Nasopharyngeal metabolomics and machine learning approach for the diagnosis of influenza

A Comprehensive Machine Learning Approach for COVID-19 Target Discovery in the Small-Molecule Metabolome

Background/Objectives: Respiratory viruses, including Influenza, RSV, and COVID-19, cause various respiratory infections. Distinguishing these viruses relies on diagnostic methods such as PCR testing. Challenges stem from overlapping symptoms and the emergence of new strains. Advanced diagnostics are crucial for accurate detection and effective management. This study leveraged nasopharyngeal metabolome data to predict respiratory virus scenarios including control vs. RSV, control vs. Influenza A, control vs. COVID-19, control vs. all respiratory viruses, and COVID-19 vs. Influenza A/RSV. Method: We proposed a stacking-based ensemble technique, integrating the top three best-performing ML models from the initial results to enhance prediction accuracy by leveraging the strengths of multiple base learners. Key techniques such as feature ranking, standard scaling, and SMOTE were used to address class imbalances, thus enhancing model robustness. SHAP analysis identified crucial metabolites influencing positive predictions, thereby providing valuable insights into diagnostic markers. Results: Our approach not only outperformed existing methods but also revealed top dominant features for predicting COVID-19, including Lysophosphatidylcholine acyl C18:2, Kynurenine, Phenylalanine, Valine, Tyrosine, and Aspartic Acid (Asp). Conclusions: This study demonstrates the effectiveness of leveraging nasopharyngeal metabolome data and stacking-based ensemble techniques for predicting respiratory virus scenarios. The proposed approach enhances prediction accuracy, provides insights into key diagnostic markers, and offers a robust framework for managing respiratory infections.

Application of machine learning algorithms to identify serological predictors of COVID-19 severity and outcomes

BACKGROUND

Critically ill hospitalized patients with COVID-19 have greater antibody titers than those with mild to moderate illness, but their association with recovery or death from COVID-19 has not been characterized.

METHODS

In a cohort study of 178 COVID-19 patients, 73 non-hospitalized and 105 hospitalized patients, mucosal swabs and plasma samples were collected at hospital enrollment and up to 3 months post-enrollment (MPE) to measure virus RNA, cytokines/chemokines, binding antibodies, ACE2 binding inhibition, and Fc effector antibody responses against SARS-CoV-2. The association of demographic variables and more than 20 serological antibody measures with intubation or death due to COVID-19 was determined using machine learning algorithms.

RESULTS

Predictive models reveal that IgG binding and ACE2 binding inhibition responses at 1 MPE are positively and anti-Spike antibody-mediated complement activation at enrollment is negatively associated with an increased probability of intubation or death from COVID-19 within 3 MPE.

CONCLUSIONS

At enrollment, serological antibody measures are more predictive than demographic variables of subsequent intubation or death among hospitalized COVID-19 patients.

A transformer-based deep learning survival prediction model and an explainable XGBoost anti-PD-1/PD-L1 outcome prediction model based on the cGAS-STING-centered pathways in hepatocellular carcinoma

Recent studies suggest cGAS-STING pathway may play a crucial role in the genesis and development of hepatocellular carcinoma (HCC), closely associated with classical pathways and tumor immunity. We aimed to develop models predicting survival and anti-PD-1/PD-L1 outcomes centered on the cGAS-STING pathway in HCC. We identified classical pathways highly correlated with cGAS-STING pathway and constructed transformer survival model preserving raw structure of pathways. We also developed explainable XGBoost model for predicting anti-PD-1/PD-L1 outcomes using SHAP algorithm. We trained and validated transformer survival model on pan-cancer cohort and tested it on three independent HCC cohorts. Using 0.5 as threshold across cohorts, we divided each HCC cohort into two groups and calculated P values with log-rank test. TCGA-LIHC: C-index = 0.750, P = 1.52e-11; ICGC-LIRI-JP: C-index = 0.741, P = .00138; GSE144269: C-index = 0.647, P = .0233. We trained and validated [area under the receiver operating characteristic curve (AUC) = 0.777] XGBoost model on immunotherapy datasets and tested it on GSE78220 (AUC = 0.789); we also tested XGBoost model on HCC anti-PD-L1 cohort (AUC = 0.719). Our deep learning model and XGBoost model demonstrate potential in predicting survival risks and anti-PD-1/PD-L1 outcomes in HCC. We deployed these two prediction models to the GitHub repository and provided detailed instructions for their usage: deep learning survival model, https://github.com/mlwalker123/CSP_survival_model; XGBoost immunotherapy model, https://github.com/mlwalker123/CSP_immunotherapy_model.

Enhancing Influenza Detection through Integrative Machine Learning and Nasopharyngeal Metabolomic Profiling: A Comprehensive Study

Background/Objectives: Nasal and nasopharyngeal swabs are commonly used for detecting respiratory viruses, including influenza, which significantly alters host cell metabolites. This study aimed to develop a machine learning model to identify biomarkers that differentiate between influenza-positive and -negative cases using clinical metabolomics data. Method: A publicly available dataset of 236 nasopharyngeal samples screened via liquid chromatography-quadrupole time-of-flight (LC/Q-TOF) mass spectrometry was used. Among these, 118 samples tested positive for influenza (40 A H1N1, 39 A H3N2, 39 Influenza B), while 118 were negative controls. A stacking-based model was proposed using the top 20 selected features. Thirteen machine learning models were initially trained, and the top three were combined using predicted probabilities to form a stacking classifier. Results: The ExtraTrees stacking model outperformed other models, achieving 97.08% accuracy. External validation on a prospective cohort of 96 symptomatic individuals (48 positive and 48 negatives for influenza) showed 100% accuracy. SHAP values were used to enhance model explainability. Metabolites such as Pyroglutamic Acid (retention time: 0.81 min, m/z: 84.0447) and its in-source fragment ion (retention time: 0.81 min, m/z: 130.0507) showed minimal impact on influenza-positive cases. On the other hand, metabolites with a retention time of 10.34 min and m/z 106.0865, and a retention time of 8.65 min and m/z 211.1376, demonstrated significant positive contributions. Conclusions: This study highlights the effectiveness of integrating metabolomics data with machine learning for accurate influenza diagnosis. The stacking-based model, combined with SHAP analysis, provided robust performance and insights into key metabolites influencing predictions.

Influenza and the gut microbiota: A hidden therapeutic link

Background

The extensive community of gut microbiota significantly influences various biological functions throughout the body, making its characterization a focal point in biomedicine research. Over the past few decades, studies have revealed a potential link between specific gut bacteria, their associated metabolic pathways, and influenza. Bacterial metabolites can communicate directly or indirectly with organs beyond the gut via the intestinal barrier, thereby impacting the physiological functions of the host. As the microbiota increasingly emerges as a 'gut signature' in influenza, gaining a deeper understanding of its role may offer new insights into its pathophysiological relevance and open avenues for novel therapeutic targets. In this Review, we explore the differences in gut microbiota between healthy individuals and those with influenza, the relationship between gut microbiota metabolites and influenza, and potential strategies for preventing and treating influenza through the regulation of gut microbiota and its metabolites, including fecal microbiota transplantation and microecological preparations.

Methods

We utilized PubMed and Web of Science as our search databases, employing keywords such as "influenza," "gut microbiota," "traditional Chinese medicine," "metabolites," "prebiotics," "probiotics," and "machine learning" to retrieve studies examining the potential therapeutic connections between the modulation of gut microbiota and its metabolites in the treatment of influenza. The search encompassed literature from the inception of the databases up to December 2023.

Results

Fecal microbiota transplantation (FMT), microbial preparations (probiotics and prebiotics), and traditional Chinese medicine have unique advantages in regulating intestinal microbiota and its metabolites to improve influenza outcomes. The primary mechanism involves increasing beneficial intestinal bacteria such as Bacteroidetes and Bifidobacterium while reducing harmful bacteria such as Proteobacteria. These interventions act directly or indirectly on metabolites such as short-chain fatty acids (SCFAs), amino acids (AAs), bile acids, and monoamines to alleviate lung inflammation, reduce viral load, and exert anti-influenza virus effects.

Conclusion

The gut microbiota and its metabolites have direct or indirect therapeutic effects on influenza, presenting broad research potential for providing new directions in influenza research and offering references for clinical prevention and treatment. Future research should focus on identifying key strains, specific metabolites, and immune regulation mechanisms within the gut microbiota to accurately target microbiota interventions and prevent respiratory viral infections such as influenza.

Exploring metabolic anomalies in COVID-19 and post-COVID-19: a machine learning approach with explainable artificial intelligence

The COVID-19 pandemic, caused by SARS-CoV-2, has led to significant challenges worldwide, including diverse clinical outcomes and prolonged post-recovery symptoms known as Long COVID or Post-COVID-19 syndrome. Emerging evidence suggests a crucial role of metabolic reprogramming in the infection's long-term consequences. This study employs a novel approach utilizing machine learning (ML) and explainable artificial intelligence (XAI) to analyze metabolic alterations in COVID-19 and Post-COVID-19 patients. Samples were taken from a cohort of 142 COVID-19, 48 Post-COVID-19, and 38 control patients, comprising 111 identified metabolites. Traditional analysis methods, like PCA and PLS-DA, were compared with ML techniques, particularly eXtreme Gradient Boosting (XGBoost) enhanced by SHAP (SHapley Additive exPlanations) values for explainability. XGBoost, combined with SHAP, outperformed traditional methods, demonstrating superior predictive performance and providing new insights into the metabolic basis of the disease's progression and aftermath. The analysis revealed metabolomic subgroups within the COVID-19 and Post-COVID-19 conditions, suggesting heterogeneous metabolic responses to the infection and its long-term impacts. Key metabolic signatures in Post-COVID-19 include taurine, glutamine, alpha-Ketoglutaric acid, and LysoPC a C16:0. This study highlights the potential of integrating ML and XAI for a fine-grained description in metabolomics research, offering a more detailed understanding of metabolic anomalies in COVID-19 and Post-COVID-19 conditions.

Artificial Intelligence in Metabolomics: A Current Review

Metabolomics and artificial intelligence (AI) form a synergistic partnership. Metabolomics generates large datasets comprising hundreds to thousands of metabolites with complex relationships. AI, aiming to mimic human intelligence through computational modeling, possesses extraordinary capabilities for big data analysis. In this review, we provide a recent overview of the methodologies and applications of AI in metabolomics studies in the context of systems biology and human health. We first introduce the AI concept, history, and key algorithms for machine learning and deep learning, summarizing their strengths and weaknesses. We then discuss studies that have successfully used AI across different aspects of metabolomic analysis, including analytical detection, data preprocessing, biomarker discovery, predictive modeling, and multi-omics data integration. Lastly, we discuss the existing challenges and future perspectives in this rapidly evolving field. Despite limitations and challenges, the combination of metabolomics and AI holds great promises for revolutionary advancements in enhancing human health.

Adjusting for covariates and assessing modeling fitness in machine learning using MUVR2

Motivation

Machine learning (ML) methods are frequently used in Omics research to examine associations between molecular data and for example exposures and health conditions. ML is also used for feature selection to facilitate biological interpretation. Our previous MUVR algorithm was shown to generate predictions and variable selections at state-of-the-art performance. However, a general framework for assessing modeling fitness is still lacking. In addition, enabling to adjust for covariates is a highly desired, but largely lacking trait in ML. We aimed to address these issues in the new MUVR2 framework.

Results

The MUVR2 algorithm was developed to include the regularized regression framework elastic net in addition to partial least squares and random forest modeling. Compared with other cross-validation strategies, MUVR2 consistently showed state-of-the-art performance, including variable selection, while minimizing overfitting. Testing on simulated and real-world data, we also showed that MUVR2 allows for the adjustment for covariates using elastic net modeling, but not using partial least squares or random forest.

Availability and implementation

Algorithms, data, scripts, and a tutorial are open source under GPL-3 license and available in the MUVR2 R package at https://github.com/MetaboComp/MUVR2.

NMR and MS reveal characteristic metabolome atlas and optimize esophageal squamous cell carcinoma early detection

Metabolic changes precede malignant histology. However, it remains unclear whether detectable characteristic metabolome exists in esophageal squamous cell carcinoma (ESCC) tissues and biofluids for early diagnosis. Here, we conduct NMR- and MS-based metabolomics on 1,153 matched ESCC tissues, normal mucosae, pre- and one-week post-operative sera and urines from 560 participants across three hospitals, with machine learning and WGCNA. Aberrations in 'alanine, aspartate and glutamate metabolism' proved to be prevalent throughout the ESCC evolution, consistently identified by NMR and MS, and reflected in 16 serum and 10 urine metabolic signatures in both discovery and validation sets. NMR-based simplified panels of any five serum or urine metabolites outperform clinical serological tumor markers (AUC = 0.984 and 0.930, respectively), and are effective in distinguishing early-stage ESCC in test set (serum accuracy = 0.994, urine accuracy = 0.879). Collectively, NMR-based biofluid screening can reveal characteristic metabolic events of ESCC and be feasible for early detection (ChiCTR2300073613).

Application of machine learning models to identify serological predictors of COVID-19 severity and outcomes

Critically ill people with COVID-19 have greater antibody titers than those with mild to moderate illness, but their association with recovery or death from COVID-19 has not been characterized. In 178 COVID-19 patients, 73 non-hospitalized and 105 hospitalized patients, mucosal swabs and plasma samples were collected at hospital enrollment and up to 3 months post-enrollment (MPE) to measure virus RNA, cytokines/chemokines, binding antibodies, ACE2 binding inhibition, and Fc effector antibody responses against SARS-CoV-2. The association of demographic variables and >20 serological antibody measures with intubation or death due to COVID-19 was determined using machine learning algorithms. Predictive models revealed that IgG binding and ACE2 binding inhibition responses at 1 MPE were positively and C1q complement activity at enrollment was negatively associated with an increased probability of intubation or death from COVID-19 within 3 MPE. Serological antibody measures were more predictive than demographic variables of intubation or death among COVID-19 patients.

Metabolomics and machine learning approaches for diagnostic and prognostic biomarkers screening in sepsis

BACKGROUND

Sepsis is a life-threatening disease with a poor prognosis, and metabolic disorders play a crucial role in its development. This study aims to identify key metabolites that may be associated with the accurate diagnosis and prognosis of sepsis.

METHODS

Septic patients and healthy individuals were enrolled to investigate metabolic changes using non-targeted liquid chromatography-high-resolution mass spectrometry metabolomics. Machine learning algorithms were subsequently employed to identify key differentially expressed metabolites (DEMs). Prognostic-related DEMs were then identified using univariate and multivariate Cox regression analyses. The septic rat model was established to verify the effect of phenylalanine metabolism-related gene MAOA on survival and mean arterial pressure after sepsis.

RESULTS

A total of 532 DEMs were identified between healthy control and septic patients using metabolomics. The main pathways affected by these DEMs were amino acid biosynthesis, phenylalanine metabolism, tyrosine metabolism, glycine, serine and threonine metabolism, and arginine and proline metabolism. To identify sepsis diagnosis-related biomarkers, support vector machine (SVM) and random forest (RF) algorithms were employed, leading to the identification of four biomarkers. Additionally, analysis of transcriptome data from sepsis patients in the GEO database revealed a significant up-regulation of the phenylalanine metabolism-related gene MAOA in sepsis. Further investigation showed that inhibition of MAOA using the inhibitor RS-8359 reduced phenylalanine levels and improved mean arterial pressure and survival rate in septic rats. Finally, using univariate and multivariate cox regression analysis, six DEMs were identified as prognostic markers for sepsis.

CONCLUSIONS

This study employed metabolomics and machine learning algorithms to identify differential metabolites that are associated with the diagnosis and prognosis of sepsis patients. Unraveling the relationship between metabolic characteristics and sepsis provides new insights into the underlying biological mechanisms, which could potentially assist in the diagnosis and treatment of sepsis.

TRIAL REGISTRATION

This human study was approved by the Ethics Committee of the Research Institute of Surgery (2021-179) and was registered by the Chinese Clinical Trial Registry (Date: 09/12/2021, ChiCTR2200055772).

Supervised Machine Learning Methods for Seasonal Influenza Diagnosis

Influenza has been a stationary disease in Mexico since 2009, and this causes a high cost for the national public health system, including its detection using RT-qPCR tests, treatments, and absenteeism in the workplace. Despite influenza's relevance, the main clinical features to detect the disease defined by international institutions like the World Health Organization (WHO) and the United States Centers for Disease Control and Prevention (CDC) do not follow the same pattern in all populations. The aim of this work is to find a machine learning method to facilitate decision making in the clinical differentiation between positive and negative influenza patients, based on their symptoms and demographic features. The research sample consisted of 15480 records, including clinical and demographic data of patients with a positive/negative RT-qPCR influenza tests, from 2010 to 2020 in the public healthcare institutions of Mexico City. The performance of the methods for classifying influenza cases were evaluated with indices like accuracy, specificity, sensitivity, precision, the f1-measure and the area under the curve (AUC). Results indicate that random forest and bagging classifiers were the best supervised methods; they showed promise in supporting clinical diagnosis, especially in places where performing molecular tests might be challenging or not feasible.

Artificial Intelligence and Infectious Disease Imaging

The mass production of the graphics processing unit and the coronavirus disease 2019 (COVID-19) pandemic have provided the means and the motivation, respectively, for rapid developments in artificial intelligence (AI) and medical imaging techniques. This has led to new opportunities to improve patient care but also new challenges that must be overcome before these techniques are put into practice. In particular, early AI models reported high performances but failed to perform as well on new data. However, these mistakes motivated further innovation focused on developing models that were not only accurate but also stable and generalizable to new data. The recent developments in AI in response to the COVID-19 pandemic will reap future dividends by facilitating, expediting, and informing other medical AI applications and educating the broad academic audience on the topic. Furthermore, AI research on imaging animal models of infectious diseases offers a unique problem space that can fill in evidence gaps that exist in clinical infectious disease research. Here, we aim to provide a focused assessment of the AI techniques leveraged in the infectious disease imaging research space, highlight the unique challenges, and discuss burgeoning solutions.

Targeted plasma metabolomics combined with machine learning for the diagnosis of severe acute respiratory syndrome virus type 2

Introduction

The routine clinical diagnosis of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is largely restricted to real-time reverse transcription quantitative PCR (RT-qPCR), and tests that detect SARS-CoV-2 nucleocapsid antigen. Given the diagnostic delay and suboptimal sensitivity associated with these respective methods, alternative diagnostic strategies are needed for acute infection.

Methods

We studied the use of a clinically validated liquid chromatography triple quadrupole method (LC/MS–MS) for detection of amino acids from plasma specimens. We applied machine learning models to distinguish between SARS-CoV-2-positive and negative samples and analyzed amino acid feature importance.

Results

A total of 200 samples were tested, including 70 from individuals with COVID-19, and 130 from negative controls. The top performing model overall allowed discrimination between SARS-CoV-2-positive and negative control samples with an area under the receiver operating characteristic curve (AUC) of 0.96 (95%CI 0.91, 1.00), overall sensitivity of 0.99 (95%CI 0.92, 1.00), and specificity of 0.92 (95%CI 0.85, 0.95).

Discussion

This approach holds potential as an alternative to existing methods for the rapid and accurate diagnosis of acute SARS-CoV-2 infection.

Interpretable machine learning with tree-based shapley additive explanations: Application to metabolomics datasets for binary classification

Machine learning (ML) models are used in clinical metabolomics studies most notably for biomarker discoveries, to identify metabolites that discriminate between a case and control group. To improve understanding of the underlying biomedical problem and to bolster confidence in these discoveries, model interpretability is germane. In metabolomics, partial least square discriminant analysis (PLS-DA) and its variants are widely used, partly due to the model's interpretability with the Variable Influence in Projection (VIP) scores, a global interpretable method. Herein, Tree-based Shapley Additive explanations (SHAP), an interpretable ML method grounded in game theory, was used to explain ML models with local explanation properties. In this study, ML experiments (binary classification) were conducted for three published metabolomics datasets using PLS-DA, random forests, gradient boosting, and extreme gradient boosting (XGBoost). Using one of the datasets, PLS-DA model was explained using VIP scores, while one of the best-performing models, a random forest model, was interpreted using Tree SHAP. The results show that SHAP has a more explanation depth than PLS-DA's VIP, making it a powerful method for rationalizing machine learning predictions from metabolomics studies.

Multi-omics association analysis reveals interactions between the oropharyngeal microbiome and the metabolome in pediatric patients with influenza A virus pneumonia

Children are at high risk for influenza A virus (IAV) infections, which can develop into severe illnesses. However, little is known about interactions between the microbiome and respiratory tract metabolites and their impact on the development of IAV pneumonia in children. Using a combination of liquid chromatography tandem mass spectrometry (LC-MS/MS) and 16S rRNA gene sequencing, we analyzed the composition and metabolic profile of the oropharyngeal microbiota in 49 pediatric patients with IAV pneumonia and 42 age-matched healthy children. The results indicate that compared to healthy children, children with IAV pneumonia exhibited significant changes in the oropharyngeal macrobiotic structure (p = 0.001), and significantly lower microbial abundance and diversity (p < 0.05). These changes came with significant disturbances in the levels of oropharyngeal metabolites. Intergroup differences were observed in 204 metabolites mapped to 36 metabolic pathways. Significantly higher levels of sphingolipid (sphinganine and phytosphingosine) and propanoate (propionic acid and succinic acid) metabolism were observed in patients with IAV pneumonia than in healthy controls. Using Spearman’s rank-correlation analysis, correlations between IAV pneumonia-associated discriminatory microbial genera and metabolites were evaluated. The results indicate significant correlations and consistency in variation trends between Streptococcus and three sphingolipid metabolites (phytosphingosine, sphinganine, and sphingosine). Besides these three sphingolipid metabolites, the sphinganine-to-sphingosine ratio and the joint analysis of the three metabolites indicated remarkable diagnostic efficacy in children with IAV pneumonia. This study confirmed significant changes in the characteristics and metabolic profile of the oropharyngeal microbiome in pediatric patients with IAV pneumonia, with high synergy between the two factors. Oropharyngeal sphingolipid metabolites may serve as potential diagnostic biomarkers of IAV pneumonia in children.

Precision Medicine Approaches with Metabolomics and Artificial Intelligence

Recent technological innovations in the field of mass spectrometry have supported the use of metabolomics analysis for precision medicine. This growth has been allowed also by the application of algorithms to data analysis, including multivariate and machine learning methods, which are fundamental to managing large number of variables and samples. In the present review, we reported and discussed the application of artificial intelligence (AI) strategies for metabolomics data analysis. Particularly, we focused on widely used non-linear machine learning classifiers, such as ANN, random forest, and support vector machine (SVM) algorithms. A discussion of recent studies and research focused on disease classification, biomarker identification and early diagnosis is presented. Challenges in the implementation of metabolomics-AI systems, limitations thereof and recent tools were also discussed.

Treatment of Recurrent Nasopharyngeal Carcinoma: A Sequential Challenge

Recurrent nasopharyngeal carcinoma (NPC), which occurs in 10-20% of patients with primary NPC after the initial treatment modality of intensity-modulated radiation therapy (IMRT), is one of the major causes of death among NPC patients. Patients with recurrent disease without distant metastases still have a chance to be saved, but re-treatment often carries more serious toxicities or higher risks. For this group of patients, both otolaryngologists and oncologists are committed to developing more appropriate treatment regimens that can prolong patient survival and improve survival therapy. Currently, there are no international guidelines for the treatment of patients with recurrent NPC. In this article, we summarize past publications on clinical research and mechanistic studies related to recurrent NPC, combined with the experience and lessons learned by our institutional multidisciplinary team in the treatment of recurrent NPC. We propose an objective protocol for the treatment of recurrent NPC.

Layer recurrent neural network-based diagnosis of Parkinson’s disease using voice features

Parkinson’s disease (PD), a slow-progressing neurological disease, affects a large percentage of the world’s elderly population, and this population is expected to grow over the next decade. As a result, early detection is crucial for community health and the future of the globe in order to take proper safeguards and have a less arduous treatment procedure. Recent research has begun to focus on the motor system deficits caused by PD. Because practically most of the PD patients suffer from voice abnormalities, researchers working on automated diagnostic systems investigate vocal impairments. In this paper, we undertake extensive experiments with features extracted from voice signals. We propose a layer Recurrent Neural Network (RNN) based diagnosis for PD. To prove the efficiency of the model, different network models are compared. To the best of our knowledge, several neural network topologies, namely RNN, Cascade Forward Neural Networks (CFNN), and Feed Forward Neural Networks (FFNN), are used and compared for voice-based PD detection for the first time. In addition, the impacts of data normalization and feature selection (FS) are thoroughly examined. The findings reveal that normalization increases classifier performance and Laplacian-based FS outperforms. The proposed RNN model with 300 voice features achieves 99.74% accuracy.

Rapid triage for ischemic stroke: a machine learning-driven approach in the context of predictive, preventive and personalised medicine

BACKGROUND

Recognising the early signs of ischemic stroke (IS) in emergency settings has been challenging. Machine learning (ML), a robust tool for predictive, preventive and personalised medicine (PPPM/3PM), presents a possible solution for this issue and produces accurate predictions for real-time data processing.

METHODS

This investigation evaluated 4999 IS patients among a total of 10,476 adults included in the initial dataset, and 1076 IS subjects among 3935 participants in the external validation dataset. Six ML-based models for the prediction of IS were trained on the initial dataset of 10,476 participants (split participants into a training set [80%] and an internal validation set [20%]). Selected clinical laboratory features routinely assessed at admission were used to inform the models. Model performance was mainly evaluated by the area under the receiver operating characteristic (AUC) curve. Additional techniques-permutation feature importance (PFI), local interpretable model-agnostic explanations (LIME), and SHapley Additive exPlanations (SHAP)-were applied for explaining the black-box ML models.

RESULTS

Fifteen routine haematological and biochemical features were selected to establish ML-based models for the prediction of IS. The XGBoost-based model achieved the highest predictive performance, reaching AUCs of 0.91 (0.90-0.92) and 0.92 (0.91-0.93) in the internal and external datasets respectively. PFI globally revealed that demographic feature age, routine haematological parameters, haemoglobin and neutrophil count, and biochemical analytes total protein and high-density lipoprotein cholesterol were more influential on the model's prediction. LIME and SHAP showed similar local feature attribution explanations.

CONCLUSION

In the context of PPPM/3PM, we used the selected predictors obtained from the results of common blood tests to develop and validate ML-based models for the diagnosis of IS. The XGBoost-based model offers the most accurate prediction. By incorporating the individualised patient profile, this prediction tool is simple and quick to administer. This is promising to support subjective decision making in resource-limited settings or primary care, thereby shortening the time window for the treatment, and improving outcomes after IS.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s13167-022-00283-4.