Feed-forward neural networks using cerebral MR spectroscopy and DTI might predict neurodevelopmental outcome in preterm neonates

Developing a practical neurodevelopmental prediction model for targeting high-risk very preterm infants during visit after NICU: a retrospective national longitudinal cohort study

BACKGROUND

Follow-up visits for very preterm infants (VPI) after hospital discharge is crucial for their neurodevelopmental trajectories, but ensuring their attendance before 12 months corrected age (CA) remains a challenge. Current prediction models focus on future outcomes at discharge, but post-discharge data may enhance predictions of neurodevelopmental trajectories due to brain plasticity. Few studies in this field have utilized machine learning models to achieve this potential benefit with transparency, explainability, and transportability.

METHODS

We developed four prediction models for cognitive or motor function at 24 months CA separately at each follow-up visits, two for the 6-month and two for the 12-month CA visits, using hospitalized and follow-up data of VPI from the Taiwan Premature Infant Follow-up Network from 2010 to 2017. Regression models were employed at 6 months CA, defined as a decline in The Bayley Scales of Infant Development 3rd edition (BSIDIII) composite score > 1 SD between 6- and 24-month CA. The delay models were developed at 12 months CA, defined as a BSIDIII composite score < 85 at 24 months CA. We used an evolutionary-derived machine learning method (EL-NDI) to develop models and compared them to those built by lasso regression, random forest, and support vector machine.

RESULTS

One thousand two hundred forty-four VPI were in the developmental set and the two validation cohorts had 763 and 1347 VPI, respectively. EL-NDI used only 4-10 variables, while the others required 29 or more variables to achieve similar performance. For models at 6 months CA, the area under the receiver operating curve (AUC) of EL-NDI were 0.76-0.81(95% CI, 0.73-0.83) for cognitive regress with 4 variables and 0.79-0.83 (95% CI, 0.76-0.86) for motor regress with 4 variables. For models at 12 months CA, the AUC of EL-NDI were 0.75-0.78 (95% CI, 0.72-0.82) for cognitive delay with 10 variables and 0.73-0.82 (95% CI, 0.72-0.85) for motor delay with 4 variables.

CONCLUSIONS

Our EL-NDI demonstrated good performance using simpler, transparent, explainable models for clinical purpose. Implementing these models for VPI during follow-up visits may facilitate more informed discussions between parents and physicians and identify high-risk infants more effectively for early intervention.

1H-magnetic resonance spectroscopy and its role in predicting neurodevelopmental impairment in preterm neonates: A systematic review

To assess the diagnostic utility of proton (1H) magnetic resonance spectroscopy in early diagnosis of neurodevelopmental impairment in preterm newborns. Systematic review performed in compliance with the PRISMA statements. Eligible articles were searched in MEDLINE, Scopus, and ISI Web of Science databases using the following medical subject headings and terms: "magnetic resonance spectroscopy," "infant," and "newborn." Studies of any design published until 20 December 2021 and fulfilling the following criteria were selected: (1) studies including newborns with gestational age at birth <37 weeks which underwent at least one 1H-MRS scan within 52 weeks' postmenstrual age and neurodevelopmental assessment within 4 years of age; (2) studies in which preterm newborns with congenital infections, genetic disorders, and brain congenital anomalies were clearly excluded. Data regarding the relationship between metabolite ratios in basal ganglia, thalamus, and white matter, and neurodevelopment were analysed. The quality assessment of included studies was performed according to the criteria from the QUADAS-2. N-acetylaspartate (NAA)/choline (Cho) was the most studied metabolite ratio. Lower NAA/Cho ratio in basal ganglia and thalamus was associated with adverse motor, cognitive, and language outcomes, and worse global neurodevelopment. Lower NAA/Cho ratio in white matter was associated with cognitive impairment. However, some associations came from single studies or were discordant among studies. The quality of included studies was low. 1H-MRS could be a promising tool for early diagnosis of neurodevelopmental impairment. However, further studies of good quality are needed to define the relationship between metabolite ratios and neurodevelopment.

Endoscopic Surgical Treatment of Osteoarthritis and Prognostic Model Construction

Purpose

Osteoarthritis (OA) is a degenerative disease of joints. Currently, there is still a lack of effective tools to predict the long-term efficacy of surgical treatment of OA. The purpose of this study was to explore the prognostic factors of endoscopic surgery for OA and to predict the long-term efficacy of this type of surgery for OA by establishing a prognostic model.

Methods

Baseline and follow-up data on 236 OA patients who underwent surgery in our hospital from January 2017 to December 2021 were selected and patients were randomly assigned to a training set (n = 165) and a test set (n = 71). The Pearson correlation coefficient was used to analyze the correlation between features. Feature selection was performed by recursive feature elimination (RFE) and linear regression. K-means clustering analysis was performed on the selected features to obtain the number of output layers. Finally, a single hidden layer error backpropagation (BP)-artificial neural network (ANN) model was established on the training set, and receiver operating characteristic (ROC) curve was drawn on the test set for verification.

Results

Correlation analysis revealed no redundancy among features. RFE and linear regression screened out the features associated with postoperative prognosis under endoscopic surgery: sex, age, BMI, region, morning stiffness time, step count, and osteophyte area. K-means clustering yielded that the optimal number of categories was three, the same as the number of categories for the outcome variable. Therefore, a 7-1-3 BP neural network model was established based on these 7 features, and this model could predict the postoperative situation within one year to a relatively accurate extent: area under curve values (AUC) were 0.814, 0.700, and 0.761 in patients with worse, unchanged, and improved conditions one year after surgery, respectively, higher than the multiclass AUC value (0.646).

Conclusion

The prognostic model of endoscopic surgery for OA constructed in this study can well predict the disease progression of patients within one year after surgery.

Machine Learning Prediction Models for Neurodevelopmental Outcome After Preterm Birth: A Scoping Review and New Machine Learning Evaluation Framework

BACKGROUND AND OBJECTIVES

Outcome prediction of preterm birth is important for neonatal care, yet prediction performance using conventional statistical models remains insufficient. Machine learning has a high potential for complex outcome prediction. In this scoping review, we provide an overview of the current applications of machine learning models in the prediction of neurodevelopmental outcomes in preterm infants, assess the quality of the developed models, and provide guidance for future application of machine learning models to predict neurodevelopmental outcomes of preterm infants.

METHODS

A systematic search was performed using PubMed. Studies were included if they reported on neurodevelopmental outcome prediction in preterm infants using predictors from the neonatal period and applying machine learning techniques. Data extraction and quality assessment were independently performed by 2 reviewers.

RESULTS

Fourteen studies were included, focusing mainly on very or extreme preterm infants, predicting neurodevelopmental outcome before age 3 years, and mostly assessing outcomes using the Bayley Scales of Infant Development. Predictors were most often based on MRI. The most prevalent machine learning techniques included linear regression and neural networks. None of the studies met all newly developed quality assessment criteria. Studies least prone to inflated performance showed promising results, with areas under the curve up to 0.86 for classification and R2 values up to 91% in continuous prediction. A limitation was that only 1 data source was used for the literature search.

CONCLUSIONS

Studies least prone to inflated prediction results are the most promising. The provided evaluation framework may contribute to improved quality of future machine learning models.

Lactylation may be a Novel Posttranslational Modification in Inflammation in Neonatal Hypoxic-Ischemic Encephalopathy

Perinatal hypoxia-ischemia remains the most common cause of acute neonatal brain injury and is associated with a high death rate and long-term neurological abnormalities such as memory and cognitive deficits and dyskinesia. Hypoxia-ischemia triggers an inflammatory cascade in the brain that is amplified by the activation of immune cells and the influx of peripheral immune cells into the brain parenchyma in response to cellular injury. Thus, acute cerebral hypoxic-ischemic inflammation is a major contributor to the pathogenesis of newborn hypoxic-ischemic brain injury. Lactate is a glycolysis end product that can regulate inflammation through histone lactylation, a unique posttranslational modification that was identified in recent studies. The purpose of this review is to outline the recent improvements in our understanding of microglia-mediated hypoxic-ischemic inflammation and to further discuss how histone lactylation regulates inflammation by affecting macrophage activation. These findings may suggest that epigenetic reprogramming-associated lactate input is linked to disease outcomes such as acute neonatal brain injury pathogenesis and the therapeutic effects of drugs and other strategies in relieving neonatal hypoxic-ischemic brain injury. Therefore, improving our knowledge of the reciprocal relationships between histone lactylation and inflammation could lead to the development of new immunomodulatory therapies for brain damage in newborns.

A Diffusion Tensor Imaging Analysis of Frontal Lobe White Matter Microstructure in Trigonocephaly Patients

BACKGROUND

Children with trigonocephaly are at risk for neurodevelopmental disorders. The aim of this study is to investigate white matter properties of the frontal lobes in young, unoperated patients with metopic synostosis as compared to healthy controls using diffusion tension imaging (DTI).

METHODS

Preoperative DTI data sets of 46 patients with trigonocephaly with a median age of 0.49 (interquartile range: 0.38) years were compared with 21 controls with a median age of 1.44 (0.98) years. White matter metrics of the tracts in the frontal lobe were calculated using FMRIB Software Library (FSL). The mean value of tract-specific fractional anisotropy (FA) and mean diffusivity (MD) were estimated for each subject and compared to healthy controls. By linear regression, FA and MD values per tract were assessed by trigonocephaly, sex, and age.

RESULTS

The mean FA and MD values in the frontal lobe tracts of untreated trigonocephaly patients, younger than 3 years, were not significantly different in comparison to controls, where age showed to be a significant associated factor.

CONCLUSIONS

Microstructural parameters of white matter tracts of the frontal lobe of patients with trigonocephaly are comparable to those of controls aged 0-3 years.

Pharmacological Neuroprotection of the Preterm Brain: Current Evidence and Perspectives

Despite improvements in viability, the long-term neurodevelopmental outcomes of preterm babies remain serious concern as a significant percentage of these infants develop neurological and/or intellectual impairment, and they are also at increased risk of psychiatric illnesses later in life. The current challenge is to develop neuroprotective approaches to improve adverse outcomes in preterm survivors. The purpose of this review was to provide an overview of the current evidence on pharmacological agents targeting the neuroprotection of the preterm brain. Among them, magnesium sulfate, given antenatally to pregnant women with imminent preterm birth before 30 to 34 weeks of gestation, as well as caffeine administered to preterm infants after birth, exhibited neuroprotective effects for human preterm brain. Erythropoietin treatment of preterm infants did not result in neuroprotection at 2 years of age in two out of three published large randomized controlled trials; however, long-term follow-up of these infants is needed to come to definite conclusions. Further studies are also required to assess whether melatonin, neurosteroids, inhaled nitric oxide, allopurinol, or dietary supplements (omega-3 fatty acids, choline, curcumin, etc.) could be implemented as neuroprotectants in clinical practice. Furthermore, other pharmacological agents showing promising signs of neuroprotective efficacy in preclinical studies (growth factors, hyaluronidase inhibitors or treatment, antidiabetic drugs, cannabidiol, histamine-H3 receptor antagonists, etc.), as well as stem cell- or exosomal-based therapies and nanomedicine, may prove useful in the future as potential neuroprotective approaches for human preterm brain. KEY POINTS: · Magnesium and caffeine have neuroprotective effects for the preterm brain.. · Follow-up of infants treated with erythropoietin is needed.. · Neuroprotective efficacy of several drugs in animals needs to be shown in humans..