain segmentation results and extract radiomics features. The most optimal features were then selected to build 5 machine learning classifiers to distinguish CD activity. The performance of the automatic segmentation model was evaluated using the Dice similarity coefficient, while the performance of the machine learning classifier was evaluated using the area under the curve, sensitivity, specificity, and accuracy.

RESULTS

The segmentation dataset had 84 CTE examinations of CD patients (mean age 31 ± 13 years , 60 males), and the classification dataset had 193 (mean age 31 ± 12 years , 136 males). The deep learning segmentation model achieved a Dice similarity coefficient of 0.824 on the testing set. The logistic regression model showed the best performance among the 5 classifiers in the testing set, with an area under the curve, sensitivity, specificity, and accuracy of 0.862, 0.697, 0.840, and 0.759, respectively.

CONCLUSION

The automated segmentation model accurately segments CD lesions, and machine learning classifier distinguishes CD activity well. This method can assist radiologists in promptly and precisely evaluating CD activity.

CTE-Based Radiomics Models Can Identify Mucosal Healing in Patients with Crohn's Disease

RATIONALE AND OBJECTIVES

To develop computed tomography enterography (CTE)-based radiomics models to assess mucosal healing (MH) in patients with Crohn's disease (CD).

MATERIALS AND METHODS

CTE images were retrospectively collected from 92 confirmed cases of CD at the post-treatment review. Patients were randomly divided into developing (n = 73) and testing (n = 19) groups. Radiomics features were extracted from the enteric phase images, and the least absolute shrinkage and selection operator (LASSO) logistic regression was applied for feature selection using 5-fold cross-validation on the developing group. The selected features were further identified from the top-ranked features and used to create improved radiomics models. Machine learning models were constructed to compare radiomics models with different radiomics features. The area under the ROC curve (AUC) was calculated to assess the predictive performance for identifying MH in CD.

RESULTS

Among the 92 CD patients included in our study, 36 patients achieved MH. The AUC of the radiomics model 1, which was based on the 26 selected radiomics features, was 0.976 for evaluating MH in the testing cohort. The AUCs of radiomics models 2 and 4, based on the top 10 and top 5 positive and negative radiomics features, were 0.974 and 0.952 in the testing cohort, respectively. The AUC of the radiomics model 3, built by removing features with r > 0.5, was 0.956 in the testing cohort. The clinical utility of the clinical radiomics nomogram was confirmed by the decision curve analysis (DCA).

CONCLUSION

The CTE-based radiomics models have demonstrated favorable performance in assessing MH in patients with CD. Radiomics features can be used as a promising imaging biomarker for MH.

A Novel Radiomics Model Integrating Luminal and Mesenteric Features to Predict Mucosal Activity and Surgery Risk in Crohn's Disease Patients: A Multicenter Study

BACKGROUND

To investigate the feasibility of integrating radiomics and morphological features based on computed tomography enterography (CTE) for developing a noninvasive grading model for mucosal activity and surgery risk of Crohn's disease (CD) patients.

METHODS

A total of 167 patients from three centers were enrolled. Radiomics and image morphological features were extracted to quantify segmental and global simple endoscopic score for Crohn's disease (SES-CD). An image-fusion-based support vector machine (SVM) classifier was used for grading SES-CD and identifying moderate-to-severe SES-CD. The performance of the predictive model was assessed using the area under the receiver operating characteristic curve (AUC). A multiparametric model was developed to predict surgical progression in CD patients by combining sum-image scores and clinical data.

RESULTS

The AUC values of the multicategorical segmental SES-CD fusion radiomic model based on a combination of luminal and mesenteric radiomics were 0.828 and 0.709 in training and validation cohorts. The image fusion model integrating the fusion radiomics and morphological features could accurately distinguish bowel segments with moderate-to-severe SES-CD in both the training cohort (AUC = 0.847, 95% confidence interval (CI): 0.784-0.902) and the validation cohort (AUC = 0.896, 95% CI: 0.812-0.960). A predictive nomogram for interval surgery was developed based on multivariable cox analysis.

CONCLUSIONS

This study demonstrated the feasibility of integrating lumen and mesentery radiomic features to develop a promising noninvasive grading model for mucosal activity of CD. In combination with clinical data, the fusion-image score may yield an accurate prognostic model for time to surgery.

Differentiating gastrointestinal tuberculosis and Crohn's disease- a comprehensive review

Gastrointestinal Tuberculosis (GITB) and Crohn's disease (CD) are both chronic granulomatous diseases with a predilection to involve primarily the terminal ileum. GITB is often considered a disease of the developing world, while CD and inflammatory bowel disease are considered a disease of the developed world. But in recent times, the epidemiology of both diseases has changed. Differentiating GITB from CD is of immense clinical importance as the management of both diseases differs. While GITB needs anti-tubercular therapy (ATT), CD needs immunosuppressive therapy. Misdiagnosis or a delay in diagnosis can lead to catastrophic consequences. Most of the clinical features, endoscopic findings, and imaging features are not pathognomonic for either of these two conditions. The definitive diagnosis of GITB can be clinched only in a fraction of cases with microbiological positivity (acid-fast bacilli, mycobacterial culture, or PCR-based tests). In most cases, the diagnosis is often based on consistent clinical, endoscopic, imaging, and histological findings. Similarly, no single finding can conclusively diagnose CD. Multiparametric-based predictive models incorporating clinical, endoscopy findings, histology, radiology, and serology have been used to differentiate GITB from CD with varied results. However, it is limited by the lack of validation studies for most such models. Many patients, especially in TB endemic regions, are initiated on a trial of ATT to see for an objective response to therapy. Early mucosal response assessed at two months is an objective marker of response to ATT. Prolonged ATT in CD is recognized to have a fibrotic effect. Therefore, early discrimination may be vital in preventing the delay in the diagnosis of CD and avoiding a complicated course.

The Role of Radiomics in Fibrosis Crohn's Disease: A Review

Inflammatory bowel disease (IBD) is a global health concern that has been on the rise in recent years. In addition, imaging is the established method of care for detecting, diagnosing, planning treatment, and monitoring the progression of IBD. While conventional imaging techniques are limited in their ability to provide comprehensive information, cross-sectional imaging plays a crucial role in the clinical management of IBD. However, accurately characterizing, detecting, and monitoring fibrosis in Crohn's disease remains a challenging task for clinicians. Recent advances in artificial intelligence technology, machine learning, computational power, and radiomic emergence have enabled the automated evaluation of medical images to generate prognostic biomarkers and quantitative diagnostics. Radiomics analysis can be achieved via deep learning algorithms or by extracting handcrafted radiomics features. As radiomic features capture pathophysiological and biological data, these quantitative radiomic features have been shown to offer accurate and rapid non-invasive tools for IBD diagnostics, treatment response monitoring, and prognosis. For these reasons, the present review aims to provide a comprehensive review of the emerging radiomics methods in intestinal fibrosis research that are highlighted and discussed in terms of challenges and advantages.

Computed tomography enterography-based multiregional radiomics model for differential diagnosis of Crohn's disease from intestinal tuberculosis

PURPOSE

To build computed tomography enterography (CTE)-based multiregional radiomics model for distinguishing Crohn's disease (CD) from intestinal tuberculosis (ITB).

MATERIALS AND METHODS

A total of 105 patients with CD and ITB who underwent CTE were retrospectively enrolled. Volume of interest segmentation were performed on CTE and radiomic features were obtained separately from the intestinal wall of lesion, the largest lymph node (LN), and region surrounding the lesion in the ileocecal region. The most valuable radiomic features was selected by the selection operator and least absolute shrinkage. We established nomogram combining clinical factors, endoscopy results, CTE features, and radiomic score through multivariate logistic regression analysis. Receiver operating characteristic (ROC) curves and decision curve analysis (DCA) were used to evaluate the prediction performance. DeLong test was applied to compare the performance of the models.

RESULTS

The clinical-radiomic combined model comprised of four variables including one radiomic signature from intestinal wall, one radiomic signature from LN, involved bowel segments on CTE, and longitudinal ulcer on endoscopy. The combined model showed good diagnostic performance with an area under the ROC curve (AUC) of 0.975 (95% CI 0.953-0.998) in the training cohort and 0.958 (95% CI 0.925-0.991) in the validation cohort. The combined model showed higher AUC than that of the clinical model in cross-validation set (0.958 vs. 0.878, P = 0.004). The DCA showed the highest benefit for the combined model.

CONCLUSION

Clinical-radiomic combined model constructed by combining CTE-based radiomics from the intestinal wall of lesion and LN, endoscopy results, and CTE features can accurately distinguish CD from ITB.

Development of a simplified model and nomogram in preoperative diagnosis of pediatric chronic cholangitis with pancreaticobiliary maljunction using clinical variables and MRI radiomics

OBJECTIVE

The aim of this study was to develop a model that combines clinically relevant features with radiomics signature based on magnetic-resonance imaging (MRI) for diagnosis of chronic cholangitis in pancreaticobiliary maljunction (PBM) children.

METHODS

A total of 144 subjects from two institutions confirmed PBM were included in this study. Clinical characteristics and MRI features were evaluated to build a clinical model. Radiomics features were extracted from the region of interest manually delineated on T2-weighted imaging. A radiomics signature was developed by the selected radiomics features using the least absolute shrinkage and selection operator and then a radiomics score (Rad-score) was calculated. We constructed a combined model incorporating clinical factors and Rad-score by multivariate logistic regression analysis. The combined model was visualized as a radiomics nomogram to achieve model visualization and provide clinical utility. Receiver operating curve analysis and decision curve analysis (DCA) were used to evaluate the diagnostic performance.

RESULTS

Jaundice, protein plug, and ascites were selected as key clinical variables. Eight radiomics features were combined to construct the radiomics signature. The combined model showed superior predictive performance compared with the clinical model alone (AUC in the training cohort: 0.891 vs. 0.767, the validation cohort: 0.858 vs. 0.731), and the difference was significant (p = 0.002, 0.028) in the both cohorts. DCA confirmed the clinical utility of the radiomics nomogram.

CONCLUSION

The proposed model that combines key clinical variables and radiomics signature is helpful in the diagnosis of chronic cholangitis in PBM children.

Development and validation of a clinical radiomics nomogram to predict secondary loss of response to infliximab in Crohn's disease patients

Background

Infliximab (IFX) is the first-line treatment for Crohn's disease (CD). However, the secondary loss of response (LOR) is common in IFX therapy. Therefore, non-invasive assessment of LOR in CD patients is the goal pursued by clinicians.

Methods

A multicenter study involving 181 CD patients was conducted, with patients being split into a training cohort (n = 102), testing cohort (n = 45), and validation cohort (n = 34). The study evaluated various clinical factors to establish a clinical model, and a radiomics signature was constructed based on reproducible features from computed tomography enterography (CTE). Logistic regression modeling was used to create models based on the radiomics signature and significant clinical factors, with the receiver operating characteristic curve (ROC) used to compare their performance.

Results

The study found that 64 of the 181 CD patients included experienced secondary LOR. The radiomics signature performed well in predicting secondary LOR, showing good discrimination in the training cohort (AUC [area under the curve], 0.947; 95% confidence interval [CI], 0.910-0.976), the testing cohort (AUC, 0.860; 95% CI, 0.768-0.941), and the validation cohort (AUC, 0.921; 95% CI: 0.831-1.000). Decision curve analysis (DCA) demonstrated the clinical value of the radiomics nomogram.

Conclusions

The CTE-based radiomics model showed good performance in predicting secondary LOR in CD patients. The nomogram can help clinicians choose alternative biologics early for CD patients.

A novel clinical radiomics nomogram at baseline to predict mucosal healing in Crohn's disease patients treated with infliximab

OBJECTIVES

Mucosal healing (MH) is currently the gold standard in Crohn's disease (CD) management. Noninvasive assessment of MH in CD patients is increasingly a concern of clinicians.

METHODS

This retrospective study included 106 patients with confirmed CD who were divided into a training cohort (n = 73) and a testing cohort (n = 33). Patient demographics were evaluated to establish a clinical model. Radiomics features were extracted from computed tomography enterography (CTE) images. A radiomics signature was constructed, and a radiomics score (Rad-score) was calculated by using the radiomics signature-based formula. A clinical radiomics nomogram was then built by incorporating the Rad-score and significant clinical features. The diagnostic performance of the three models was evaluated using receiver operating characteristic (ROC) curve analysis.

RESULTS

Of the 106 patients with CD, 37 exhibited MH after 26 weeks of infliximab (IFX) treatment. The area under the ROC curve (AUC) of the clinical radiomics nomogram for distinguishing MH from non-MH, which was based on the disease duration and Rad-score, was 0.880 (95% confidence interval [CI]: 0.809-0.943) in the training cohort and 0.877 (95% CI: 0.745-0.983) in the testing cohort. Decision curve analysis (DCA) confirmed the clinical utility of the clinical radiomics nomogram.

CONCLUSIONS

This is a preliminary study suggesting that this CTE-based radiomics model has potential value for predicting MH in CD patients. A nomogram constructed by combining radiomics signatures and clinical features can help clinicians select appropriate therapeutic strategies for CD patients.

KEY POINTS

• The disease duration (odds ratio (OR) = 0.969, 95% confidence interval (CI) = 0.943-0.995, p = 0.021) was an independent predictor of MH in the clinical model. • The AUC of the radiomics model constructed by the five radiomics features was 0.846 (95% CI: 0.759-0.921) in the training cohort and 0.817 (95% CI: 0.665-0.945) in the testing cohort for differentiating MH from non-MH. • The AUC of the clinical radiomics nomogram was 0.880 (95% CI: 0.809-0.943) in the training cohort and 0.877 (95% CI: 0.745-0.983) in the testing cohort for distinguishing MH from non-MH.