NOD2 Polymorphisms May Direct a Crohn Disease Phenotype in Patients With Very Early-Onset Inflammatory Bowel Disease.

NOD2/CARD15 was the first susceptibility gene recognized for adult-onset Crohn's (or Crohn) disease (CD). Recessive inheritance of NOD2 polymorphisms has been implicated as a mechanistic driver of pediatric-onset CD. In patients with very early-onset inflammatory bowel disease (VEO-IBD), however, the clinical relevance of NOD2 polymorphisms has not been fully established. Ten VEO-IBD patients with NOD2 polymorphisms ( NOD2 +) were compared to 16 VEO-IBD patients without genetic variants in NOD2 or any other VEO-IBD susceptibility genes ( NOD2 -). The majority of NOD2 + patients exhibited a CD-like phenotype (90%), linear growth impairment (90%), and arthropathy (60%), all of which were significantly more common than in the NOD2 - group ( P = 0.037, P = 0.004, P = 0.026, respectively). We propose that the presence of NOD2 polymorphisms in patients with VEO-IBD might confer a CD-like phenotype, linear growth impairment, and arthropathy. These findings should be validated in larger cohorts and may guide precision medicine for patients with VEO-IBD in the future.

Utilization of Antitumor Necrosis Factor Biologics in Very Early Onset Inflammatory Bowel Disease: A Multicenter Retrospective Cohort Study From North America.

BACKGROUND: Research on the utilization and effectiveness of antitumor necrosis factor (TNF) biologics in children with very early onset inflammatory bowel disease (VEOIBD) is urgently needed. Here we describe anti-TNF use and durability in a multicenter cohort. METHODS: We performed a retrospective cohort study of patients diagnosed with VEOIBD (<6 years) between 2008 and 2013 at 25 North American centers. We performed chart abstraction at diagnosis and 1, 3, and 5 years after diagnosis. We examined the rate of initiation and durability of infliximab and adalimumab and evaluated associations between treatment durability and the following covariates with multivariate Cox proportional hazard regression: age at diagnosis, sex, disease duration, disease classification, and presence of combined immunomodulatory treatment versus monotherapy. RESULTS: Of 294 children with VEOIBD, 120 initiated treatment with anti-TNF therapy and 101 had follow-up data recorded [50% Crohn disease (CD), 31% ulcerative colitis (UC), and 19% IBD unclassified (IBD-U)]. The cumulative probability of anti-TNF treatment was 15% at 1 year, 30% at 3 years, and 45% at 5 years from diagnosis; 56 (55%) were treated between 0 and 6 years old. Anti-TNF durability was 90% at 1 year, 75% at 3 years, and 55% at 5 years. The most common reason for discontinuation of anti-TNF were loss of response in 24 (57%) children. Children with UC/IBD-U had lower durability than those with CD (hazard ratio [HR] 0.17; 95% confidence interval [CI], 0.06-0.51; P = 0.001). CONCLUSIONS: Utilization and durability of anti-TNF in VEOIBD is relatively high and comparable with older children. Having Crohn disease (compared with UC/IBD-U) is associated with greater durability.

High Seroconversion Rate Against Severe Acute Respiratory Syndrome Coronavirus 2 in Symptomatic Pediatric Inflammatory Bowel Disease Patients.

ABSTRACT: Understanding coronavirus disease 2019 (COVID-19) in pediatric inflammatory bowel disease (PIBD) is important. We describe a single-center cohort of COVID-19 PIBD patients where seroconversion against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was examined.Immunosuppressed PIBD patients at Texas Children's Hospital who tested positive for SARS-CoV-2 by nasopharyngeal reverse transcriptase polymerase chain reaction were included in the study. The clinical course of IBD, concurrent medications, COVID-19 related symptoms, SARS-CoV-2 testing date, and SARS-CoV-2 immunoglobulin G (IgG) antibody testing date and result were examined. Of 14 SARS-CoV-2 positive PIBD patients, 12 were tested for qualitative anti-SARS-CoV-2 IgG (seven with transient COVID-19 symptoms, five asymptomatic). All symptomatic (7/7) and 60% of asymptomatic (3/5) patients seroconverted. No patients required hospitalization attributed to COVID-19.High rates of COVID-19 seroconversion occurred in immunosuppressed and symptomatic PIBD patients. More research to evaluate the significance of COVID-19 seroconversion is needed.

Early-onset lymphoproliferation and autoimmunity caused by germline STAT3 gain-of-function mutations.

Germline loss-of-function mutations in the transcription factor signal transducer and activator of transcription 3 (STAT3) cause immunodeficiency, whereas somatic gain-of-function mutations in STAT3 are associated with large granular lymphocytic leukemic, myelodysplastic syndrome, and aplastic anemia. Recently, germline mutations in STAT3 have also been associated with autoimmune disease. Here, we report on 13 individuals from 10 families with lymphoproliferation and early-onset solid-organ autoimmunity associated with 9 different germline heterozygous mutations in STAT3. Patients exhibited a variety of clinical features, with most having lymphadenopathy, autoimmune cytopenias, multiorgan autoimmunity (lung, gastrointestinal, hepatic, and/or endocrine dysfunction), infections, and short stature. Functional analyses demonstrate that these mutations confer a gain-of-function in STAT3 leading to secondary defects in STAT5 and STAT1 phosphorylation and the regulatory T-cell compartment. Treatment targeting a cytokine pathway that signals through STAT3 led to clinical improvement in 1 patient, suggesting a potential therapeutic option for such patients. These results suggest that there is a broad range of autoimmunity caused by germline STAT3 gain-of-function mutations, and that hematologic autoimmunity is a major component of this newly described disorder. Some patients for this study were enrolled in a trial registered at www.clinicaltrials.gov as #NCT00001350.

Effects of Recording Food Intake Using Cell Phone Camera Pictures on Energy Intake and Food Choice.

BACKGROUND: The well-documented increases in obesity and unhealthy dietary practices substantiate the need for evidence-based tools that can help people improve their dietary habits. The current spread of mobile phone-embedded cameras offers new opportunities for recording food intake. Moreover, the act of taking pictures of food consumed may enhance visual consciousness of food choice and quantity. The present study aimed to assess the effect of using cell phone pictures to record food intake on energy intake and food choice in college students. The effectiveness and acceptability of cell phone picture-based diet recording also was assessed. METHODS: A repeated measures crossover design was used. One group of participants entered their food intake online during 3 days based on their memory, although a second group recorded their food intake using cell phone pictures as their reference. Participants then crossed over to complete 3 more days of diet recording using the alternate method. Focus groups were conducted to obtain feedback on the effectiveness and acceptability of cell phone picture-based diet recording. RESULTS: Intake of meat and vegetable servings were significantly higher in the memory period compared with the cell phone period, regardless of the order. Results from the focus group indicated a positive attitude toward the use of cell phone pictures in recording food intake and an increased awareness of food choice and portion size. LINKING EVIDENCE TO ACTION: Cell phone pictures may be an easy, relevant, and accessible method of diet self-monitoring when aiming at dietary changes. Future trials should combine this technique with healthy eating education.

Neuronavigation using three-dimensional proton magnetic resonance spectroscopy data.

BACKGROUND: Applications in clinical medicine can benefit from fusion of spectroscopy data with anatomical imagery. For example, new 3-dimensional (3D) spectroscopy techniques allow for improved correlation of metabolite profiles with specific regions of interest in anatomical tumor images, which can be useful in characterizing and treating heterogeneous tumors that appear structurally homogeneous. OBJECTIVES: We sought to develop a clinical workflow and uniquely capable custom software tool to integrate advanced 3-tesla 3D proton magnetic resonance spectroscopic imaging ((1)H-MRSI) into industry standard image-guided neuronavigation systems, especially for use in brain tumor surgery. METHODS: (1)H-MRSI spectra from preoperative scanning on 15 patients with recurrent or newly diagnosed meningiomas were processed and analyzed, and specific voxels were selected based on their chemical contents. 3D neuronavigation overlays were then generated and applied to anatomical image data in the operating room. The proposed 3D methods fully account for scanner calibration and comprise tools that we have now made publicly available. RESULTS: The new methods were quantitatively validated through a phantom study and applied successfully to mitigate biopsy uncertainty in a clinical study of meningiomas. CONCLUSIONS: The proposed methods improve upon the current state of the art in neuronavigation through the use of detailed 3D (1)H-MRSI data. Specifically, 3D MRSI-based overlays provide comprehensive, quantitative visual cues and location information during neurosurgery, enabling a progressive new form of online spectroscopy-guided neuronavigation.

Reduced-complexity Convolutional Neural Network in the compressed domain.

Deep neural networks have achieved outstanding performance in computer vision tasks. Convolutional Neural Networks (CNNs) typically operate in the spatial domain with raw images, but in practice, images are usually stored and transmitted in their compressed representation where JPEG is one of the most widely used encoder. Also, these networks are computationally intensive and slow. This paper proposes performing the learning and inference processes in the compressed domain in order to reduce the computational complexity and improve the speed of popular CNNs. For this purpose, a novel graph-based frequency channel selection method is proposed to identify and select the most important frequency channels. The computational complexity is reduced by retaining the important frequency components and discarding the insignificant ones as well as eliminating the unnecessary layers of the network. Experimental results show that the modified ResNet-50 operating in the compressed domain is up to 70% faster than the spatial-based traditional ResNet-50 while resulting in similar classification accuracy. Moreover, this paper proposes a preprocessing step with partial encoding to improve the resilience to distortions caused by low-quality encoded images. Finally, we show that training a network with highly compressed data can achieve a good classification accuracy with up to 93% reduction in the storage requirements of the training data.

A Single-Center Study of Long-Term Effectiveness of Vedolizumab in Anti-TNF Refractory Pediatric Inflammatory Bowel Disease.

Vedolizumab is an anti-α4ß7 integrin antibody that has been used successfully in the treatment of adult-onset inflammatory bowel diseases (IBDs: Crohn disease [CD] and ulcerative colitis [UC]). Its off-label use in the pediatric IBD (PIBD) population is increasing, but knowledge on durability beyond 6 months of treatment is limited. Methods: A real-life, single-center, retrospective study of PIBD patients treated with vedolizumab was performed. Data on demographics, prior and concomitant treatments, and disease activity were obtained at 14 weeks, 26 weeks, 1 year, and 2 years of therapy. Primary outcome was corticosteroid- and other biologic-free remission (based on pediatric ulcerative colitis activity index [PUCAI]). Results: Thirty-nine patients were studied. By 1 year, 65% of CD and 68% of UC patients continued on vedolizumab therapy. Corticosteroid- and other biologic-free remission was 29% in CD and 16% in UC. By 2 years, 36% of CD and 47% of UC patients continued therapy. Corticosteroid- and other biologic-free remission was 21% in CD and 40% in UC. By 2 years, 80% of CD and 100% of UC patients were on intensified treatment regimen compared to the manufacturer guidance. Nine patients (23%) required surgical intervention within 26 months of starting vedolizumab indicating the severity of IBD in this cohort. Conclusions: Vedolizumab is a useful therapeutic modality in PIBD patients refractory to anti-TNF therapy, although with declining effectiveness by 2 years. Intensified treatment regimens are associated with long-term durability. Larger prospective trials in children are warranted.

Frequency-Tuned Universal Adversarial Attacks on Texture Recognition.

Although deep neural networks (DNNs) have been shown to be susceptible to image-agnostic adversarial attacks on natural image classification problems, the effects of such attacks on DNN-based texture recognition have yet to be explored. As part of our work, we find that limiting the perturbation's lp norm in the spatial domain may not be a suitable way to restrict the perceptibility of universal adversarial perturbations for texture images. Based on the fact that human perception is affected by local visual frequency characteristics, we propose a frequency-tuned universal attack method to compute universal perturbations in the frequency domain. Our experiments indicate that our proposed method can produce less perceptible perturbations yet with a similar or higher white-box fooling rates on various DNN texture classifiers and texture datasets as compared to existing universal attack techniques. We also demonstrate that our approach can improve the attack robustness against defended models as well as the cross-dataset transferability for texture recognition problems.

Infliximab Can Be Effective in Pediatric Patients with Ulcerative Colitis and Primary Failure of Adalimumab.

OBJECTIVE: The FDA approved adalimumab, an anti-tumor necrosis factor alpha agent, for the treatment of moderate to severe pediatric ulcerative colitis (UC) in February 2021. There are no real-world publications, however, on adalimumab as the first-line biologic in pediatric UC, a form of chronic inflammatory bowel disease (IBD). METHODS: A retrospective review was conducted to characterize the clinical courses of nine patients with moderate to severe pediatric UC who received adalimumab as their first biologic and had documented drug monitoring trough levels. RESULTS: Seven of the nine patients, or 78%, were switched from adalimumab to another therapy due to continued symptoms or steroid-dependence at an average of 5 months from initiation. Six of these seven patients, or 86%, had adalimumab drug trough levels in the consensus therapeutic range. Three patients were successfully switched to infliximab. CONCLUSIONS: Both the Crohn's disease-based dosing and the new FDA-approved pediatric UC dosing of adalimumab were ineffective in inducing remission in the majority of patients in our case series. This study indicates that further real-world observations are needed to optimize and position adalimumab in the treatment paradigm of moderate to severe pediatric UC.

Untargeted metabolomic profiling in a patient with glycogen storage disease Ib receiving empagliflozin treatment.

Glycogen storage disease type Ib (GSD-Ib) is a rare inborn error of glycogen metabolism uniquely associated with neutropenia and neutrophil dysfunction, causing severe infections, inflammatory bowel disease (IBD), and impaired wound healing. Recently, kidney sodium-glucose co-transporter-2 (SGLT2) inhibitors such as empagliflozin known to reduce plasma levels of 1,5-anhydroglucitol (1,5-AG) and its toxic derivatives in neutrophils, have been described as a new treatment option in case reports of patients with GSD-Ib from Europe and Asia. We report our experience with an 11-year-old girl with GSD-Ib presenting with short fasting hypoglycemia, neutropenia with neutrophil dysfunction, recurrent infections, suboptimal growth, iron-deficiency anemia, and IBD. Treatment with daily empagliflozin improved neutrophil counts and function with a significant reduction in G-CSF needs. Significant improvement in IBD has led to weight gain with improved nutritional markers and improved fasting tolerance. Reduction of maximum empagliflozin dose was needed due to arthralgia. No other significant side effects of empagliflozin were observed. This report uniquely highlights the novel use of untargeted metabolomics profiling for monitoring plasma levels of 1,5-AG to assess empagliflozin dose responsiveness and guide dietary management and G-CSF therapy. Clinical improvement correlated to rapid normalization of 1,5-AG levels in plasma sustained after dose reduction. In conclusion, empagliflozin appeared to be a safe treatment option for GSD-Ib-associated neutropenia and neutrophil dysfunction. Global untargeted metabolomics is an efficient method to assess biochemical responsiveness to treatment.

It GAN Do Better: GAN-Based Detection of Objects on Images With Varying Quality.

In this paper, we propose a novel generative framework which uses Generative Adversarial Networks (GANs) to generate features that provide robustness for object detection on reduced-quality images. The proposed GAN-based Detection of Objects (GAN-DO) framework is not restricted to any particular architecture and can be generalized to several deep neural network (DNN) based architectures. The resulting deep neural network maintains the exact architecture as the selected baseline model without adding to the model parameter complexity or inference speed. We first evaluate the effect of image quality on both object classification and object bounding box regression. We then test the models resulting from our proposed GAN-DO framework, using two state-of-the-art object detection architectures as the baseline models. We also evaluate the effect of the number of re-trained parameters in the generator of GAN-DO on the accuracy of the final trained model. Performance results provided using GAN-DO on object detection datasets establish an improved robustness to varying image quality and a higher mAP compared to the existing approaches.

Natural History of Very Early Onset Inflammatory Bowel Disease in North America: A Retrospective Cohort Study.

BACKGROUND: The incidence of very early onset inflammatory bowel disease (VEOIBD) is increasing, yet the phenotype and natural history of VEOIBD are not well described. METHODS: We performed a retrospective cohort study of patients diagnosed with VEOIBD (6 years of age and younger) between 2008 and 2013 at 25 North American centers. Eligible patients at each center were randomly selected for chart review. We abstracted data at diagnosis and at 1, 3, and 5 years after diagnosis. We compared the clinical features and outcomes with VEOIBD diagnosed younger than 3 years of age with children diagnosed with VEOIBD at age 3 to 6 years. RESULTS: The study population included 269 children (105 [39%] Crohn's disease, 106 [39%] ulcerative colitis, and 58 [22%] IBD unclassified). The median age of diagnosis was 4.2 years (interquartile range 2.9-5.2). Most (94%) Crohn's disease patients had inflammatory disease behavior (B1). Isolated colitis (L2) was the most common disease location (70% of children diagnosed younger than 3 years vs 43% of children diagnosed 3 years and older; P = 0.10). By the end of follow-up, stricturing/penetrating occurred in 7 (6.6%) children. The risk of any bowel surgery in Crohn's disease was 3% by 1 year, 12% by 3 years, and 15% by 5 years and did not differ by age at diagnosis. Most ulcerative colitis patients had pancolitis (57% of children diagnosed younger than 3 years vs 45% of children diagnosed 3 years and older; P = 0.18). The risk of colectomy in ulcerative colitis/IBD unclassified was 0% by 1 year, 3% by 3 years, and 14% by 5 years and did not differ by age of diagnosis. CONCLUSIONS: Very early onset inflammatory bowel disease has a distinct phenotype with predominantly colonic involvement and infrequent stricturing/penetrating disease. The cumulative risk of bowel surgery in children with VEOIBD was approximately 14%-15% by 5 years. These data can be used to provide anticipatory guidance in this emerging patient population.

Progression to colectomy in the era of biologics: A single center experience with pediatric ulcerative colitis.

BACKGROUND/PURPOSE: Clinical outcomes in pediatric ulcerative colitis (UC) in the era of biologic agents are poorly defined. We aimed to describe risk factors for colectomy in pediatric UC in the era of infliximab therapy. METHODS: We reviewed 217 pediatric patients at Texas Children's Hospital with newly diagnosed UC between 2003 and 2015; 117 had a minimum of 5 years of follow-up. Extent of disease at diagnosis, medication exposure, the presence of extraintestinal manifestations (EIMs), and need for surgery were noted. RESULTS: Average length of follow up was 5.02 ±â€¯2.27 years. Forty-two percent presented with pancolitis. Infliximab was used in 39%, immunomodulators in 65%, and steroids in 89% of patients. EIMs occurred in 24.9% of patients. The cumulative rate of colectomy was 12.9% at 5 years. Children presenting as E2 (Paris Classification) and children prescribed oral steroid monotherapy at diagnosis progressed to surgery faster than any other group. Of the children who received infliximab, females and children less than 5 years old were less likely to respond to therapy. CONCLUSIONS: The natural course of pediatric UC remains aggressive despite the addition of infliximab to the standard of care and suggests a need for early aggressive clinical intervention. LEVEL-OF-EVIDENCE RATING: Level IV.

A no-reference objective image sharpness metric based on the notion of just noticeable blur (JNB).

This work presents a perceptual-based no-reference objective image sharpness/blurriness metric by integrating the concept of just noticeable blur into a probability summation model. Unlike existing objective no-reference image sharpness/blurriness metrics, the proposed metric is able to predict the relative amount of blurriness in images with different content. Results are provided to illustrate the performance of the proposed perceptual-based sharpness metric. These results show that the proposed sharpness metric correlates well with the perceived sharpness being able to predict with high accuracy the relative amount of blurriness in images with different content.

DeepCorrect: Correcting DNN Models Against Image Distortions.

In recent years, the widespread use of deep neural networks (DNNs) has facilitated great improvements in performance for computer vision tasks like image classification and object recognition. In most realistic computer vision applications, an input image undergoes some form of image distortion such as blur and additive noise during image acquisition or transmission. Deep networks trained on pristine images perform poorly when tested on such distortions. In this paper, we evaluate the effect of image distortions like Gaussian blur and additive noise on the activations of pre-trained convolutional filters. We propose a metric to identify the most noise susceptible convolutional filters and rank them in order of the highest gain in classification accuracy upon correction. In our proposed approach called DeepCorrect, we apply small stacks of convolutional layers with residual connections at the output of these ranked filters and train them to correct the worst distortion affected filter activations, while leaving the rest of the pre-trained filter outputs in the network unchanged. Performance results show that applying DeepCorrect models for common vision tasks like image classification (ImageNet), object recognition (Caltech-101, Caltech-256), and scene classification (SUN-397), significantly improves the robustness of DNNs against distorted images and outperforms other alternative approaches.

Liver biopsy results in patients with sickle cell disease on chronic transfusions: poor correlation with ferritin levels.

BACKGROUND: Chronic transfusions are effective in preventing stroke and other complications of sickle cell disease. The aim of this study was to determine whether serum ferritin levels correlated with liver iron content in sickle cell patients on chronic transfusion. PROCEDURE: Forty-four liver biopsy specimens from 38 patients with homozygous sickle cell anemia (HbSS) and one patient with sickle thalassemia receiving chronic transfusions were studied. Five patients underwent a second liver biopsy for follow up. Three ferritin measurements were used to calculate a mean for each patient. The association between serum ferritin levels and liver iron quantitation was measured using the Spearman rank correlation, and sensitivity and specificity were determined for selected threshold values of serum ferritin. RESULTS: Serum ferritin levels ranged from 515 to 6076 ng/ml, liver iron concentration ranged from 1.8 to 67.97 mg/g dry weight. The amount of iron per gram liver dry weight was moderately correlated with serum ferritin values (r = 0.46). The correlation of duration of transfusion with serum ferritin (r = 0.40) and with liver iron content (r = 0.41) also indicated moderate correlation. Liver biopsy results led to changes in the management after 29/44 (66%) of the biopsies. Serum ferritin >/=2500 ng/ml predicted high liver iron content (>/=7 mg/g), with a sensitivity of 62.5% and a specificity of 77.8%. CONCLUSION: We found a poor correlation between serum ferritin levels and liver iron content (LIC). Despite being on chelation therapy, many patients on chronic transfusion had high levels of liver iron. Measurement of LIC is highly recommended in these patients.

Visual Saliency Prediction Using a Mixture of Deep Neural Networks.

Visual saliency models have recently begun to incorporate deep learning to achieve predictive capacity much greater than previous unsupervised methods. However, most existing models predict saliency without explicit knowledge of global scene semantic information. We propose a model (MxSalNet) that incorporates global scene semantic information in addition to local information gathered by a convolutional neural network. Our model is formulated as a mixture of experts. Each expert network is trained to predict saliency for a set of closely related images. The final saliency map is computed as a weighted mixture of the expert networks' output, with weights determined by a separate gating network. This gating network is guided by global scene information to predict weights. The expert networks and the gating network are trained simultaneously in an end-toend manner. We show that our mixture formulation leads to improvement in performance over an otherwise identical nonmixture model that does not incorporate global scene information. Additionally, we show that our model achieves better performance than several other visual saliency models.

Quality Robust Mixtures of Deep Neural Networks.

We study deep neural networks for classification of images with quality distortions. Deep network performance on poor quality images can be greatly improved if the network is fine-tuned with distorted data. However, it is difficult for a single fine-tuned network to perform well across multiple distortion types. We propose a mixture of experts based ensemble method, MixQualNet, that is robust to multiple different types of distortions. The "experts" in our model are trained on a particular type of distortion. The output of the model is a weighted sum of the expert models, where the weights are determined by a separate gating network. The gating network is trained to predict weights for a particular distortion type and level. During testing, the network is blind to the distortion level and type, yet can still assign appropriate weights to the expert models. In order to reduce the computational complexity, we introduce weight sharing into the MixQualNet. We utilize the TreeNet weight sharing architecture as well as introduce the Inverted TreeNet architecture. While both weight sharing architectures reduce memory requirements, our proposed Inverted TreeNet also achieves improved accuracy.