Improving the Diagnosis of Phenylketonuria by Using a Machine Learning-Based Screening Model of Neonatal MRM Data

Fast Label-Free Metabolic Profile Recognition Identifies Phenylketonuria and Subtypes

Phenylketonuria (PKU) is the most common inherited metabolic disease in humans. Clinical screening of newborn heel blood samples for PKU is costly and time-consuming because it requires multiple procedures, like isotope labeling and derivatization, and PKU subtype identification requires an additional urine sample. Delayed diagnosis of PKU, or subtype identification can result in mental disability. Here, plasmonic silver nanoshells are used for laser desorption/ionization mass spectrometry (MS) detection of PKU with label-free assay by recognizing metabolic profile in dried blood spot (DBS) samples. A total of 1100 subjects are recruited and each DBS sample can be processed in seconds. This platform achieves PKU screening with a sensitivity of 0.985 and specificity of 0.995, which is comparable to existing clinical liquid chromatography MS (LC-MS) methods. This method can process 360 samples per hour, compared with the LC-MS method which processes only 30 samples per hour. Moreover, this assay enables precise identification of PKU subtypes without the need for a urine sample. It is demonstrated that this platform enables high-performance and fast, low-cost PKU screening and subtype identification. This approach might be suitable for the detection of other clinically relevant biomarkers in blood or other clinical samples.

The significance of machine learning in neonatal screening for inherited metabolic diseases

Background

Neonatal screening for inherited metabolic diseases (IMDs) has been revolutionized by tandem mass spectrometry (MS/MS). This study aimed to enhance neonatal screening for IMDs using machine learning (ML) techniques.

Methods

The study involved the analysis of a comprehensive dataset comprising 309,102 neonatal screening records collected in the Ningbo region, China. An advanced ML system model, encompassing nine distinct algorithms, was employed for the purpose of predicting the presence of 31 different IMDs. The model was compared with traditional cutoff schemes to assess its diagnostic efficacy. Additionally, 180 suspected positive cases underwent further evaluation.

Results

The ML system exhibited a significantly reduced positive rate, from 1.17% to 0.33%, compared to cutoff schemes in the initial screening, minimizing unnecessary recalls and associated stress. In suspected positive cases, the ML system identified 142 true positives with high sensitivity (93.42%) and improved specificity (78.57%) compared to the cutoff scheme. While false negatives emerged, particularly in heterozygous carriers, our study revealed the potential of the ML system to detect asymptomatic cases.

Conclusion

This research provides valuable insights into the potential of ML in pediatric medicine for IMD diagnosis through neonatal screening, emphasizing the need for accurate carrier detection and further research in this domain.

Pediatrics in Artificial Intelligence Era: A Systematic Review on Challenges, Opportunities, and Explainability

BACKGROUND

The emergence of artificial intelligence (AI) tools such as ChatGPT and Bard is disrupting a broad swathe of fields, including medicine. In pediatric medicine, AI is also increasingly being used across multiple subspecialties. However, the practical application of AI still faces a number of key challenges. Consequently, there is a requirement for a concise overview of the roles of AI across the multiple domains of pediatric medicine, which the current study seeks to address.

AIM

To systematically assess the challenges, opportunities, and explainability of AI in pediatric medicine.

METHODOLOGY

A systematic search was carried out on peer-reviewed databases, PubMed Central, Europe PubMed Central, and grey literature using search terms related to machine learning (ML) and AI for the years 2016 to 2022 in the English language. A total of 210 articles were retrieved that were screened with PRISMA for abstract, year, language, context, and proximal relevance to research aims. A thematic analysis was carried out to extract findings from the included studies.

RESULTS

Twenty articles were selected for data abstraction and analysis, with three consistent themes emerging from these articles. In particular, eleven articles address the current state-of-the-art application of AI in diagnosing and predicting health conditions such as behavioral and mental health, cancer, syndromic and metabolic diseases. Five articles highlight the specific challenges of AI deployment in pediatric medicines: data security, handling, authentication, and validation. Four articles set out future opportunities for AI to be adapted: the incorporation of Big Data, cloud computing, precision medicine, and clinical decision support systems. These studies collectively critically evaluate the potential of AI in overcoming current barriers to adoption.

CONCLUSION

AI is proving disruptive within pediatric medicine and is presently associated with challenges, opportunities, and the need for explainability. AI should be viewed as a tool to enhance and support clinical decision-making rather than a substitute for human judgement and expertise. Future research should consequently focus on obtaining comprehensive data to ensure the generalizability of research findings.

Machine Learning Methods Improve Specificity in Newborn Screening for Isovaleric Aciduria

Isovaleric aciduria (IVA) is a rare disorder of leucine metabolism and part of newborn screening (NBS) programs worldwide. However, NBS for IVA is hampered by, first, the increased birth prevalence due to the identification of individuals with an attenuated disease variant (so-called "mild" IVA) and, second, an increasing number of false positive screening results due to the use of pivmecillinam contained in the medication. Recently, machine learning (ML) methods have been analyzed, analogous to new biomarkers or second-tier methods, in the context of NBS. In this study, we investigated the application of machine learning classification methods to improve IVA classification using an NBS data set containing 2,106,090 newborns screened in Heidelberg, Germany. Therefore, we propose to combine two methods, linear discriminant analysis, and ridge logistic regression as an additional step, a digital-tier, to traditional NBS. Our results show that this reduces the false positive rate by 69.9% from 103 to 31 while maintaining 100% sensitivity in cross-validation. The ML methods were able to classify mild and classic IVA from normal newborns solely based on the NBS data and revealed that besides isovalerylcarnitine (C5), the metabolite concentration of tryptophan (Trp) is important for improved classification. Overall, applying ML methods to improve the specificity of IVA could have a major impact on newborns, as it could reduce the newborns' and families' burden of false positives or over-treatment.

The effects of a self-learning package on mothers' knowledge and practices towards caring for their children with phenylketonuria

The objective of this study was to evaluate the effect of a self-learning package on mothers' knowledge and practices towards caring for their children with phenylketonuria. A pre/post quasi-experimental study was conducted, including 128 mothers of children diagnosed with phenylketonuria. A specifically designed and validated questionnaire was used to evaluate mothers' knowledge and reported practices toward their children before and after participating in the educational program. There was a highly positive association between knowledge and reported practice (.674 and .398). The self-learning package had a positive impact on mothers' knowledge and practices. Consequently, educational programs should be provided to all mothers of newly diagnosed cases to improve their children's adherence to the therapeutic regimen.

Random forest classifier improving phenylketonuria screening performance in two Chinese populations

Phenylketonuria (PKU) is a genetic disorder with amino acid metabolic defect, which does great harms to the development of newborns and children. Early diagnosis and treatment can effectively prevent the disease progression. Here we developed a PKU screening model using random forest classifier (RFC) to improve PKU screening performance with excellent sensitivity, false positive rate (FPR) and positive predictive value (PPV) in all the validation dataset and two testing Chinese populations. RFC represented outstanding advantages comparing several different classification models based on machine learning and the traditional logistic regression model. RFC is promising to be applied to neonatal PKU screening.

Big Data in Laboratory Medicine—FAIR Quality for AI?

Laboratory medicine is a digital science. Every large hospital produces a wealth of data each day—from simple numerical results from, e.g., sodium measurements to highly complex output of “-omics” analyses, as well as quality control results and metadata. Processing, connecting, storing, and ordering extensive parts of these individual data requires Big Data techniques. Whereas novel technologies such as artificial intelligence and machine learning have exciting application for the augmentation of laboratory medicine, the Big Data concept remains fundamental for any sophisticated data analysis in large databases. To make laboratory medicine data optimally usable for clinical and research purposes, they need to be FAIR: findable, accessible, interoperable, and reusable. This can be achieved, for example, by automated recording, connection of devices, efficient ETL (Extract, Transform, Load) processes, careful data governance, and modern data security solutions. Enriched with clinical data, laboratory medicine data allow a gain in pathophysiological insights, can improve patient care, or can be used to develop reference intervals for diagnostic purposes. Nevertheless, Big Data in laboratory medicine do not come without challenges: the growing number of analyses and data derived from them is a demanding task to be taken care of. Laboratory medicine experts are and will be needed to drive this development, take an active role in the ongoing digitalization, and provide guidance for their clinical colleagues engaging with the laboratory data in research.

A multiplex protein panel assay for severity prediction and outcome prognosis in patients with COVID-19: An observational multi-cohort study

BACKGROUND

Global healthcare systems continue to be challenged by the COVID-19 pandemic, and there is a need for clinical assays that can help optimise resource allocation, support treatment decisions, and accelerate the development and evaluation of new therapies.

METHODS

We developed a multiplexed proteomics assay for determining disease severity and prognosis in COVID-19. The assay quantifies up to 50 peptides, derived from 30 known and newly introduced COVID-19-related protein markers, in a single measurement using routine-lab compatible analytical flow rate liquid chromatography and multiple reaction monitoring (LC-MRM). We conducted two observational studies in patients with COVID-19 hospitalised at Charité - Universitätsmedizin Berlin, Germany before (from March 1 to 26, 2020, n=30) and after (from April 4 to November 19, 2020, n=164) dexamethasone became standard of care. The study is registered in the German and the WHO International Clinical Trials Registry (DRKS00021688).

FINDINGS

The assay produces reproducible (median inter-batch CV of 10.9%) absolute quantification of 47 peptides with high sensitivity (median LLOQ of 143 ng/ml) and accuracy (median 96.8%). In both studies, the assay reproducibly captured hallmarks of COVID-19 infection and severity, as it distinguished healthy individuals, mild, moderate, and severe COVID-19. In the post-dexamethasone cohort, the assay predicted survival with an accuracy of 0.83 (108/130), and death with an accuracy of 0.76 (26/34) in the median 2.5 weeks before the outcome, thereby outperforming compound clinical risk assessments such as SOFA, APACHE II, and ABCS scores.

INTERPRETATION

Disease severity and clinical outcomes of patients with COVID-19 can be stratified and predicted by the routine-applicable panel assay that combines known and novel COVID-19 biomarkers. The prognostic value of this assay should be prospectively assessed in larger patient cohorts for future support of clinical decisions, including evaluation of sample flow in routine setting. The possibility to objectively classify COVID-19 severity can be helpful for monitoring of novel therapies, especially in early clinical trials.

FUNDING

This research was funded in part by the European Research Council (ERC) under grant agreement ERC-SyG-2020 951475 (to M.R) and by the Wellcome Trust (IA 200829/Z/16/Z to M.R.). The work was further supported by the Ministry of Education and Research (BMBF) as part of the National Research Node 'Mass Spectrometry in Systems Medicine (MSCoresys)', under grant agreements 031L0220 and 161L0221. J.H. was supported by a Swiss National Science Foundation (SNSF) Postdoc Mobility fellowship (project number 191052). This study was further supported by the BMBF grant NaFoUniMedCOVID-19 - NUM-NAPKON, FKZ: 01KX2021. The study was co-funded by the UK's innovation agency, Innovate UK, under project numbers 75594 and 56328.

Inferring Time-Lagged Causality Using the Derivative of Single-Cell Expression

Many computational methods have been developed to infer causality among genes using cross-sectional gene expression data, such as single-cell RNA sequencing (scRNA-seq) data. However, due to the limitations of scRNA-seq technologies, time-lagged causal relationships may be missed by existing methods. In this work, we propose a method, called causal inference with time-lagged information (CITL), to infer time-lagged causal relationships from scRNA-seq data by assessing the conditional independence between the changing and current expression levels of genes. CITL estimates the changing expression levels of genes by “RNA velocity”. We demonstrate the accuracy and stability of CITL for inferring time-lagged causality on simulation data against other leading approaches. We have applied CITL to real scRNA data and inferred 878 pairs of time-lagged causal relationships. Furthermore, we showed that the number of regulatory relationships identified by CITL was significantly more than that expected by chance. We provide an R package and a command-line tool of CITL for different usage scenarios.