Early prediction of pediatric asthma in the Canadian Healthy Infant Longitudinal Development (CHILD) birth cohort using machine learning.

BACKGROUND: Early identification of children at risk of asthma can have significant clinical implications for effective intervention and treatment. This study aims to disentangle the relative timing and importance of early markers of asthma. METHODS: Using the CHILD Cohort Study, 132 variables measured in 1754 multi-ethnic children were included in the analysis for asthma prediction. Data up to 4 years of age was used in multiple machine learning models to predict physician-diagnosed asthma at age 5 years. Both predictive performance and variable importance was assessed in these models. RESULTS: Early-life data (≤1 year) has limited predictive ability for physician-diagnosed asthma at age 5 years (area under the precision-recall curve (AUPRC) < 0.35). The earliest reliable prediction of asthma is achieved at age 3 years, (area under the receiver-operator curve (AUROC) > 0.90) and (AUPRC > 0.80). Maternal asthma, antibiotic exposure, and lower respiratory tract infections remained highly predictive throughout childhood. Wheezing status and atopy are the most important predictors of early childhood asthma from among the factors included in this study. CONCLUSIONS: Childhood asthma is predictable from non-biological measurements from the age of 3 years, primarily using parental asthma and patient history of wheezing, atopy, antibiotic exposure, and lower respiratory tract infections. IMPACT: Machine learning models can predict physician-diagnosed asthma in early childhood (AUROC > 0.90 and AUPRC > 0.80) using ≥3 years of non-biological and non-genetic information, whereas prediction with the same patient information available before 1 year of age is challenging. Wheezing, atopy, antibiotic exposure, lower respiratory tract infections, and the child's mother having asthma were the strongest early markers of 5-year asthma diagnosis, suggesting an opportunity for earlier diagnosis and intervention and focused assessment of patients at risk for asthma, with an evolving risk stratification over time.

Predicting obstruction risk using common ultrasonography parameters in paediatric hydronephrosis with machine learning.

OBJECTIVE: To sensitively predict the risk of renal obstruction on diuretic renography using routine reported ultrasonography (US) findings, coupled with machine learning approaches, and determine safe criteria for deferral of diuretic renography. PATIENTS AND METHODS: Patients from two institutions with isolated hydronephrosis who underwent a diuretic renogram within 3 months following renal US were included. Age, sex, and routinely reported US findings (laterality, kidney length, anteroposterior diameter, Society for Fetal Urology [SFU] grade) were abstracted. The drainage half-times were collected from renography and stratified as low risk (<20 min, primary outcome), intermediate risk (20-60 min), and high risk of obstruction (>60 min). A random Forest model was trained to classify obstruction risk, here named the 'Artificial intelligence Evaluation of Renogram Obstruction' (AERO). Model performance was determined by measuring area under the receiver-operating-characteristic curve (AUROC) and decision curve analysis. RESULTS: A total of 304 patients met the inclusion criteria, with a median (interquartile range) age of diuretic renogram at 4 (2-7) months. Of all patients, 48 (16%) were low risk, 102 (33%) were intermediate risk, 156 (51%) were high risk of obstruction based on diuretic renogram. The AERO achieved a binary AUROC of 0.84, multi-class AUROC of 0.74 that was superior to the SFU grade, and external validation (n = 64) binary AUROC of 0.76. The most important features for prediction included age, anteroposterior diameter, and SFU grade. We deployed our application in an easy-to-use application (https://sickkidsurology.shinyapps.io/AERO/). At a threshold probability of 30%, the AERO would allow 66 more patients per 1000 to safely avoid a renogram without missing significant obstruction compared to a strategy in which a renogram is routinely performed for SFU Grade ≥3. CONCLUSIONS: Coupled with machine learning, routine US findings can improve the criteria to determine in which children with isolated hydronephrosis a diuretic renogram can be safely avoided. Further optimisation and validation are required prior to implementation into clinical practice.

Deep-learning computer vision can identify increased nuchal translucency in the first trimester of pregnancy.

OBJECTIVE: Many fetal anomalies can already be diagnosed by ultrasound in the first trimester of pregnancy. Unfortunately, in clinical practice, detection rates for anomalies in early pregnancy remain low. Our aim was to use an automated image segmentation algorithm to detect one of the most common fetal anomalies: a thickened nuchal translucency (NT), which is a marker for genetic and structural anomalies. METHODS: Standardized mid-sagittal ultrasound images of the fetal head and chest were collected for 560 fetuses between 11 and 13 weeks and 6 days of gestation, 88 (15.7%) of whom had an NT thicker than 3.5 mm. Image quality was graded as high or low by two fetal medicine experts. Images were divided into a training-set (n = 451, 55 thick NT) and a test-set (n = 109, 33 thick NT). We then trained a U-Net convolutional neural network to segment the fetus and the NT region and computed the NT:fetus ratio of these regions. The ability of this ratio to separate thick (anomalous) NT regions from healthy, typical NT regions was first evaluated in ground-truth segmentation to validate the metric and then with predicted segmentation to validate our algorithm, both using the area under the receiver operator curve (AUROC). RESULTS: The ground-truth NT:fetus ratio detected thick NTs with 0.97 AUROC in both the training and test sets. The fetus and NT regions were detected with a Dice score of 0.94 in the test set. The NT:fetus ratio based on model segmentation detected thick NTs with an AUROC of 0.96 relative to clinician labels. At a 91% specificity, 94% of thick NT cases were detected (sensitivity) in the test set. The detection rate was statistically higher (p = 0.003) in high versus low-quality images (AUROC 0.98 vs. 0.90, respectively). CONCLUSION: Our model provides an explainable deep-learning method for detecting increased NT. This technique can be used to screen for other fetal anomalies in the first trimester of pregnancy.

Machine learning-based mortality prediction models using national liver transplantation registries are feasible but have limited utility across countries.

Many countries curate national registries of liver transplant (LT) data. These registries are often used to generate predictive models; however, potential performance and transferability of these models remain unclear. We used data from 3 national registries and developed machine learning algorithm (MLA)-based models to predict 90-day post-LT mortality within and across countries. Predictive performance and external validity of each model were assessed. Prospectively collected data of adult patients (aged ≥18 years) who underwent primary LTs between January 2008 and December 2018 from the Canadian Organ Replacement Registry (Canada), National Health Service Blood and Transplantation (United Kingdom), and United Network for Organ Sharing (United States) were used to develop MLA models to predict 90-day post-LT mortality. Models were developed using each registry individually (based on variables inherent to the individual databases) and using all 3 registries combined (variables in common between the registries [harmonized]). The model performance was evaluated using area under the receiver operating characteristic (AUROC) curve. The number of patients included was as follows: Canada, n = 1214; the United Kingdom, n = 5287; and the United States, n = 59,558. The best performing MLA-based model was ridge regression across both individual registries and harmonized data sets. Model performance diminished from individualized to the harmonized registries, especially in Canada (individualized ridge: AUROC, 0.74; range, 0.73-0.74; harmonized: AUROC, 0.68; range, 0.50-0.73) and US (individualized ridge: AUROC, 0.71; range, 0.70-0.71; harmonized: AUROC, 0.66; range, 0.66-0.66) data sets. External model performance across countries was poor overall. MLA-based models yield a fair discriminatory potential when used within individual databases. However, the external validity of these models is poor when applied across countries. Standardization of registry-based variables could facilitate the added value of MLA-based models in informing decision making in future LTs.

Dynamic risk profiling of HCC recurrence after curative intent liver resection.

BACKGROUND AND AIMS: Following liver resection (LR) for HCC, the likelihood of survival is dynamic, in that multiple recurrences and/or metastases are possible, each having variable impacts on outcomes. We sought to evaluate the natural progression, pattern, and timing of various disease states after LR for HCC using multistate modeling and to create a practical calculator to provide prognostic information for patients and clinicians. APPROACH AND RESULTS: Adult patients undergoing LR for HCC between January 2000 and December 2018 were retrospectively identified at a single center. Multistate analysis modeled post-LR tumor progression by describing transitions between distinct disease states. In this model, the states included surgery, intrahepatic recurrence (first, second, third, fourth, fifth), distant metastasis with or without intrahepatic recurrence, and death. Of the 486 patients included, 169 (34.8%) remained recurrence-free, 205 (42.2%) developed intrahepatic recurrence, 80 (16.5%) developed distant metastasis, and 32 (7%) died. For an average patient having undergone LR, there was a 33.1% chance of remaining disease-free, a 31.0% chance of at least one intrahepatic recurrence, a 16.3% chance of distant metastasis, and a 19.8% chance of death within the first 60 months post-LR. The transition probability from surgery to first intrahepatic recurrence, without a subsequent state transition, increased from 3% (3 months) to 17.4% (30 months) and 17.2% (60 months). Factors that could modify these probabilities included tumor size, satellite lesions, and microvascular invasion. The online multistate model calculator can be found on https://multistatehcc.shinyapps.io/home/. CONCLUSIONS: In contrast to standard single time-to-event estimates, multistate modeling provides more realistic prognostication of outcomes after LR for HCC by taking into account many postoperative disease states and transitions between them. Our multistate modeling calculator can provide meaningful data to guide the management of patients undergoing postoperative surveillance and therapy.

Systemic lupus erythematosus phenotypes formed from machine learning with a specific focus on cognitive impairment.

OBJECTIVE: To phenotype SLE based on symptom burden (disease damage, system involvement and patient reported outcomes), with a specific focus on objective and subjective cognitive function. METHODS: SLE patients ages 18-65 years underwent objective cognitive assessment using the ACR Neuropsychological Battery (ACR-NB) and data were collected on demographic and clinical variables, disease burden/activity, health-related quality of life (HRQoL), depression, anxiety, fatigue and perceived cognitive deficits. Similarity network fusion (SNF) was used to identify patient subtypes. Differences between the subtypes were evaluated using Kruskal-Wallis and χ2 tests. RESULTS: Of the 238 patients, 90% were female, with a mean age of 41 years (s.d. 12) and a disease duration of 14 years (s.d. 10) at the study visit. The SNF analysis defined two subtypes (A and B) with distinct patterns in objective and subjective cognitive function, disease burden/damage, HRQoL, anxiety and depression. Subtype A performed worst on all significantly different tests of objective cognitive function (P < 0.03) compared with subtype B. Subtype A also had greater levels of subjective cognitive function (P < 0.001), disease burden/damage (P < 0.04), HRQoL (P < 0.001) and psychiatric measures (P < 0.001) compared with subtype B. CONCLUSION: This study demonstrates the complexity of cognitive impairment (CI) in SLE and that individual, multifactorial phenotypes exist. Those with greater disease burden, from SLE-specific factors or other factors associated with chronic conditions, report poorer cognitive functioning and perform worse on objective cognitive measures. By exploring different ways of phenotyping SLE we may better define CI in SLE. Ultimately this will aid our understanding of personalized CI trajectories and identification of appropriate treatments.

Using a machine learning algorithm to predict outcome of primary cytoreductive surgery in advanced ovarian cancer.

OBJECTIVE: To develop a machine learning (ML) algorithm to predict outcome of primary cytoreductive surgery (PCS) in patients with advanced ovarian cancer (AOC) METHODS: This retrospective cohort study included patients with AOC undergoing PCS between January 2017 and February 2021. Using radiologic criteria, patient factors (age, CA-125, performance status, BRCA) and surgical complexity scores, we trained a random forest model to predict the dichotomous outcome of optimal cytoreduction (<1 cm) and no gross residual (RD = 0 mm) using JMP-Pro 15 (SAS). This model is available at https://ipm-ml.ccm.sickkids.ca. RESULTS: One hundred and fifty-one patients underwent PCS and randomly assigned to train (n = 92), validate (n = 30), or test (n = 29) the model. The median age was 58 (27-83). Patients with suboptimal cytoreduction were more likely to have an Eastern Cooperative Oncology Group 3-4 (11% vs. 0.75%, p = 0.004), lower albumin (38 vs. 41, p = 0.02), and higher CA125 (1126 vs. 388, p = 0.012) than patients with optimal cytoreduction (n = 133). There were no significant differences in age, histology, stage, or BRCA status between groups. The bootstrap random forest model had AUCs of 99.8% (training), 89.6%(validation), and 89.0% (test). The top five contributors were CA125, albumin, diaphragmatic disease, age, and ascites. For RD = 0 mm, the AUCs were 94.4%, 52%, and 84%, respectively. CONCLUSION: Our ML algorithm demonstrated high accuracy in predicting optimal cytoreduction in patients with AOC selected for PCS and may assist decision-making.

Deep learning imaging features derived from kidney ultrasounds predict chronic kidney disease progression in children with posterior urethral valves.

BACKGROUND: We sought to use deep learning to extract anatomic features from postnatal kidney ultrasounds and evaluate their performance in predicting the risk and timing of chronic kidney disease (CKD) progression for boys with posterior urethral valves (PUV). We hypothesized that these features would predict CKD progression better than clinical characteristics such as nadir creatinine alone. METHODS: We performed a retrospective cohort study of boys with PUV treated at two pediatric health systems from 1990 to 2021. Features of kidneys were extracted from initial postnatal kidney ultrasound images using a deep learning model. Three time-to-event prediction models were built using random survival forests. The Imaging Model included deep learning imaging features, the Clinical Model included clinical data, and the Ensemble Model combined imaging features and clinical data. Separate models were built to include time-dependent clinical data that were available at 6 months, 1 year, 3 years, and 5 years. RESULTS: Two-hundred and twenty-five patients were included in the analysis. All models performed well with C-indices of 0.7 or greater. The Clinical Model outperformed the Imaging Model at all time points with nadir creatinine driving the performance of the Clinical Model. Combining the 6-month Imaging Model (C-index 0.7; 95% confidence interval [CI] 0.6, 0.79) with the 6-month Clinical Model (C-index 0.79; 95% CI 0.71, 0.86) resulted in a 6-month Ensemble Model that performed better (C-index 0.82; 95% CI 0.77, 0.88) than either model alone. CONCLUSIONS: Deep learning imaging features extracted from initial postnatal kidney ultrasounds may improve early prediction of CKD progression among children with PUV. A higher resolution version of the Graphical abstract is available as Supplementary information.

Large scale genotype- and phenotype-driven machine learning in Von Hippel-Lindau disease.

Von Hippel-Lindau (VHL) disease is a hereditary cancer syndrome where individuals are predisposed to tumor development in the brain, adrenal gland, kidney, and other organs. It is caused by pathogenic variants in the VHL tumor suppressor gene. Standardized disease information has been difficult to collect due to the rarity and diversity of VHL patients. Over 4100 unique articles published until October 2019 were screened for germline genotype-phenotype data. Patient data were translated into standardized descriptions using Human Genome Variation Society gene variant nomenclature and Human Phenotype Ontology terms and has been manually curated into an open-access knowledgebase called Clinical Interpretation of Variants in Cancer. In total, 634 unique VHL variants, 2882 patients, and 1991 families from 427 papers were captured. We identified relationship trends between phenotype and genotype data using classic statistical methods and spectral clustering unsupervised learning. Our analyses reveal earlier onset of pheochromocytoma/paraganglioma and retinal angiomas, phenotype co-occurrences and genotype-phenotype correlations including hotspots. It confirms existing VHL associations and can be used to identify new patterns and associations in VHL disease. Our database serves as an aggregate knowledge translation tool to facilitate sharing information about the pathogenicity of VHL variants.

The Toronto Postliver Transplantation Hepatocellular Carcinoma Recurrence Calculator: A Machine Learning Approach.

Liver transplantation (LT) listing criteria for hepatocellular carcinoma (HCC) remain controversial. To optimize the utility of limited donor organs, this study aims to leverage machine learning to develop an accurate posttransplantation HCC recurrence prediction calculator. Patients with HCC listed for LT from 2000 to 2016 were identified, with 739 patients who underwent LT used for modeling. Data included serial imaging, alpha-fetoprotein (AFP), locoregional therapies, treatment response, and posttransplantation outcomes. We compared the CoxNet (regularized Cox regression), survival random forest, survival support vector machine, and DeepSurv machine learning algorithms via the mean cross-validated concordance index. We validated the selected CoxNet model by comparing it with other currently available recurrence risk algorithms on a held-out test set (AFP, Model of Recurrence After Liver Transplant [MORAL], and Hazard Associated with liver Transplantation for Hepatocellular Carcinoma [HALT-HCC score]). The developed CoxNet-based recurrence prediction model showed a satisfying overall concordance score of 0.75 (95% confidence interval [CI], 0.64-0.84). In comparison, the recalibrated risk algorithms' concordance scores were as follows: AFP score 0.64 (outperformed by the CoxNet model, 1-sided 95% CI, >0.01; P = 0.04) and MORAL score 0.64 (outperformed by the CoxNet model 1-sided 95% CI, >0.02; P = 0.03). The recalibrated HALT-HCC score performed well with a concordance of 0.72 (95% CI, 0.63-0.81) and was not significantly outperformed (1-sided 95% CI, ≥0.05; P = 0.29). Developing a comprehensive posttransplantation HCC recurrence risk calculator using machine learning is feasible and can yield higher accuracy than other available risk scores. Further research is needed to confirm the utility of machine learning in this setting.

Multi-institutional Validation of Improved Vesicoureteral Reflux Assessment With Simple and Machine Learning Approaches.

PURPOSE: Vesicoureteral reflux grading from voiding cystourethrograms is highly subjective with low reliability. We aimed to demonstrate improved reliability for vesicoureteral reflux grading with simple and machine learning approaches using ureteral tortuosity and dilatation on voiding cystourethrograms. MATERIALS AND METHODS: Voiding cystourethrograms were collected from our institution for training and 5 external data sets for validation. Each voiding cystourethrogram was graded by 5-7 raters to determine a consensus vesicoureteral reflux grade label and inter- and intra-rater reliability was assessed. Each voiding cystourethrogram was assessed for 4 features: ureteral tortuosity, proximal, distal, and maximum ureteral dilatation. The labels were then assigned to the combination of the 4 features. A machine learning-based model, qVUR, was trained to predict vesicoureteral reflux grade from these features and model performance was assessed by AUROC (area under the receiver-operator-characteristic). RESULTS: A total of 1,492 kidneys and ureters were collected from voiding cystourethrograms resulting in a total of 8,230 independent gradings. The internal inter-rater reliability for vesicoureteral reflux grading was 0.44 with a median percent agreement of 0.71 and low intra-rater reliability. Higher values for each feature were associated with higher vesicoureteral reflux grade. qVUR performed with an accuracy of 0.62 (AUROC=0.84) with stable performance across all external data sets. The model improved vesicoureteral reflux grade reliability by 3.6-fold compared to traditional grading (P < .001). CONCLUSIONS: In a large pediatric population from multiple institutions, we show that machine learning-based assessment for vesicoureteral reflux improves reliability compared to current grading methods. qVUR is generalizable and robust with similar accuracy to clinicians but the added prognostic value of quantitative measures warrants further study.

Machine learning classification of multiple sclerosis in children using optical coherence tomography.

BACKGROUND: In children, multiple sclerosis (MS) is the ultimate diagnosis in only 1/5 to 1/3 of cases after a first episode of central nervous system (CNS) demyelination. As the visual pathway is frequently affected in MS and other CNS demyelinating disorders (DDs), structural retinal imaging such as optical coherence tomography (OCT) can be used to differentiate MS. OBJECTIVE: This study aimed to investigate the utility of machine learning (ML) based on OCT features to identify distinct structural retinal features in children with DDs. METHODS: This study included 512 eyes from 187 (neyes = 374) children with demyelinating diseases and 69 (neyes = 138) controls. Input features of the analysis comprised of 24 auto-segmented OCT features. RESULTS: Random Forest classifier with recursive feature elimination yielded the highest predictive values and identified DDs with 75% and MS with 80% accuracy, while multiclass distinction between MS and monophasic DD was performed with 64% accuracy. A set of eight retinal features were identified as the most important features in this classification. CONCLUSION: This study demonstrates that ML based on OCT features can be used to support a diagnosis of MS in children.

Pre-versus postnatal presentation of posterior urethral valves: a multi-institutional experience.

OBJECTIVE: To compare the outcomes of pre- vs postnatally diagnosed posterior urethral valves (PUV) at two large paediatric centres in North America to ascertain if the prenatal diagnosis of PUV is associated with better outcomes. PATIENTS AND METHODS: All boys with PUV were identified at two large paediatric institutions in North America between 2000 and 2020 (The Hospital for Sick Children [SickKids, SK] and Children's Hospital of Philadelphia [CHOP]). Baseline characteristics and outcome measures were compared between those diagnosed pre- vs postnatally. Main outcomes of interest included progression of chronic kidney disease (CKD), the need for renal replacement therapy (RRT), and bladder function compromise, as determined by need for clean intermittent catheterisation (CIC). Time-to-event analyses were completed when possible. RESULTS: During the study period, 152 boys with PUV were treated at the SK (39% prenatal) and 216 were treated at the CHOP (71% prenatal). At the SK, there was no difference between the pre- and postnatal groups in the proportion of boys who required RRT, progressed to CKD Stage ≥3, or who were managed with CIC when comparing the timing of diagnosis. The time to event for RRT and CIC was significantly younger for prenatally detected PUV. At the CHOP, significantly more prenatal boys required RRT; however, there was no significant difference in the age this outcome was reached. The proportion of boys managed with CIC was not different but the time to event was significantly earlier in the prenatal group. CONCLUSION: This study represents the largest multi-institutional series of boys with PUV and failed to identify any difference in the outcomes of pre- vs postnatal detection of PUV. A multidisciplinary approach with standardisation of the treatment pathways will help in understanding the true impact of prenatal/early detection on outcomes of PUV.

Personalized application of machine learning algorithms to identify pediatric patients at risk for recurrent ureteropelvic junction obstruction after dismembered pyeloplasty.

PURPOSE: To develop a model that predicts whether a child will develop a recurrent obstruction after pyeloplasty, determine their survival risk score, and expected time to re-intervention using machine learning (ML). METHODS: We reviewed patients undergoing pyeloplasty from 2008 to 2020 at our institution, including all children and adolescents younger than 18 years. We developed a two-stage machine learning model from 34 clinical fields, which included patient characteristics, ultrasound findings, and anatomical variation. We fit and trained with a logistic lasso model for binary cure model and subsequent survival model. Feature importance on the model was determined with post-selection inference. Performance metrics included area under the receiver-operating-characteristic (AUROC), concordance, and leave-one-out cross validation. RESULTS: A total of 543 patients were identified, with a median preoperative and postoperative anteroposterior diameter of 23 and 10 mm, respectively. 39 of 232 patients included in the survival model required re-intervention. The cure and survival models performed well with a leave-one-out cross validation AUROC and concordance of 0.86 and 0.78, respectively. Post-selective inference showed that larger anteroposterior diameter at the second post-op follow-up, and anatomical variation in the form of concurrent anomalies were significant model features predicting negative outcomes. The model can be used at https://sickkidsurology.shinyapps.io/PyeloplastyReOpRisk/ . CONCLUSION: Our ML-based model performed well in predicting the risk of and time to re-intervention after pyeloplasty. The implementation of this ML-based approach is novel in pediatric urology and will likely help achieve personalized risk stratification for patients undergoing pyeloplasty. Further real-world validation is warranted.

Posterior Urethral Valves Outcomes Prediction (PUVOP): a machine learning tool to predict clinically relevant outcomes in boys with posterior urethral valves.

BACKGROUND: Early kidney and anatomic features may be predictive of future progression and need for additional procedures in patients with posterior urethral valve (PUV). The objective of this study was to use machine learning (ML) to predict clinically relevant outcomes in these patients. METHODS: Patients diagnosed with PUV with kidney function measurements at our institution between 2000 and 2020 were included. Pertinent clinical measures were abstracted, including estimated glomerular filtration rate (eGFR) at each visit, initial vesicoureteral reflux grade, and renal dysplasia at presentation. ML models were developed to predict clinically relevant outcomes: progression in CKD stage, initiation of kidney replacement therapy (KRT), and need for clean-intermittent catheterization (CIC). Model performance was assessed by concordance index (c-index) and the model was externally validated. RESULTS: A total of 103 patients were included with a median follow-up of 5.7 years. Of these patients, 26 (25%) had CKD progression, 18 (17%) required KRT, and 32 (31%) were prescribed CIC. Additionally, 22 patients were included for external validation. The ML model predicted CKD progression (c-index = 0.77; external C-index = 0.78), KRT (c-index = 0.95; external C-index = 0.89) and indicated CIC (c-index = 0.70; external C-index = 0.64), and all performed better than Cox proportional-hazards regression. The models have been packaged into a simple easy-to-use tool, available at https://share.streamlit.io/jcckwong/puvop/main/app.py CONCLUSION: ML-based approaches for predicting clinically relevant outcomes in PUV are feasible. Further validation is warranted, but this implementable model can act as a decision-making aid. A higher resolution version of the Graphical abstract is available as Supplementary information.

A Research Ethics Framework for the Clinical Translation of Healthcare Machine Learning.

The application of artificial intelligence and machine learning (ML) technologies in healthcare have immense potential to improve the care of patients. While there are some emerging practices surrounding responsible ML as well as regulatory frameworks, the traditional role of research ethics oversight has been relatively unexplored regarding its relevance for clinical ML. In this paper, we provide a comprehensive research ethics framework that can apply to the systematic inquiry of ML research across its development cycle. The pathway consists of three stages: (1) exploratory, hypothesis-generating data access; (2) silent period evaluation; (3) prospective clinical evaluation. We connect each stage to its literature and ethical justification and suggest adaptations to traditional paradigms to suit ML while maintaining ethical rigor and the protection of individuals. This pathway can accommodate a multitude of research designs from observational to controlled trials, and the stages can apply individually to a variety of ML applications.

Retraining an Artificial Intelligence Algorithm to Calculate Left Ventricular Ejection Fraction in Pediatrics.

OBJECTIVES: Identifying patients with low left ventricular ejection fraction (LVEF) and monitoring LVEF responses to treatment are important clinical goals. Can a deep-learning algorithm predict pediatric LVEF within clinically acceptable error? DESIGN: The study authors wanted to fine-tune an adult deep-learning algorithm to calculate LVEF in pediatric patients. A priori, their objective was to refine the algorithm to perform LVEF calculation with a mean absolute error (MAE) ≤5%. SETTING: A quaternary pediatric hospital PARTICIPANTS: A convenience sample (n = 321) of echocardiograms from newborns to 18 years old with normal cardiac anatomy or hemodynamically insignificant anomalies. Echocardiograms were chosen from a group of healthy controls with known normal LVEF (n = 267) and a dilated cardiomyopathy patient group with reduced LVEF (n = 54). INTERVENTIONS: The artificial intelligence model EchoNet-Dynamic was tested on this data set and then retrained, tested, and further validated to improve LVEF calculation. The gold standard value was LVEF calculated by clinical experts. MEASUREMENTS AND MAIN RESULTS: In a random subset of subjects (n = 40) not analyzed prior to selection of the final model, EchoNet-Dynamic calculated LVEF with a MAE of 8.39%, R2 = 0.47 without, and MAE 4.47%, R2 = 0.87 with fine-tuning. Bland-Altman analysis suggested that the model slightly underestimates LVEF (bias = -2.42%). The 95% limits of agreement between actual and calculated values were -12.32% to 7.47%. CONCLUSIONS: The fine-tuned model calculates LVEF in a range of pediatric patients within clinically acceptable error. Potential advantages include reducing operator error in LVEF calculation and supporting independent LVEF assessment by inexperienced users.

Data governance functions to support responsible data stewardship in pediatric radiology research studies using artificial intelligence.

The integration of human and machine intelligence promises to profoundly change the practice of medicine. The rapidly increasing adoption of artificial intelligence (AI) solutions highlights its potential to streamline physician work and optimize clinical decision-making, also in the field of pediatric radiology. Large imaging databases are necessary for training, validating and testing these algorithms. To better promote data accessibility in multi-institutional AI-enabled radiologic research, these databases centralize the large volumes of data required to effect accurate models and outcome predictions. However, such undertakings must consider the sensitivity of patient information and therefore utilize requisite data governance measures to safeguard data privacy and security, to recognize and mitigate the effects of bias and to promote ethical use. In this article we define data stewardship and data governance, review their key considerations and applicability to radiologic research in the pediatric context, and consider the associated best practices along with the ramifications of poorly executed data governance. We summarize several adaptable data governance frameworks and describe strategies for their implementation in the form of distributed and centralized approaches to data management.

Predicting the risk of mortality during hospitalization in sick severely malnourished children using daily evaluation of key clinical warning signs.

BACKGROUND: Despite adherence to WHO guidelines, inpatient mortality among sick children admitted to hospital with complicated severe acute malnutrition (SAM) remains unacceptably high. Several studies have examined risk factors present at admission for mortality. However, risks may evolve during admission with medical and nutritional treatment or deterioration. Currently, no specific guidance exists for assessing daily treatment response. This study aimed to determine the prognostic value of monitoring clinical signs on a daily basis for assessing mortality risk during hospitalization in children with SAM. METHODS: This is a secondary analysis of data from a randomized trial (NCT02246296) among 843 hospitalized children with SAM. Daily clinical signs were prospectively collected during ward rounds. Multivariable extended Cox regression using backward feature selection was performed to identify daily clinical warning signs (CWS) associated with time to death within the first 21 days of hospitalization. Predictive models were subsequently developed, and their prognostic performance evaluated using Harrell's concordance index (C-index) and time-dependent area under the curve (tAUC). RESULTS: Inpatient case fatality ratio was 16.3% (n=127). The presence of the following CWS during daily assessment were found to be independent predictors of inpatient mortality: symptomatic hypoglycemia, reduced consciousness, chest indrawing, not able to complete feeds, nutritional edema, diarrhea, and fever. Daily risk scores computed using these 7 CWS together with MUAC<10.5cm at admission as additional CWS predict survival outcome of children with SAM with a C-index of 0.81 (95% CI 0.77-0.86). Moreover, counting signs among the top 5 CWS (reduced consciousness, symptomatic hypoglycemia, chest indrawing, not able to complete foods, and MUAC<10.5cm) provided a simpler tool with similar prognostic performance (C-index of 0.79; 95% CI 0.74-0.84). Having 1 or 2 of these CWS on any day during hospitalization was associated with a 3 or 11-fold increased mortality risk compared with no signs, respectively. CONCLUSIONS: This study provides evidence for structured monitoring of daily CWS as recommended clinical practice as it improves prediction of inpatient mortality among sick children with complicated SAM. We propose a simple counting-tool to guide healthcare workers to assess treatment response for these children. TRIAL REGISTRATION: NCT02246296.

Examination of the shared genetic basis of anorexia nervosa and obsessive-compulsive disorder.

Anorexia nervosa (AN) and obsessive-compulsive disorder (OCD) are often comorbid and likely to share genetic risk factors. Hence, we examine their shared genetic background using a cross-disorder GWAS meta-analysis of 3495 AN cases, 2688 OCD cases, and 18,013 controls. We confirmed a high genetic correlation between AN and OCD (rg = 0.49 ± 0.13, p = 9.07 × 10-7) and a sizable SNP heritability (SNP h2 = 0.21 ± 0.02) for the cross-disorder phenotype. Although no individual loci reached genome-wide significance, the cross-disorder phenotype showed strong positive genetic correlations with other psychiatric phenotypes (e.g., rg = 0.36 with bipolar disorder and 0.34 with neuroticism) and negative genetic correlations with metabolic phenotypes (e.g., rg = -0.25 with body mass index and -0.20 with triglycerides). Follow-up analyses revealed that although AN and OCD overlap heavily in their shared risk with other psychiatric phenotypes, the relationship with metabolic and anthropometric traits is markedly stronger for AN than for OCD. We further tested whether shared genetic risk for AN/OCD was associated with particular tissue or cell-type gene expression patterns and found that the basal ganglia and medium spiny neurons were most enriched for AN-OCD risk, consistent with neurobiological findings for both disorders. Our results confirm and extend genetic epidemiological findings of shared risk between AN and OCD and suggest that larger GWASs are warranted.