Development of a multimodal machine-learning fusion model to non-invasively assess ileal Crohn's disease endoscopic activity.

BACKGROUND AND OBJECTIVE: Recurrent attentive non-invasive observation of intestinal inflammation is essential for the proper management of Crohn's disease (CD). The goal of this study was to develop and evaluate a multi-modal machine-learning (ML) model to assess ileal CD endoscopic activity by integrating information from Magnetic Resonance Enterography (MRE) and biochemical biomarkers. METHODS: We obtained MRE, biochemical and ileocolonoscopy data from the multi-center ImageKids study database. We developed an optimized multimodal fusion ML model to non-invasively assess terminal ileum (TI) endoscopic disease activity in CD from MRE data. We determined the most informative features for model development using a permutation feature importance technique. We assessed model performance in comparison to the clinically recommended linear-regression MRE model in an experimental setup that consisted of stratified 2-fold validation, repeated 50 times, with the ileocolonoscopy-based Simple Endoscopic Score for CD at the TI (TI SES-CD) as a reference. We used the predictions' mean-squared-error (MSE) and the receiver operation characteristics (ROC) area under curve (AUC) for active disease classification (TI SEC-CD≥3) as performance metrics. RESULTS: 121 subjects out of the 240 subjects in the ImageKids study cohort had all required information (Non-active CD: 62 [51%], active CD: 59 [49%]). Length of disease segment and normalized biochemical biomarkers were the most informative features. The optimized fusion model performed better than the clinically recommended model determined by both a better median test MSE distribution (7.73 vs. 8.8, Wilcoxon test, p<1e-5) and a better aggregated AUC over the folds (0.84 vs. 0.8, DeLong's test, p<1e-9). CONCLUSIONS: Optimized ML models for ileal CD endoscopic activity assessment have the potential to enable accurate and non-invasive attentive observation of intestinal inflammation in CD patients. The presented model is available at https://tcml-bme.github.io/ML_SESCD.html.

Development, Validation, and Evaluation of the Pediatric Inflammatory Crohn's Magnetic Resonance Enterography Index From the ImageKids Study.

BACKGROUND & AIMS: Cross-sectional imaging is important in the assessment of transmural inflammation in Crohn's disease (CD). Small bowel involvement is often more extensive in pediatric CD, requiring a panentering measuring tool. We undertook to develop a magnetic resonance enterography (MRE)-based index that would measure inflammation in all segments of the intestine, without rectal contrast. METHODS: Children with CD underwent ileocolonoscopy and MRE and half were prospectively followed for 18 months when MRE was repeated. Item generation and reduction were performed by a Delphi panel of pediatric radiologists, a systematic literature review, a cross-sectional study of 48 MREs, and a steering committee. Formatting and weighting were performed using multivariate modeling adjusted by a steering committee. MREs were read locally and centrally. Reliability, validity, and responsiveness were determined using several clinimetric and psychometric approaches. RESULTS: Thirty items were initially generated and reduced to 5 using regression analysis on 159 MREs: wall thickness, wall diffusion weighted imaging, ulcerations, mesenteric edema, and comb sign. In the validation cohort of 81 MREs, the weighted global PICMI correlated well with the radiologist global assessment (r = 0.85; P < .001) and with the simple endoscopic score in a subsample with ileocolonic disease (r = 0.63; P < .001). Interobserver and test-retest reliability were high (interclass correlation coefficients, 0.84; 95% confidence interval [CI], 0.79-0.87; and 0.81, 95% CI, 0.65-0.90, respectively; both P < .001). Excellent responsiveness was found at repeated visits (n = 116 MREs; area under the receiver operating characteristic curve 0.96; 95% CI, 0.93-0.99). Transmural healing was defined as PICMI ≤10 and response as a change of >20 points with excellent discriminative validity (area under the receiver operating characteristic curve = 0.96; 95% CI, 0.93-0.99). CONCLUSIONS: The PICMI is a valid, reliable, and responsive index for assessing transmural inflammation in pediatric CD. It scores the entire bowel length and does not require intravenous contrast or rectal enema and, therefore, is suitable for use in children. (ClinicalTrials.gov, Number: NCT01881490.).

Development and Validation of a Pediatric MRI-Based Perianal Crohn Disease (PEMPAC) Index-A Report from the ImageKids Study.

BACKGROUND: As part of the prospective multicenter ImageKids study, we aimed to develop and validate the pediatric MRI-based perianal Crohn disease (PEMPAC) index. METHODS: Children with Crohn disease with any clinical perianal findings underwent pelvic magnetic resonance imaging at 21 sites globally. The site radiologist and 2 central radiologists provided a radiologist global assessment (RGA) on a 100 mm visual analog scale and scored the items selected by a Delphi group of 35 international radiologists and a review of the literature. Two weighted multivariable statistical models were constructed against the RGA. RESULTS: Eighty children underwent 95 pelvic magnetic resonance imaging scans; 64 were used for derivation and 31 for validation. The following items were included: fistula number, location, length and T2 hyperintensity; abscesses; rectal wall involvement; and fistula branching. The last 2 items had negative beta scores and thus were excluded in a contending basic model. In the validation cohort, the full and the basic models had the same strong correlation with the RGA (r = 0.75; P < 0.01) and with the adult Van Assche index (VAI; r = 0.93 and 0.92; P < 0.001). The correlation of the VAI with the RGA was similar (r = 0.77; P < 0.01). The 2 models and the VAI had a similar ability to differentiate remission from active disease (area under the receiver operating characteristic curve, 0.91-0.94). The PEMPAC index had good responsiveness to change (area under the receiver operating characteristic curve, 0.89; 95% confidence interval, 0.69-1.00). CONCLUSIONS: Using a blended judgmental and mathematical approach, we developed and validated an index for quantifying the severity of perianal disease in children with CD. The adult VAI may also be used with confidence in children.

Clinical-stage Approaches for Imaging Chronic Inflammation and Fibrosis in Crohn's Disease.

The number of imaging-based indices developed for inflammatory bowel disease as research tools, objectively measuring ileocolonic and perianal activity and treatment response, has expanded in the past 2 decades. Created primarily to assess Crohn's disease (CD), there is increasing adoption of these indices into the clinical realm to guide patient care. This translation has been facilitated by validation in adult and pediatric populations, prompted by simplification of score calculations needed for practical application outside the research environment. The majority of these indices utilize magnetic resonance imaging (MRI), specifically MR enterography (MRE) and pelvic MRI, and more recently ultrasound. This review explores validated indices by modality, anatomic site and indication, including for documentation of the presence and extent of CD, disease progression, complications, and treatment response, highlighting those in clinical use or with the potential to be. As well, it details index imaging features used to quantify chronic inflammatory activity, severity, and to lesser extent fibrosis, in addition to their reference standards and any modifications. Validation in the pediatric population of indices primarily developed in adult cohorts such as the Magnetic Resonance Index of Activity (MaRIA), the Simplified Magnetic Resonance Index of Activity (MARIAs), and the MRE global score (MEGS), together with newly developed pediatric-specific indices, are discussed. Indices that may be predictive of disease course and investigational techniques with the potential to provide future imaging biomarkers, such as multiparametric MRI, are also briefly considered.

Pleural-Based Intrathoracic Cystic Lymphangioma in an Infant Mimicking a Pneumonia.

Cystic lymphangioma is an uncommon benign tumor that occurs primarily in children in the cervical region. We report the first case of a pleural-based cystic lymphangioma in an infant. The patient was admitted for RUL pneumonia. Because of the persistence of the radiographic findings despite clinical improvement, a computed tomography (CT) and a magnetic resonance imaging (MRI) scan were performed. They showed a multiloculated cystic lesion in the superior posterior right hemithorax. A surgical procedure was performed with complete resection of the tumor. Histopathological examination showed a pleural-based intrathoracic multicystic lymphangioma. One year after the surgery, the patient feels well without any sign of recurrence.

A Simple Endoscopic Score Modified for the Upper Gastrointestinal Tract in Crohn's Disease [UGI-SES-CD]: A Report From the ImageKids Study.

OBJECTIVE: There is no standardized endoscopic description of upper gastrointestinal [UGI] disease in Crohn's disease [CD]. We prospectively applied the Simple Endoscopic Score for CD [SES-CD] to the UGI tract as a planned sub-study of the multicentre prospective ImageKids study. We aimed to assess the utility of the UGI-SES-CD and its clinical significance in paediatric CD. DESIGN: Patients underwent an oesophagogastroduodenoscopy [EGD], ileocolonoscopy, and magnetic resonance enterography [MRE] with explicit clinical data recorded. SES-CD was scored at each region [oesophagus, stomach body, antrum, and duodenum]. Half of the patients were followed for 18 months, when a repeat MRE was performed. RESULTS: A total of 202 children were included 56% males, mean age 11.5 ± 3.2 years, median weighted Paediatric Crohn's Disease Activity Index [wPCDAI 25]). UGI-SES-CD score ranged 0-17, with 95 [47%] having a UGI-SES-CD ≥1; no narrowing was detected. UGI-SES-CD ≥1 was associated with higher: wPCDAI [32.5 vs 20; p = 0.03]; Physician's Global Assessment [PGA] of inflammation (45 mm visual analogue score [VAS] vs 30 mm VAS; p = 0.04); ileocolonoscopic SES-CD [10 vs 7; p = 0.004], faecal calprotectin [717 µg/g vs 654 µ/g; p= 0.046]; and radiological global assessment of damage by MRE [7 mm VAS vs 0; p = 0.04]. In all, 81 patients were followed for 18 months and no association was identified between initial UGI SES-CD and markers of disease course such as surgery, MRE assessment, or treatment escalation. CONCLUSION: UGI-SES-CD is an easily reported objective scoring system and is associated with a more severe disease phenotype but not with disease course.

Magnetic resonance enterography has good inter-rater agreement and diagnostic accuracy for detecting inflammation in pediatric Crohn disease.

BACKGROUND: Magnetic resonance enterography (MRE) is increasingly relied upon for noninvasive assessment of intestinal inflammation in Crohn disease. However very few studies have examined the diagnostic accuracy of individual MRE signs in children. OBJECTIVE: We have created an MR-based multi-item measure of intestinal inflammation in children with Crohn disease - the Pediatric Inflammatory Crohn's MRE Index (PICMI). To inform item selection for this instrument, we explored the inter-rater agreement and diagnostic accuracy of individual MRE signs of inflammation in pediatric Crohn disease and compared our findings with the reference standards of the weighted Pediatric Crohn's Disease Activity Index (wPCDAI) and C-reactive protein (CRP). MATERIALS AND METHODS: In this cross-sectional single-center study, MRE studies in 48 children with diagnosed Crohn disease (66% male, median age 15.5 years) were reviewed by two independent radiologists for the presence of 15 MRE signs of inflammation. Using kappa statistics we explored inter-rater agreement for each MRE sign across 10 anatomical segments of the gastrointestinal tract. We correlated MRE signs with the reference standards using correlation coefficients. Radiologists measured the length of inflamed bowel in each segment of the gastrointestinal tract. In each segment, MRE signs were scored as either binary (0-absent, 1-present), or ordinal (0-absent, 1-mild, 2-marked). These segmental scores were weighted by the length of involved bowel and were summed to produce a weighted score per patient for each MRE sign. Using a combination of wPCDAI≥12.5 and CRP≥5 to define active inflammation, we calculated area under the receiver operating characteristic curve (AUC) for each weighted MRE sign. RESULTS: Bowel wall enhancement, wall T2 hyperintensity, wall thickening and wall diffusion-weighted imaging (DWI) hyperintensity were most commonly identified. Inter-rater agreement was best for decreased motility and wall DWI hyperintensity (kappa≥0.64). Correlation between MRE signs and wPCDAI was higher than with CRP. AUC was highest (≥0.75) for ulcers, wall enhancement, wall thickening, wall T2 hyperintensity and wall DWI hyperintensity. CONCLUSION: Some MRE signs had good inter-rater agreement and AUC for detection of inflammation in children with Crohn disease.

Non-invasive diagnosis of liver fibrosis and cirrhosis.

The evaluation and follow up of liver fibrosis and cirrhosis have been traditionally performed by liver biopsy. However, during the last 20 years, it has become evident that this "gold-standard" is imperfect; even according to its proponents, it is only "the best" among available methods. Attempts at uncovering non-invasive diagnostic tools have yielded multiple scores, formulae, and imaging modalities. All are better tolerated, safer, more acceptable to the patient, and can be repeated essentially as often as required. Most are much less expensive than liver biopsy. Consequently, their use is growing, and in some countries the number of biopsies performed, at least for routine evaluation of hepatitis B and C, has declined sharply. However, the accuracy and diagnostic value of most, if not all, of these methods remains controversial. In this review for the practicing physician, we analyze established and novel biomarkers and physical techniques. We may be witnessing in recent years the beginning of the end of the first phase for the development of non-invasive markers. Early evidence suggests that they might be at least as good as liver biopsy. Novel experimental markers and imaging techniques could produce a dramatic change in diagnosis in the near future.

Pediatric vasitis: A rare complication of epididymitis.

Vasitis represents an inflammation of the vas deferens. This is a rare entity seen mostly in adult males following local surgery (e.g., vasectomy, hernia repair). Children with groin masses have a wide differential diagnosis. We describe a child with a groin mass following epididymitis diagnosed with vasitis and review the known literature regarding diagnostic tools and treatment. Vasitis in children, although rare, can be seen as a complication of epididymitis.