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.

Deep Learning Radiomics Analysis of CT Imaging for Differentiating Between Crohn's Disease and Intestinal Tuberculosis.

This study aimed to develop and evaluate a CT-based deep learning radiomics model for differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB). A total of 330 patients with pathologically confirmed as CD or ITB from the First Affiliated Hospital of Zhengzhou University were divided into the validation dataset one (CD: 167; ITB: 57) and validation dataset two (CD: 78; ITB: 28). Based on the validation dataset one, the synthetic minority oversampling technique (SMOTE) was adopted to create balanced dataset as training data for feature selection and model construction. The handcrafted and deep learning (DL) radiomics features were extracted from the arterial and venous phases images, respectively. The interobserver consistency analysis, Spearman's correlation, univariate analysis, and the least absolute shrinkage and selection operator (LASSO) regression were used to select features. Based on extracted multi-phase radiomics features, six logistic regression models were finally constructed. The diagnostic performances of different models were compared using ROC analysis and Delong test. The arterial-venous combined deep learning radiomics model for differentiating between CD and ITB showed a high prediction quality with AUCs of 0.885, 0.877, and 0.800 in SMOTE dataset, validation dataset one, and validation dataset two, respectively. Moreover, the deep learning radiomics model outperformed the handcrafted radiomics model in same phase images. In validation dataset one, the Delong test results indicated that there was a significant difference in the AUC of the arterial models (p = 0.037), while not in venous and arterial-venous combined models (p = 0.398 and p = 0.265) as comparing deep learning radiomics models and handcrafted radiomics models. In our study, the arterial-venous combined model based on deep learning radiomics analysis exhibited good performance in differentiating between CD and ITB.

The Diagnosis of Esophageal Tuberculosis through an Endoscopic Ultrasound-guided Fine-needle Aspiration Biopsy.

A 54-year-old woman presented with an elevated esophageal lesion. Computed tomography (CT) and magnetic resonance imaging revealed a mass in the pancreatic head. Endoscopic ultrasound (EUS) showed a well-defined, round, hypoechoic mass, which was considered lymph node enlargement. An EUS-guided fine-needle aspiration biopsy (FNAB) was performed on the esophagus and the mass above the pancreatic head. The pathologically confirmed epithelial cells and multinucleated giant cells were positive for T-SPOT. Clinically, tuberculous lymphadenitis and esophageal tuberculosis were suspected, with successful treatment with anti-tuberculosis therapy resulting in a good response. Our findings suggest that an EUS-FNAB is useful for diagnosing esophageal tuberculosis.

Differentiation of Crohn's disease, ulcerative colitis, and intestinal tuberculosis by dual-layer spectral detector CT enterography.

AIM: To investigate the value of radiological features and energy spectrum quantitative parameters in the differential diagnosis of Crohn's disease (CD), ulcerative colitis (UC), and intestinal tuberculosis (ITB) by dual-layer spectral detector computed tomography (CT) enterography (CTE). MATERIALS AND METHODS: Clinical and CTE data were collected from 182 patients with CD, 29 with UC, and 51 with ITB. CT images were obtained at the enteric phases and portal phases. The quantitative energy spectrum parameters were iodine density (ID), normalised ID (NID), virtual non-contrast (VNC) value, and effective atomic number (Z-eff). The area under curve (AUC) of the receiver operating characteristic curve (ROC) was calculated. RESULTS: The vascular comb sign (p=0.009) and enlarged lymph nodes (p=0.001) were more common in patients with CD than UC or ITB. In the differentiation of moderate-severe active CD from UC, enteric phase NID (AUC, 0.938; p<0.001) and portal phase Z-eff (AUC, 0.925; p<0.001) had the highest accuracy, which were compared separately. In the differentiation of moderate-severe active CD from ITB, enteric phase NID (AUC, 0.906; p<0.001) and portal phase Z-eff (AUC, 0.947; p<0.001) had the highest accuracy; however, the AUC value was highest when the four parameters are combined (AUC, 0.989; p<0.001; AUC, 0.986; p<0.001; AUC, 0.936; p<0.001; and AUC, 0.986; p<0.001). CONCLUSION: The present study shows that the combined strategies of four parameters have higher sensitivity and specificity in differentiating CD, UC, and ITB, and may play a key role in guiding treatment.

Abdominal visceral tuberculosis: a malignancy mimic.

The purpose is to discuss abdominal tuberculosis mimicking malignancy involving the abdominal viscera. TB of the abdominal viscera is common, especially in countries where tuberculosis is endemic and in pockets of non-endemic countries. Diagnosis is challenging as clinical presentations are often non-specific. Tissue sampling may be necessary for definitive diagnosis. Awareness of the early and late disease imaging appearances of abdominal tuberculosis involving the viscera that can mimic malignancy can aid detecting TB, providing a differential diagnosis, assessing extent of spread, guiding biopsy, and evaluating response.

Duodenal tuberculosis.

Gastrointestinal tuberculosis (TB) is a rare disease and only involves the duodenum in 2-2,5% of all cases. A 60-year-old female with no reported medical history, presented with constitutional syndrome with a 10 kg weight loss in three months, epigastric pain, bloating and vomiting. She denied fever or respiratory symptoms. Laboratory examination revealed elevated C-reactive protein levels and low prealbumin. Abdominal computed tomography (CT) showed duodenal wall thickening, mainly in its third part, with infiltration of the root of the mesentery and numerous subcentimeter adenopathies at that level.

Intestinal and perianal tuberculosis: an uncommon clinical presentation and challenging diagnosis.

We present the case of a 40-year-old immunocompetent man with complaints of perianal abscess, diarrhea, and weight loss for 3 months. He denied respiratory symptoms. Colonoscopy revealed ulcers and histopathology showed non-necrotizing granulomas and excluded acid-fast bacilli. Initially, the clinical and histological findings favored the diagnosis of Crohn's disease, however, further investigation by performing chest radiography in the context of a pre-immunomodulatory checklist, revealed pulmonary cavitations confirmed on computed tomography scan. Further mycobacteriological studies suggested the diagnosis of active pulmonary tuberculosis. Demonstration of the presence of Mycobacterium tuberculosis in the colon was possible by RT-PCR. Complete resolution of intestinal and perianal symptoms was achieved 2 weeks after treatment with anti-bacillary agents.

Applying logistic LASSO regression for the diagnosis of atypical Crohn's disease.

In countries with a high incidence of tuberculosis, the typical clinical features of Crohn's disease (CD) may be covered up after tuberculosis infection, and the identification of atypical Crohn's disease and intestinal tuberculosis (ITB) is still a dilemma for clinicians. Least absolute shrinkage and selection operator (LASSO) regression has been applied to select variables in disease diagnosis. However, its value in discriminating ITB and atypical Crohn's disease remains unknown. A total of 400 patients were enrolled from January 2014 to January 2019 in second Xiangya hospital Central South University.Among them, 57 indicators including clinical manifestations, laboratory results, endoscopic findings, computed tomography enterography features were collected for further analysis. R software version 3.6.1 (glmnet package) was used to perform the LASSO logistic regression analysis. SPSS 20.0 was used to perform Pearson chi-square test and binary logistic regression analysis. In the variable selection step, LASSO regression and Pearson chi-square test were applied to select the most valuable variables as candidates for further logistic regression analysis. Secondly, variables identified from step 1 were applied to construct binary logistic regression analysis. Receiver operating characteristic (ROC) curve analysis was performed on these models to assess the ability and the optimal cutoff value for diagnosis. The area under the ROC curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy rate, together with their 95% confidence and intervals (CIs) were calculated. MedCalc software (Version 16.8) was applied to analyze the ROC curves of models. 332 patients were eventually enrolled to build a binary logistic regression model to discriminate CD (including comprehensive CD and tuberculosis infected CD) and ITB. However, we did not get a satisfactory diagnostic value via applying the binary logistic regression model of comprehensive CD and ITB to predict tuberculosis infected CD and ITB (accuracy rate:79.2%VS 65.1%). Therefore, we further established a binary logistic regression model to discriminate atypical CD from ITB, based on Pearsonchi-square test (model1) and LASSO regression (model 2). Model 1 showed 89.9% specificity, 65.9% sensitivity, 88.5% PPV, 68.9% NPV, 76.9% diagnostic accuracy, and an AUC value of 0.811, and model 2 showed 80.6% specificity, 84.4% sensitivity, 82.3% PPV, 82.9% NPV, 82.6% diagnostic accuracy, and an AUC value of 0.887. The comparison of AUCs between model1 and model2 was statistically different (P < 0.05). Tuberculosis infection increases the difficulty of discriminating CD from ITB. LASSO regression showed a more efficient ability than Pearson chi-square test based logistic regression on differential diagnosing atypical CD and ITB.

Deep-learning system for real-time differentiation between Crohn's disease, intestinal Behçet's disease, and intestinal tuberculosis.

BACKGROUND AND AIM: Pattern analysis of big data can provide a superior direction for the clinical differentiation of diseases with similar endoscopic findings. This study aimed to develop a deep-learning algorithm that performs differential diagnosis between intestinal Behçet's disease (BD), Crohn's disease (CD), and intestinal tuberculosis (ITB) using colonoscopy images. METHODS: The typical pattern for each disease was defined as a typical image. We implemented a convolutional neural network (CNN) using Pytorch and visualized a deep-learning model through Gradient-weighted Class Activation Mapping. The performance of the algorithm was evaluated using the area under the receiver operating characteristic curve (AUROC). RESULTS: A total of 6617 colonoscopy images of 211 CD, 299 intestinal BD, and 217 ITB patients were used. The accuracy of the algorithm for discriminating the three diseases (all-images: 65.15% vs typical images: 72.01%, P = 0.024) and discriminating between intestinal BD and CD (all-images: 78.15% vs typical images: 85.62%, P = 0.010) was significantly different between all-images and typical images. The CNN clearly differentiated colonoscopy images of the diseases (AUROC from 0.7846 to 0.8586). Algorithmic prediction AUROC for typical images ranged from 0.8211 to 0.9360. CONCLUSION: This study found that a deep-learning model can discriminate between colonoscopy images of intestinal BD, CD, and ITB. In particular, the algorithm demonstrated superior discrimination ability for typical images. This approach presents a beneficial method for the differential diagnosis of the diseases.

FDG-PET-CT Enterography Helps Determine Clinical Significance of Suspected Ileocecal Thickening: A Prospective Study.

BACKGROUND: Ileocecal thickening (ICT) on imaging could result from diverse etiologies but may also be clinically insignificant. AIM: Evaluation of role of combined 2-deoxy-2-fluorine-18-fluoro-D-glucose(18F-FDG)-positron emission tomography and computed tomographic enterography (PET-CTE) for determination of clinical significance of suspected ICT. METHODS: This prospective study enrolled consecutive patients with suspected ICT on ultrasound. Patients were evaluated with PET-CTE and colonoscopy. The patients were divided into: Group A (clinically significant diagnosis) or Group B (clinically insignificant diagnosis) and compared for various clinical and radiological findings. The two groups were compared for maximum standardized uptake values of terminal ileum, ileo-cecal valve, cecum and overall. RESULTS: Of 34 patients included (23 males, mean age: 40.44 ± 15.40 years), 12 (35.3%) had intestinal tuberculosis, 11 (32.4%) Crohn's disease, 3 (8.8%) other infections, 1 (2.9%) malignancy, 4 (11.8%) non-specific terminal ileitis while 3 (8.8%) had normal colonoscopy and histology. The maximum standardized uptake value of the ileocecal area overall (SUVmax-ICT-overall) was significantly higher in Group A (7.16 ± 4.38) when compared to Group B (3.62 ± 9.50, P = 0.003). A cut-off of 4.50 for SUVmax-ICT-overall had a sensitivity of 70.37% and a specificity of 100% for prediction of clinically significant diagnosis. Using decision tree model, the SUVmax-ICT with a cut-off of 4.75 was considered appropriate for initial decision followed by the presence of mural thickening in the next node. CONCLUSION: PET-CTE can help in discrimination of clinically significant and insignificant diagnosis. It may help guide the need for colonoscopy in patients suspected to have ICT on CT.

Gastrointestinal tuberculosis, the great simulator. From inflammatory disease to a tumor.

A 52-year-old Senegalese male presented due to abdominal pain and fever, with findings suggestive of a stenosing tumor of the right colon. Biopsies during colonoscopy were compatible with adenocarcinoma and infiltration into neighboring organs was observed during the surgery. New biopsies were taken that did not show dysplasia but granulomatous foci that were suggestive of a non-filiated infection. Thus, tuberculosis was ruled out.

Tuberculosis esofágica simulando malignidad en la PET/TC 18F-FDG / Esophageal tuberculosis mimicking malignancy on 18F-FDG PET/CT

No disponible

Hodkgin lymphoma concomitant of tuberculosis, a therapeutic challenge for multidisciplinary management.

Hodgkin lymphoma (HL) is a disease characterized by a high curability rate, and the treatment benefit-risk balance must be carefully addressed to achieve complete disease control with low risk of long-term toxicities. Most patients are treated with a combination of chemotherapy and radiotherapy, after disease staging and response to treatment evaluated by FDG PET/CT. We report the case of a 28-year-old patient concomitantly diagnosed of a Hodgkin lymphoma and active tuberculosis. Initial staging was difficult due to pulmonary and abdominal tuberculosis localization that induced FDG PET/CT hypermetabolism. Anti-tuberculosis treatment was first started, allowing secondary an early accurate Hodgkin lymphoma staging by FDG PET/CT. The patient was then treated by chemotherapy and radiotherapy. Helical TomoTherapy® was used with involved site (IS) irradiation volume was performed to decrease the high doses to organs-at-risk (OAR), especially lungs in this context of tuberculosis.