Machine learning guided postnatal gestational age assessment using new-born screening metabolomic data in South Asia and sub-Saharan Africa

Current Status of Newborn Bloodspot Screening Worldwide 2024: A Comprehensive Review of Recent Activities (2020-2023)

Newborn bloodspot screening (NBS) began in the early 1960s based on the work of Dr. Robert "Bob" Guthrie in Buffalo, NY, USA. His development of a screening test for phenylketonuria on blood absorbed onto a special filter paper and transported to a remote testing laboratory began it all. Expansion of NBS to large numbers of asymptomatic congenital conditions flourishes in many settings while it has not yet been realized in others. The need for NBS as an efficient and effective public health prevention strategy that contributes to lowered morbidity and mortality wherever it is sustained is well known in the medical field but not necessarily by political policy makers. Acknowledging the value of national NBS reports published in 2007, the authors collaborated to create a worldwide NBS update in 2015. In a continuing attempt to review the progress of NBS globally, and to move towards a more harmonized and equitable screening system, we have updated our 2015 report with information available at the beginning of 2024. Reports on sub-Saharan Africa and the Caribbean, missing in 2015, have been included. Tables popular in the previous report have been updated with an eye towards harmonized comparisons. To emphasize areas needing attention globally, we have used regional tables containing similar listings of conditions screened, numbers of screening laboratories, and time at which specimen collection is recommended. Discussions are limited to bloodspot screening.

Untargeted metabolomic profiling reveals molecular signatures associated with type 2 diabetes in Nigerians

BACKGROUND

Type 2 diabetes (T2D) has reached epidemic proportions globally, including in Africa. However, molecular studies to understand the pathophysiology of T2D remain scarce outside Europe and North America. The aims of this study are to use an untargeted metabolomics approach to identify: (a) metabolites that are differentially expressed between individuals with and without T2D and (b) a metabolic signature associated with T2D in a population of Sub-Saharan Africa (SSA).

METHODS

A total of 580 adult Nigerians from the Africa America Diabetes Mellitus (AADM) study were studied. The discovery study included 310 individuals (210 without T2D, 100 with T2D). Metabolites in plasma were assessed by reverse phase, ultra-performance liquid chromatography and mass spectrometry (RP)/UPLC-MS/MS methods on the Metabolon Platform. Welch's two-sample t-test was used to identify differentially expressed metabolites (DEMs), followed by the construction of a biomarker panel using a random forest (RF) algorithm. The biomarker panel was evaluated in a replication sample of 270 individuals (110 without T2D and 160 with T2D) from the same study.

RESULTS

Untargeted metabolomic analyses revealed 280 DEMs between individuals with and without T2D. The DEMs predominantly belonged to the lipid (51%, 142/280), amino acid (21%, 59/280), xenobiotics (13%, 35/280), carbohydrate (4%, 10/280) and nucleotide (4%, 10/280) super pathways. At the sub-pathway level, glycolysis, free fatty acid, bile metabolism, and branched chain amino acid catabolism were altered in T2D individuals. A 10-metabolite biomarker panel including glucose, gluconate, mannose, mannonate, 1,5-anhydroglucitol, fructose, fructosyl-lysine, 1-carboxylethylleucine, metformin, and methyl-glucopyranoside predicted T2D with an area under the curve (AUC) of 0.924 (95% CI: 0.845-0.966) and a predicted accuracy of 89.3%. The panel was validated with a similar AUC (0.935, 95% CI 0.906-0.958) in the replication cohort. The 10 metabolites in the biomarker panel correlated significantly with several T2D-related glycemic indices, including Hba1C, insulin resistance (HOMA-IR), and diabetes duration.

CONCLUSIONS

We demonstrate that metabolomic dysregulation associated with T2D in Nigerians affects multiple processes, including glycolysis, free fatty acid and bile metabolism, and branched chain amino acid catabolism. Our study replicated previous findings in other populations and identified a metabolic signature that could be used as a biomarker panel of T2D risk and glycemic control thus enhancing our knowledge of molecular pathophysiologic changes in T2D. The metabolomics dataset generated in this study represents an invaluable addition to publicly available multi-omics data on understudied African ancestry populations.

Radiology weather forecast: A retrospective analysis of predictability of median daily polytrauma-CT occurrence based on weather data

OBJECTIVES

Resource planning is a crucial component in hospitals, particularly in radiology departments. Since weather conditions are often described to correlate with emergency room visits, we aimed to forecast the amount of polytrauma-CTs using weather information.

DESIGN

All polytrauma-CTs between 01/01/2011 and 12/31/2022 (n = 6638) were retrieved from the radiology information system. Local weather data was downloaded from meteoblue.com. The data was normalized and smoothened. Daily polytrauma-CT occurrence was stratified into below median and above median number of daily polytrauma-CTs. Logistic regression and machine learning algorithms (neural network, random forest classifier, support vector machine, gradient boosting classifier) were employed as prediction models. Data from 2012 to 2020 was used for training, data from 2021 to 2022 for validation.

RESULTS

More polytrauma-CTs were acquired in summer compared with winter months, demonstrating a seasonal change (median: 2.35; IQR 1.60-3.22 vs. 2.08; IQR 1.36-3.03; p <.001). Temperature (rs = 0.45), sunshine duration (rs = 0.38) and ultraviolet light amount (rs = 0.37) correlated positively, wind velocity (rs = -0.57) and cloudiness (rs = -0.28) correlated negatively with polytrauma-CT occurrence (all p <.001). The logistic regression model for identification of days with above median number of polytrauma-CTs achieved an accuracy of 87 % on training data from 2011 to 2020. When forecasting the years 2021-2022 an accuracy of 65 % was achieved. A neural network and a support vector machine both achieved a validation accuracy of 72 %, whereas all classifiers regarded wind velocity and ultraviolet light amount as the most important parameters.

CONCLUSION

It is possible to forecast above or below median daily number of polytrauma-CTs using weather data.

CLINCICAL RELEVANCE STATEMENT

Prediction of polytrauma-CT examination volumes may be used to improve resource planning.

Small babies, big risks: global estimates of prevalence and mortality for vulnerable newborns to accelerate change and improve counting

Small newborns are vulnerable to mortality and lifelong loss of human capital. Measures of vulnerability previously focused on liveborn low-birthweight (LBW) babies, yet LBW reduction targets are off-track. There are two pathways to LBW, preterm birth and fetal growth restriction (FGR), with the FGR pathway resulting in the baby being small for gestational age (SGA). Data on LBW babies are available from 158 (81%) of 194 WHO member states and the occupied Palestinian territory, including east Jerusalem, with 113 (58%) having national administrative data, whereas data on preterm births are available from 103 (53%) of 195 countries and areas, with only 64 (33%) providing national administrative data. National administrative data on SGA are available for only eight countries. Global estimates for 2020 suggest 13·4 million livebirths were preterm, with rates over the past decade remaining static, and 23·4 million were SGA. In this Series paper, we estimated prevalence in 2020 for three mutually exclusive types of small vulnerable newborns (SVNs; preterm non-SGA, term SGA, and preterm SGA) using individual-level data (2010-20) from 23 national datasets (∼110 million livebirths) and 31 studies in 18 countries (∼0·4 million livebirths). We found 11·9 million (50% credible interval [Crl] 9·1-12·2 million; 8·8%, 50% Crl 6·8-9·0%) of global livebirths were preterm non-SGA, 21·9 million (50% Crl 20·1-25·5 million; 16·3%, 14·9-18·9%) were term SGA, and 1·5 million (50% Crl 1·2-4·2 million; 1·1%, 50% Crl 0·9-3·1%) were preterm SGA. Over half (55·3%) of the 2·4 million neonatal deaths worldwide in 2020 were attributed to one of the SVN types, of which 73·4% were preterm and the remainder were term SGA. Analyses from 12 of the 23 countries with national data (0·6 million stillbirths at ≥22 weeks gestation) showed around 74% of stillbirths were preterm, including 16·0% preterm SGA and approximately one-fifth of term stillbirths were SGA. There are an estimated 1·9 million stillbirths per year associated with similar vulnerability pathways; hence integrating stillbirths to burden assessments and relevant indicators is crucial. Data can be improved by counting, weighing, and assessing the gestational age of every newborn, whether liveborn or stillborn, and classifying small newborns by the three vulnerability types. The use of these more specific types could accelerate prevention and help target care for the most vulnerable babies.

Predicting facility-based delivery in Zanzibar: The vulnerability of machine learning algorithms to adversarial attacks

Background

Community health worker (CHW)-led maternal health programs have contributed to increased facility-based deliveries and decreased maternal mortality in sub-Saharan Africa. The recent adoption of mobile devices in these programs provides an opportunity for real-time implementation of machine learning predictive models to identify women most at risk for home-based delivery. However, it is possible that falsified data could be entered into the model to get a specific prediction result - known as an "adversarial attack". The goal of this paper is to evaluate the algorithm's vulnerability to adversarial attacks.

Methods

The dataset used in this research is from the Uzazi Salama ("Safer Deliveries") program, which operated between 2016 and 2019 in Zanzibar. We used LASSO regularized logistic regression to develop the prediction model. We used "One-At-a-Time (OAT)" adversarial attacks across four different types of input variables: binary - access to electricity at home, categorical - previous delivery location, ordinal - educational level, and continuous - gestational age. We evaluated the percent of predicted classifications that change due to these adversarial attacks.

Results

Manipulating input variables affected prediction results. The variable with the greatest vulnerability was previous delivery location, with 55.65% of predicted classifications changing when applying adversarial attacks from previously delivered at a facility to previously delivered at home, and 37.63% of predicted classifications changing when applying adversarial attacks from previously delivered at home to previously delivered at a facility.

Conclusion

This paper investigates the vulnerability of an algorithm to predict facility-based delivery when facing adversarial attacks. By understanding the effect of adversarial attacks, programs can implement data monitoring strategies to assess for and deter these manipulations. Ensuring fidelity in algorithm deployment secures that CHWs target those women who are actually at high risk of delivering at home.

Pulmonary transit time of cardiovascular magnetic resonance perfusion scans for quantification of cardiopulmonary haemodynamics

AIMS

Pulmonary transit time (PTT) is the time blood takes to pass from the right ventricle to the left ventricle via pulmonary circulation. We aimed to quantify PTT in routine cardiovascular magnetic resonance imaging perfusion sequences. PTT may help in the diagnostic assessment and characterization of patients with unclear dyspnoea or heart failure (HF).

METHODS AND RESULTS

We evaluated routine stress perfusion cardiovascular magnetic resonance scans in 352 patients, including an assessment of PTT. Eighty-six of these patients also had simultaneous quantification of N-terminal pro-brain natriuretic peptide (NTproBNP). NT-proBNP is an established blood biomarker for quantifying ventricular filling pressure in patients with presumed HF. Manually assessed PTT demonstrated low inter-rater variability with a correlation between raters >0.98. PTT was obtained automatically and correctly in 266 patients using artificial intelligence. The median PTT of 182 patients with both left and right ventricular ejection fraction >50% amounted to 6.8 s (Pulmonary transit time: 5.9-7.9 s). PTT was significantly higher in patients with reduced left ventricular ejection fraction (<40%; P < 0.001) and right ventricular ejection fraction (<40%; P < 0.0001). The area under the receiver operating characteristics curve (AUC) of PTT for exclusion of HF (NT-proBNP <125 ng/L) was 0.73 (P < 0.001) with a specificity of 77% and sensitivity of 70%. The AUC of PTT for the inclusion of HF (NT-proBNP >600 ng/L) was 0.70 (P < 0.001) with a specificity of 78% and sensitivity of 61%.

CONCLUSION

PTT as an easily, even automatically obtainable and robust non-invasive biomarker of haemodynamics might help in the evaluation of patients with dyspnoea and HF.

Machine learning for maternal health: Predicting delivery location in a community health worker program in Zanzibar

Background

Maternal and neonatal health outcomes in low- and middle-income countries (LMICs) have improved over the last two decades. However, many pregnant women still deliver at home, which increases the health risks for both the mother and the child. Community health worker programs have been broadly employed in LMICs to connect women to antenatal care and delivery locations. More recently, employment of digital tools in maternal health programs have resulted in better care delivery and served as a routine mode of data collection. Despite the availability of rich, patient-level data within these digital tools, there has been limited utilization of this type of data to inform program delivery in LMICs.

Methods

We use program data from 38,787 women enrolled in Safer Deliveries, a community health worker program in Zanzibar, to build a generalizable prediction model that accurately predicts whether a newly enrolled pregnant woman will deliver in a health facility. We use information collected during the enrollment visit, including demographic data, health characteristics and current pregnancy information. We apply four machine learning methods: logistic regression, LASSO regularized logistic regression, random forest and an artificial neural network; and three sampling techniques to address the imbalanced data: undersampling of facility deliveries, oversampling of home deliveries and addition of synthetic home deliveries using SMOTE.

Results

Our models correctly predicted the delivery location for 68%–77% of the women in the test set, with slightly higher accuracy when predicting facility delivery versus home delivery. A random forest model with a balanced training set created using undersampling of existing facility deliveries accurately identified 74.4% of women delivering at home.

Conclusions

This model can provide a “real-time” prediction of the delivery location for new maternal health program enrollees and may enable early provision of extra support for individuals at risk of not delivering in a health facility, which has potential to improve health outcomes for both mothers and their newborns. The framework presented here is applicable in other contexts and the selection of input features can easily be adapted to match data availability and other outcomes, both within and beyond maternal health.

Machine learning prediction of gestational age from metabolic screening markers resistant to ambient temperature transportation: Facilitating use of this technology in low resource settings of South Asia and East Africa

Background

Knowledge of gestational age is critical for guiding preterm neonatal care. In the last decade, metabolic gestational dating approaches emerged in response to a global health need; because in most of the developing world, accurate antenatal gestational age estimates are not feasible. These methods initially developed in North America have now been externally validated in two studies in developing countries, however, require shipment of samples at sub-zero temperature.

Methods

A subset of 330 pairs of heel prick dried blood spot samples were shipped on dry ice and in ambient temperature from field sites in Tanzania, Bangladesh and Pakistan to laboratory in Iowa (USA). We evaluated impact on recovery of analytes of shipment temperature, developed and evaluated models for predicting gestational age using a limited set of metabolic screening analytes after excluding 17 analytes that were impacted by shipment conditions of a total of 44 analytes.

Results

With the machine learning model using all the analytes, samples shipped in dry ice yielded a Root Mean Square Error (RMSE) of 1.19 weeks compared to 1.58 weeks for samples shipped in ambient temperature. Out of the 44 screening analytes, recovery of 17 analytes was significantly different between the two shipment methods and these were excluded from further machine learning model development. The final model, restricted to stable analytes provided a RMSE of 1.24 (95% confidence interval (CI) = 1.10-1.37) weeks for samples shipped on dry ice and RMSE of 1.28 (95% CI = 1.15-1.39) for samples shipped at ambient temperature. Analysis for discriminating preterm births (gestational age <37 weeks), yielded an area under curve (AUC) of 0.76 (95% CI = 0.71-0.81) for samples shipped on dry ice and AUC of 0.73 (95% CI = 0.67-0.78) for samples shipped in ambient temperature.

Conclusions

In this study, we demonstrate that machine learning algorithms developed using a sub-set of newborn screening analytes which are not sensitive to shipment at ambient temperature, can accurately provide estimates of gestational age comparable to those from published regression models from North America using all analytes. If validated in larger samples especially with more newborns <34 weeks, this technology could substantially facilitate implementation in LMICs.

Leveraging Artificial Intelligence to Improve Pregnancy Dating in Low-Resource Settings

For practicing obstetricians and pediatricians, accurate gestational age (GA) is a cornerstone of quality care during pregnancy and the newborn period. GA determines prenatal management decisions, such as eligibility for antenatal glucocorticoids if preterm delivery is anticipated or timing of delivery in complicated pregnancies to avoid stillbirth or maternal morbidity. Knowledge of GA for women in labor also allows for transfer to facilities capable of handling preterm infants and guides postnatal resuscitation practices and specialized newborn care.

Current Status and Future Directions of Neuromonitoring With Emerging Technologies in Neonatal Care

Neonatology has experienced a significant reduction in mortality rates of the preterm population and critically ill infants over the last few decades. Now, the emphasis is directed toward improving long-term neurodevelopmental outcomes and quality of life. Brain-focused care has emerged as a necessity. The creation of neonatal neurocritical care units, or Neuro-NICUs, provides strategies to reduce brain injury using standardized clinical protocols, methodologies, and provider education and training. Bedside neuromonitoring has dramatically improved our ability to provide assessment of newborns at high risk. Non-invasive tools, such as continuous electroencephalography (cEEG), amplitude-integrated electroencephalography (aEEG), and near-infrared spectroscopy (NIRS), allow screening for seizures and continuous evaluation of brain function and cerebral oxygenation at the bedside. Extended and combined uses of these techniques, also described as multimodal monitoring, may allow practitioners to better understand the physiology of critically ill neonates. Furthermore, the rapid growth of technology in the Neuro-NICU, along with the increasing use of telemedicine and artificial intelligence with improved data mining techniques and machine learning (ML), has the potential to vastly improve decision-making processes and positively impact outcomes. This article will cover the current applications of neuromonitoring in the Neuro-NICU, recent advances, potential pitfalls, and future perspectives in this field.