Interpretable prediction of necrotizing enterocolitis from machine learning analysis of premature infant stool microbiota

Artificial Intelligence-Based Diagnostic Support System for Patent Ductus Arteriosus in Premature Infants

Background: Patent ductus arteriosus (PDA) is a prevalent congenital heart defect in premature infants, associated with significant morbidity and mortality. Accurate and timely diagnosis of PDA is crucial, given the vulnerability of this population. Methods: We introduce an artificial intelligence (AI)-based PDA diagnostic support system designed to assist medical professionals in diagnosing PDA in premature infants. This study utilized electronic health record (EHR) data from 409 premature infants spanning a decade at Severance Children's Hospital. Our system integrates a data viewer, data analyzer, and AI-based diagnosis supporter, facilitating comprehensive data presentation, analysis, and early symptom detection. Results: The system's performance was evaluated through diagnostic tests involving medical professionals. This early detection model achieved an accuracy rate of up to 84%, enabling detection up to 3.3 days in advance. In diagnostic tests, medical professionals using the system with the AI-based diagnosis supporter outperformed those using the system without the supporter. Conclusions: Our AI-based PDA diagnostic support system offers a comprehensive solution for medical professionals to accurately diagnose PDA in a timely manner in premature infants. The collaborative integration of medical expertise and technological innovation demonstrated in this study underscores the potential of AI-driven tools in advancing neonatal diagnosis and care.

Proportion-based normalizations outperform compositional data transformations in machine learning applications

BACKGROUND

Normalization, as a pre-processing step, can significantly affect the resolution of machine learning analysis for microbiome studies. There are countless options for normalization scheme selection. In this study, we examined compositionally aware algorithms including the additive log ratio (alr), the centered log ratio (clr), and a recent evolution of the isometric log ratio (ilr) in the form of balance trees made with the PhILR R package. We also looked at compositionally naïve transformations such as raw counts tables and several transformations that are based on relative abundance, such as proportions, the Hellinger transformation, and a transformation based on the logarithm of proportions (which we call "lognorm").

RESULTS

In our evaluation, we used 65 metadata variables culled from four publicly available datasets at the amplicon sequence variant (ASV) level with a random forest machine learning algorithm. We found that different common pre-processing steps in the creation of the balance trees made very little difference in overall performance. Overall, we found that the compositionally aware data transformations such as alr, clr, and ilr (PhILR) performed generally slightly worse or only as well as compositionally naïve transformations. However, relative abundance-based transformations outperformed most other transformations by a small but reliably statistically significant margin.

CONCLUSIONS

Our results suggest that minimizing the complexity of transformations while correcting for read depth may be a generally preferable strategy in preparing data for machine learning compared to more sophisticated, but more complex, transformations that attempt to better correct for compositionality. Video Abstract.

Artificial Intelligence in Pediatrics: Learning to Walk Together

In this era of rapidly advancing technology, artificial intelligence (AI) has emerged as a transformative force, even being called the Fourth Industrial Revolution, along with gene editing and robotics. While it has undoubtedly become an increasingly important part of our daily lives, it must be recognized that it is not an additional tool, but rather a complex concept that poses a variety of challenges. AI, with considerable potential, has found its place in both medical care and clinical research. Within the vast field of pediatrics, it stands out as a particularly promising advancement. As pediatricians, we are indeed witnessing the impactful integration of AI-based applications into our daily clinical practice and research efforts. These tools are being used for simple to more complex tasks such as diagnosing clinically challenging conditions, predicting disease outcomes, creating treatment plans, educating both patients and healthcare professionals, and generating accurate medical records or scientific papers. In conclusion, the multifaceted applications of AI in pediatrics will increase efficiency and improve the quality of healthcare and research. However, there are certain risks and threats accompanying this advancement including the biases that may contribute to health disparities and, inaccuracies. Therefore, it is crucial to recognize and address the technical, ethical, and legal challenges as well as explore the benefits in both clinical and research fields.

Reassessing acquired neonatal intestinal diseases using unsupervised machine learning

BACKGROUND

Acquired neonatal intestinal diseases have an array of overlapping presentations and are often labeled under the dichotomous classification of necrotizing enterocolitis (which is poorly defined) or spontaneous intestinal perforation, hindering more precise diagnosis and research. The objective of this study was to take a fresh look at neonatal intestinal disease classification using unsupervised machine learning.

METHODS

Patients admitted to the University of Florida Shands Neonatal Intensive Care Unit January 2013-September 2019 diagnosed with an intestinal injury, or had imaging findings of portal venous gas, pneumatosis, abdominal free air, or had an abdominal drain placed or exploratory laparotomy during admission were included. Congenital gastroschisis, omphalocele, intestinal atresia, malrotation were excluded. Data was collected via retrospective chart review with subsequent hierarchal, unsupervised clustering analysis.

RESULTS

Five clusters of intestinal injury were identified: Cluster 1 deemed the "Low Mortality" cluster, Cluster 2 deemed the "Mature with Inflammation" cluster, Cluster 3 deemed the "Immature with High Mortality" cluster, Cluster 4 deemed the "Late Injury at Full Feeds" cluster, and Cluster 5 deemed the "Late Injury with High Rate of Intestinal Necrosis" cluster.

CONCLUSION

Unsupervised machine learning can be used to cluster acquired neonatal intestinal injuries. Future study with larger multicenter datasets is needed to further refine and classify types of intestinal diseases.

IMPACT

Unsupervised machine learning can be used to cluster types of acquired neonatal intestinal injury. Five major clusters of acquired neonatal intestinal injury are described, each with unique features. The clusters herein described deserve future, multicenter study to determine more specific early biomarkers and tailored therapeutic interventions to improve outcomes of often devastating neonatal acquired intestinal injuries.

Transforming neonatal care with artificial intelligence: challenges, ethical consideration, and opportunities

Artificial intelligence (AI) offers tremendous potential to transform neonatology through improved diagnostics, personalized treatments, and earlier prevention of complications. However, there are many challenges to address before AI is ready for clinical practice. This review defines key AI concepts and discusses ethical considerations and implicit biases associated with AI. Next we will review literature examples of AI already being explored in neonatology research and we will suggest future potentials for AI work. Examples discussed in this article include predicting outcomes such as sepsis, optimizing oxygen therapy, and image analysis to detect brain injury and retinopathy of prematurity. Realizing AI's potential necessitates collaboration between diverse stakeholders across the entire process of incorporating AI tools in the NICU to address testability, usability, bias, and transparency. With multi-center and multi-disciplinary collaboration, AI holds tremendous potential to transform the future of neonatology.

Crohn's Disease Prediction Using Sequence Based Machine Learning Analysis of Human Microbiome

Human microbiota refers to the trillions of microorganisms that inhabit our bodies and have been discovered to have a substantial impact on human health and disease. By sampling the microbiota, it is possible to generate massive quantities of data for analysis using Machine Learning algorithms. In this study, we employed several modern Machine Learning techniques to predict Inflammatory Bowel Disease using raw sequence data. The dataset was obtained from NCBI preprocessed graph representations and converted into a structured form. Seven well-known Machine Learning frameworks, including Random Forest, Support Vector Machines, Extreme Gradient Boosting, Light Gradient Boosting Machine, Gaussian Naïve Bayes, Logistic Regression, and k-Nearest Neighbor, were used. Grid Search was employed for hyperparameter optimization. The performance of the Machine Learning models was evaluated using various metrics such as accuracy, precision, fscore, kappa, and area under the receiver operating characteristic curve. Additionally, Mc Nemar's test was conducted to assess the statistical significance of the experiment. The data was constructed using k-mer lengths of 3, 4 and 5. The Light Gradient Boosting Machine model overperformed over other models with 67.24%, 74.63% and 76.47% accuracy for k-mer lengths of 3, 4 and 5, respectively. The LightGBM model also demonstrated the best performance in each metric. The study showed promising results predicting disease from raw sequence data. Finally, Mc Nemar's test results found statistically significant differences between different Machine Learning approaches.

Stability and heterogeneity in the antimicrobiota reactivity of human milk-derived immunoglobulin A

Immunoglobulin A (IgA) is secreted into breast milk and is critical for both protecting against enteric pathogens and shaping the infant intestinal microbiota. The efficacy of breast milk-derived maternal IgA (BrmIgA) is dependent upon its specificity; however, heterogeneity in BrmIgA binding ability to the infant microbiota is not known. Using a flow cytometric array, we analyzed the reactivity of BrmIgA against bacteria common to the infant microbiota and discovered substantial heterogeneity between all donors, independent of preterm or term delivery. Surprisingly, we also observed intradonor variability in the BrmIgA response to closely related bacterial isolates. Conversely, longitudinal analysis showed that the antibacterial BrmIgA reactivity was relatively stable through time, even between sequential infants, indicating that mammary gland IgA responses are durable. Together, our study demonstrates that the antibacterial BrmIgA reactivity displays interindividual heterogeneity but intraindividual stability. These findings have important implications for how breast milk shapes the development of the preterm infant microbiota and protects against necrotizing enterocolitis.

[Recent research on machine learning in the diagnosis and treatment of necrotizing enterocolitis in neonates]

Necrotizing enterocolitis (NEC), with the main manifestations of bloody stool, abdominal distension, and vomiting, is one of the leading causes of death in neonates, and early identification and diagnosis are crucial for the prognosis of NEC. The emergence and development of machine learning has provided the potential for early, rapid, and accurate identification of this disease. This article summarizes the algorithms of machine learning recently used in NEC, analyzes the high-risk predictive factors revealed by these algorithms, evaluates the ability and characteristics of machine learning in the etiology, definition, and diagnosis of NEC, and discusses the challenges and prospects for the future application of machine learning in NEC.

Development of an international core outcome set for treatment trials in necrotizing enterocolitis-a study protocol

AIM

Necrotizing enterocolitis (NEC) is the most lethal disease of the gastrointestinal tract of preterm infants. New and existing management strategies need clinical evaluation. Large heterogeneity exists in the selection, measurement, and reporting of outcome measures in NEC intervention studies. This hampers meta-analyses and the development of evidence-based management guidelines. We aim to develop a Core Outcome Set (COS) for NEC that includes the most relevant outcomes for patients and physicians, from moment of diagnosis into adulthood. This COS is designed for use in NEC treatment trials, in infants with confirmed NEC.

METHODS

This study is designed according to COS-STAD (Core Outcome Set-STAndards for Development) recommendations and the COMET (Core Outcome Measures in Effectiveness Trials) Initiative Handbook. We obtained a waiver from the Ethics Review Board and prospectively registered this study with COMET (Study 1920). We will approach 125 clinicians and/or researchers from low-middle and high-income countries based on their scientific output (using SCIVAL, a bibliometric tool). Patients and parents will be approached through local patient organisations. Participants will be separated into three panels, to assess differences in priorities between former patients and parents (1. lay panel), clinicians and researchers involved in the neonatal period (2. neonatal panel) and after the neonatal period (3. post-neonatal panel). They will be presented with outcomes currently used in NEC research, identified through a systematic review, in a Delphi process. Eligible outcome domains are also identified from the patients and parents' perspectives. Using a consensus process, including three online Delphi rounds and a final face-to-face consensus meeting, the COS will be finalised and include outcomes deemed essential to all stakeholders: health care professionals, parents and patients' representatives. The final COS will be reported in accordance with the COS-Standards for reporting (COS-STAR) statement.

CONCLUSIONS

Development of an international COS will help to improve homogeneity of outcome measure reporting in NEC, will enable adequate and efficient comparison of treatment strategies, and will help the interpretation and implementation of clinical trial results. This will contribute to high-quality evidence regarding the best treatment strategy for NEC in preterm infants.

State of the art review on machine learning and artificial intelligence in the study of neonatal necrotizing enterocolitis

Necrotizing Enterocolitis (NEC) is one of the leading causes of gastrointestinal emergency in preterm infants. Although NEC was formally described in the 1960's, there is still difficulty in diagnosis and ultimately treatment for NEC due in part to the multifactorial nature of the disease. Artificial intelligence (AI) and machine learning (ML) techniques have been applied by healthcare researchers over the past 30 years to better understand various diseases. Specifically, NEC researchers have used AI and ML to predict NEC diagnosis, NEC prognosis, discover biomarkers, and evaluate treatment strategies. In this review, we discuss AI and ML techniques, the current literature that has applied AI and ML to NEC, and some of the limitations in the field.

Enteral Feeding and Antibiotic Treatment Do Not Influence Increased Coefficient of Variation of Total Fecal Bile Acids in Necrotizing Enterocolitis

Introduction

Necrotizing enterocolitis (NEC) is the most common gastrointestinal emergency in preterm infants. In animal models, the accumulation of ileal bile acids (BAs) is a crucial component of NEC pathophysiology. Recently, we showed that the coefficient of variation of total fecal BAs (CV-TBA) was elevated in infants who develop NEC compared to matched controls. However, neither the type of enteral nutrition nor antibiotic treatments-parameters that could potentially influence BA levels-were used to match pairs. Thus, we assessed the relationships between exposure to enteral feeding types and antibiotic treatments with NEC status and CV-TBA.

Materials and methods

Serial fecal samples were collected from 79 infants born with birth weight (BW) ≤1800 gm and estimated gestational age (EGA) ≤32 weeks; eighteen of these infants developed NEC. Total fecal BA levels (TBA) were determined using a commercially available enzyme cycling kit. Relationships between CV-TBA and dichotomous variables (NEC status, demographics, early exposure variables) were assessed by independent samples t-tests. Fisher's exact tests were used to assess relationships between NEC status and categorical variables.

Results

High values for CV-TBA levels perfectly predicted NEC status among infants in this study. However, feeding type and antibiotic usage did not drive this relationship.

Conclusions

As in previous studies, high values for the CV-TBA levels in the first weeks of life perfectly predicted NEC status among infants. Importantly, feeding type and antibiotic usage-previously identified risk factors for NEC-did not drive this relationship.

Stability and heterogeneity in the anti-microbiota reactivity of human milk-derived Immunoglobulin A

Immunoglobulin A (IgA) is secreted into breast milk and is critical to both protecting against enteric pathogens and shaping the infant intestinal microbiota. The efficacy of breast milk-derived maternal IgA (BrmIgA) is dependent upon its specificity, however heterogeneity in BrmIgA binding ability to the infant microbiota is not known. Using a flow cytometric array, we analyzed the reactivity of BrmIgA against bacteria common to the infant microbiota and discovered substantial heterogeneity between all donors, independent of preterm or term delivery. We also observed intra-donor variability in the BrmIgA response to closely related bacterial isolates. Conversely, longitudinal analysis showed that the anti-bacterial BrmIgA reactivity was relatively stable through time, even between sequential infants, indicating that mammary gland IgA responses are durable. Together, our study demonstrates that the anti-bacterial BrmIgA reactivity displays inter-individual heterogeneity but intra-individual stability. These findings have important implications for how breast milk shapes the development of the infant microbiota and protects against Necrotizing Enterocolitis.

SUMMARY

We analyze the ability of breast milk-derived Immunoglobulin A (IgA) antibodies to bind the infant intestinal microbiota. We discover that each mother secretes into their breast milk a distinct set of IgA antibodies that are stably maintained over time.

Probiotic-educated Tregs are more potent than naïve Tregs for immune tolerance in stressed new-born mice

When new-born mice are subjected to acute maternal separation stress, cow-milk based formula feeding, and brief recurrent hypoxia with cold stress, they develop gut inflammation similar to the phenotype of neonatal necrotizing enterocolitis, characterised by an increase in gut mucosal effector T (Teffs) and reduced Foxp3⁺ regulatory T (Tregs) cells. The imbalance can be prevented by probiotic Limosilactobacillus reuteri DSM 17938 (LR 17938). We hypothesised that LR 17938 could potentiate a tolerogenic function of Tregs. To analyse whether LR 17938 can educate Tregs to improve their tolerogenic potency during neonatal stress, we isolated T cells (Tregs and Teffs) from 'donor' mice fed with either LR 17938 (10⁷ cfu) or control media. The cells were adoptively transferred (AT) by intraperitoneal injection (5 × 10⁵ cells/mouse) to new-born (d5) recipient mice. Mice were then separated from their dams, fed formula by gavage, and exposed to hypoxia and cold stress (NeoStress) for 4 days. We analysed the percentage of Tregs in CD4⁺T helper cells in the intestine (INT) and mesenteric lymph nodes (MLN) of recipient mice. We found that: (1) the percentage of Tregs in the INT and MLN following NeoStress were significantly reduced compared to dam-fed unstressed mice; (2) AT of either naïve Tregs or LR-educated Tregs to mice with Neostress increased the percentage of Tregs in the INT and MLN compared to the percentage in NeoStress mice without Treg treatment; however, LR-educated Tregs increased the Tregs significantly more than naïve Tregs; and (3) AT of LR-educated Tregs reduced pro-inflammatory CD44⁺Foxp3^-NonTregs and inflammatory CX3CR1⁺ dendritic cells in the intestinal mucosa of NeoStress mice. In conclusion, adoptive transfer of Tregs promotes the generation of and/or migration of endogenous Tregs in the intestinal mucosa of recipient mice. Importantly, probiotic-educated Tregs are more potent than naïve Tregs to enhance immune tolerance following neonatal stress.

Biomarkers of necrotizing enterocolitis in the era of machine learning and omics

Necrotizing enterocolitis (NEC) continues to be a major cause of morbidity and mortality in preterm infants. Despite decades of research in NEC, no reliable biomarkers can accurately diagnose NEC or predict patient prognosis. The recent emergence of multi-omics could potentially shift NEC biomarker discovery, particularly when evaluated using systems biology techniques. Furthermore, the use of machine learning and artificial intelligence in analyzing this 'big data' could enable novel interpretations of NEC subtypes, disease progression, and potential therapeutic targets, allowing for integration with personalized medicine approaches. In this review, we evaluate studies using omics technologies and machine learning in the diagnosis of NEC. Future implications and challenges inherent to the field are also discussed.

Necrotizing Enterocolitis: The Role of Hypoxia, Gut Microbiome, and Microbial Metabolites

Necrotizing enterocolitis (NEC) is a life-threatening disease that predominantly affects very low birth weight preterm infants. Development of NEC in preterm infants is accompanied by high mortality. Surgical treatment of NEC can be complicated by short bowel syndrome, intestinal failure, parenteral nutrition-associated liver disease, and neurodevelopmental delay. Issues surrounding pathogenesis, prevention, and treatment of NEC remain unclear. This review summarizes data on prenatal risk factors for NEC, the role of pre-eclampsia, and intrauterine growth retardation in the pathogenesis of NEC. The role of hypoxia in NEC is discussed. Recent data on the role of the intestinal microbiome in the development of NEC, and features of the metabolome that can serve as potential biomarkers, are presented. The Pseudomonadota phylum is known to be associated with NEC in preterm neonates, and the role of other bacteria and their metabolites in NEC pathogenesis is also discussed. The most promising approaches for preventing and treating NEC are summarized.

Using machine learning models to predict the effects of seasonal fluxes on Plesiomonas shigelloides population density

Seasonal variations (SVs) affect the population density (PD), fate, and fitness of pathogens in environmental water resources and the public health impacts. Therefore, this study is aimed at applying machine learning intelligence (MLI) to predict the impacts of SVs on P. shigelloides population density (PDP) in the aquatic milieu. Physicochemical events (PEs) and PDP from three rivers acquired via standard microbiological and instrumental techniques across seasons were fitted to MLI algorithms (linear regression (LR), multiple linear regression (MR), random forest (RF), gradient boosted machine (GBM), neural network (NN), K-nearest neighbour (KNN), boosted regression tree (BRT), extreme gradient boosting (XGB) regression, support vector regression (SVR), decision tree regression (DTR), M5 pruned regression (M5P), artificial neural network (ANN) regression (with one 10-node hidden layer (ANN10), two 6- and 4-node hidden layers (ANN64), and two 5- and 5-node hidden layers (ANN55)), and elastic net regression (ENR)) to assess the implications of the SVs of PEs on aquatic PDP. The results showed that SVs significantly influenced PDP and PEs in the water (p < 0.0001), exhibiting a site-specific pattern. While MLI algorithms predicted PDP with differing absolute flux magnitudes for the contributing variables, DTR predicted the highest PDP value of 1.707 log unit, followed by XGB (1.637 log unit), but XGB (mean-squared-error (MSE) = 0.0025; root-mean-squared-error (RMSE) = 0.0501; R² =0.998; medium absolute deviation (MAD) = 0.0275) outperformed other models in terms of regression metrics. Temperature and total suspended solids (TSS) ranked first and second as significant factors in predicting PDP in 53.3% (8/15) and 40% (6/15), respectively, of the models, based on the RMSE loss after permutations. Additionally, season ranked third among the 7 models, and turbidity (TBS) ranked fourth at 26.7% (4/15), as the primary significant factor for predicting PDP in the aquatic milieu. The results of this investigation demonstrated that MLI predictive modelling techniques can promisingly be exploited to complement the repetitive laboratory-based monitoring of PDP and other pathogens, especially in low-resource settings, in response to seasonal fluxes and can provide insights into the potential public health risks of emerging pathogens and TSS pollution (e.g., nanoparticles and micro- and nanoplastics) in the aquatic milieu. The model outputs provide low-cost and effective early warning information to assist watershed managers and fish farmers in making appropriate decisions about water resource protection, aquaculture management, and sustainable public health protection.

Artificial intelligence in the diagnosis of necrotising enterocolitis in newborns

Necrotising enterocolitis (NEC) is one of the most common diseases in neonates and predominantly affects premature or very-low-birth-weight infants. Diagnosis is difficult and needed in hours since the first symptom onset for the best therapeutic effects. Artificial intelligence (AI) may play a significant role in NEC diagnosis. A literature search on the use of AI in the diagnosis of NEC was performed. Four databases (PubMed, Embase, arXiv, and IEEE Xplore) were searched with the appropriate MeSH terms. The search yielded 118 publications that were reduced to 8 after screening and checking for eligibility. Of the eight, five used classic machine learning (ML), and three were on the topic of deep ML. Most publications showed promising results. However, no publications with evident clinical benefits were found. Datasets used for training and testing AI systems were small and typically came from a single institution. The potential of AI to improve the diagnosis of NEC is evident. The body of literature on this topic is scarce, and more research in this area is needed, especially with a focus on clinical utility. Cross-institutional data for the training and testing of AI algorithms are required to make progress in this area. IMPACT: Only a few publications on the use of AI in NEC diagnosis are available although they offer some evidence that AI may be helpful in NEC diagnosis. AI requires large, multicentre, and multimodal datasets of high quality for model training and testing. Published results in the literature are based on data from single institutions and, as such, have limited generalisability. Large multicentre studies evaluating broad datasets are needed to evaluate the true potential of AI in diagnosing NEC in a clinical setting.

Development of artificial neural networks for early prediction of intestinal perforation in preterm infants

Intestinal perforation (IP) in preterm infants is a life-threatening condition that may result in serious complications and increased mortality. Early Prediction of IP in infants is important, but challenging due to its multifactorial and complex nature of the disease. Thus, there are no reliable tools to predict IP in infants. In this study, we developed new machine learning (ML) models for predicting IP in very low birth weight (VLBW) infants and compared their performance to that of classic ML methods. We developed artificial neural networks (ANNs) using VLBW infant data from a nationwide cohort and prospective web-based registry. The new ANN models, which outperformed all other classic ML methods, showed an area under the receiver operating characteristic curve (AUROC) of 0.8832 for predicting IP associated with necrotizing enterocolitis (NEC-IP) and 0.8797 for spontaneous IP (SIP). We tested these algorithms using patient data from our institution, which were not included in the training dataset, and obtained an AUROC of 1.0000 for NEC-IP and 0.9364 for SIP. NEC-IP and SIP in VLBW infants can be predicted at an excellent performance level with these newly developed ML models. https://github.com/kdhRick2222/Early-Prediction-of-Intestinal-Perforation-in-Preterm-Infants.