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

Identification of Blood Biomarkers in Ischemic Stroke by Integrated Analysis of Metabolomics and Proteomics

We aimed to uncover the pathological mechanism of ischemic stroke (IS) using a combined analysis of untargeted metabolomics and proteomics. The serum samples from a discovery set of 44 IS patients and 44 matched controls were analyzed using a specific detection method. The same method was then used to validate metabolites and proteins in the two validation sets: one with 30 IS patients and 30 matched controls, and the other with 50 IS patients and 50 matched controls. A total of 105 and 221 differentially expressed metabolites or proteins were identified, and the association between the two omics was determined in the discovery set. Enrichment analysis of the top 25 metabolites and 25 proteins in the two-way orthogonal partial least-squares with discriminant analysis, which was employed to identify highly correlated biomarkers, highlighted 15 pathways relevant to the pathological process. One metabolite and seven proteins exhibited differences between groups in the validation set. The binary logistic regression model, which included metabolite 2-hydroxyhippuric acid and proteins APOM_O95445, MASP2_O00187, and PRTN3_D6CHE9, achieved an area under the curve of 0.985 (95% CI: 0.966-1) in the discovery set. This study elucidated alterations and potential coregulatory influences of metabolites and proteins in the blood of IS patients.

Machine Learning Approaches for Stroke Risk Prediction: Findings from the Suita Study

Stroke constitutes a significant public health concern due to its impact on mortality and morbidity. This study investigates the utility of machine learning algorithms in predicting stroke and identifying key risk factors using data from the Suita study, comprising 7389 participants and 53 variables. Initially, unsupervised k-prototype clustering categorized participants into risk clusters, while five supervised models including Logistic Regression (LR), Random Forest (RF), Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosted Machine (LightGBM) were employed to predict stroke outcomes. Stroke incidence disparities among identified risk clusters using the unsupervised k-prototype clustering method are substantial, according to the findings. Supervised learning, particularly RF, was a preferable option because of the higher levels of performance metrics. The Shapley Additive Explanations (SHAP) method identified age, systolic blood pressure, hypertension, estimated glomerular filtration rate, metabolic syndrome, and blood glucose level as key predictors of stroke, aligning with findings from the unsupervised clustering approach in high-risk groups. Additionally, previously unidentified risk factors such as elbow joint thickness, fructosamine, hemoglobin, and calcium level demonstrate potential for stroke prediction. In conclusion, machine learning facilitated accurate stroke risk predictions and highlighted potential biomarkers, offering a data-driven framework for risk assessment and biomarker discovery.

The Emergency Medical Team Operating System - a vision for field hospital data management in following the concepts of predictive, preventive, and personalized medicine

In times where sudden-onset disasters (SODs) present challenges to global health systems, the integration of predictive, preventive, and personalized medicine (PPPM / 3PM) into emergency medical responses has manifested as a critical necessity. We introduce a modern electronic patient record system designed specifically for emergency medical teams (EMTs), which will serve as a novel approach in how digital healthcare management can be optimized in crisis situations. This research is based on the principle that advanced information technology (IT) systems are key to transforming humanitarian aid by offering predictive insights, preventive strategies, and personalized care in disaster scenarios. We aim to address the critical gaps in current emergency medical response strategies, particularly in the context of SODs. Building upon a collaborative effort with European emergency medical teams, we have developed a comprehensive and scalable electronic patient record system. It not only enhances patient management during emergencies but also enables predictive analytics to anticipate patient needs, preventive guidelines to reduce the impact of potential health threats, and personalized treatment plans for the individual needs of patients. Furthermore, our study examines the possibilities of adopting PPPM-oriented IT solutions in disaster relief. By integrating predictive models for patient triage, preventive measures to mitigate health risks, and personalized care protocols, potential improvements to patient health or work efficiency could be established. This system was evaluated with clinical experts and shall be used to establish digital solutions and new forms of assistance for humanitarian aid in the future. In conclusion, to really achieve PPPM-related efforts more investment will need to be put into research and development of electronic patient records as the foundation as well as into the clinical processes along all pathways of stakeholders in disaster medicine.

Serum Olink Proteomics-Based Identification of Protein Biomarkers Associated with the Immune Response in Ischemic Stroke

The immune response is considered essential for pathology of ischemic stroke (IS), but it remains unclear which immune response-related proteins exhibit altered expression in IS patients. Here, we used Olink proteomics to examine the expression levels of 92 immune response-related proteins in the sera of IS patients (n = 88) and controls (n = 88), and we found that 59 of these proteins were differentially expressed. Feature variables were screened from the differentially expressed proteins by the least absolute shrinkage and selection operator (LASSO) and the random forest and by determining whether their proteins had an area under the curve (AUC) greater than 0.8. Ultimately, we identified six potential protein biomarkers of IS, namely, MASP1, STC1, HCLS1, CLEC4D, PTH1R, and PIK3AP1, and established a logistic regression model that used these proteins to diagnose IS. The AUCs of the models in the internal validation and the test set were 0.962 (95% confidence interval (CI): 0.895-1.000) and 0.954 (95% CI: 0.884-1.000), respectively, and the same protein detection method was performed in an external independent validation set (AUC: 0.857 (95% CI: 0.801-0.913)). These proteins may play a role in immune regulation via the C-type lectin receptor signaling pathway, the PI3K-AKT signaling pathway, and the B-cell receptor signaling pathway.

Artificial Intelligence in Optimizing the Functioning of Emergency Departments; a Systematic Review of Current Solutions

Introduction

The burgeoning burden on emergency departments is a global challenge that we have been confronting for many years. Emerging artificial intelligence (AI)-based solutions may constitute a critical component in the optimization of these units. This systematic review was conducted to thoroughly examine and summarize the currently available AI solutions, assess potential benefits from their implementation, and identify anticipated directions of further development in this fascinating and rapidly evolving field.

Methods

This systematic review utilized data compiled from three key scientific databases: PubMed (2045 publications), Scopus (877 publications), and Web of Science (2495 publications). After meticulous removal of duplicates, we conducted a detailed analysis of 2052 articles, including 147 full-text papers. From these, we selected 51 of the most pertinent and representative publications for the review.

Results

Overall the present research indicates that due to high accuracy and sensitivity of machine learning (ML) models it's reasonable to use AI in support of doctors as it can show them the potential diagnosis, which could save time and resources. However, AI-generated diagnoses should be verified by a doctor as AI is not infallible.

Conclusions

Currently available AI algorithms are capable of analysing complex medical data with unprecedented precision and speed. Despite AI's vast potential, it is still a nascent technology that is often perceived as complicated and challenging to implement. We propose that a pivotal point in effectively harnessing this technology is the close collaboration between medical professionals and AI experts. Future research should focus on further refining AI algorithms, performing comprehensive validation, and introducing suitable legal regulations and standard procedures, thereby fully leveraging the potential of AI to enhance the quality and efficiency of healthcare delivery.

Designing machine learning for big data: A study to identify factors that increase the risk of ischemic stroke and prognosis in hypertensive patients

Background

Ischemic stroke (IS) accounts large amount of stroke incidence. The aim of this study was to discover the risk and prognostic factors that affecting the occurrence of IS in hypertensive patients.

Method

Study data were obtained from the Medical Information Mart for Intensive Care (MIMIC)-IV database. To avoid biased factors selection process, several approaches were studied including logistic regression, elastic net regression, random forest, correlation analysis, and multifactor logistic regression methods. And seven different machine-learning methods are used to construct predictive models. The performance of the developed models was evaluated using AUC (Area Under the Curve), prediction accuracy, precision, recall, F1 score, PPV (Positive Predictive Value) and NPV (Negative Predictive Value). Interaction analysis was conducted to explore potential relationships between influential factors.

Results

The study included 92,514 hypertensive patients, of which 1746 hypertensive patients experienced IS. The Gradient Boosted Decision Tree (GBDT) model outperformed the other prediction model terms of prediction accuracy and AUC values in both ischemic and prognosis cases. By using the SHapley Additive exPlanations (SHAP), we found that a range of factors and corresponding interactions between factors are important risk factors for IS and its prognosis in hypertensive patients.

Conclusion

The study identified factors that increase the risk of IS and poor prognosis in hypertensive patients, which may provide guidance for clinical diagnosis and treatment.

Predicting ischemic stroke risk from atrial fibrillation based on multi-spectral fundus images using deep learning

Background

Ischemic stroke (IS) is one of the most common serious secondary diseases of atrial fibrillation (AF) within 1 year after its occurrence, both of which have manifestations of ischemia and hypoxia of the small vessels in the early phase of the condition. The fundus is a collection of capillaries, while the retina responds differently to light of different wavelengths. Predicting the risk of IS occurring secondary to AF, based on subtle differences in fundus images of different wavelengths, is yet to be explored. This study was conducted to predict the risk of IS occurring secondary to AF based on multi-spectrum fundus images using deep learning.

Methods

A total of 150 AF participants without suffering from IS within 1 year after discharge and 100 IS participants with persistent arrhythmia symptoms or a history of AF diagnosis in the last year (defined as patients who would develop IS within 1 year after AF, based on fundus pathological manifestations generally prior to symptoms of the brain) were recruited. Fundus images at 548, 605, and 810 nm wavelengths were collected. Three classical deep neural network (DNN) models (Inception V3, ResNet50, SE50) were trained. Sociodemographic and selected routine clinical data were obtained.

Results

The accuracy of all DNNs with the single-spectral or multi-spectral combination images at the three wavelengths as input reached above 78%. The IS detection performance of DNNs with 605 nm spectral images as input was relatively more stable than with the other wavelengths. The multi-spectral combination models acquired a higher area under the curve (AUC) scores than the single-spectral models.

Conclusions

The probability of IS secondary to AF could be predicted based on multi-spectrum fundus images using deep learning, and combinations of multi-spectrum images improved the performance of DNNs. Acquiring different spectral fundus images is advantageous for the early prevention of cardiovascular and cerebrovascular diseases. The method in this study is a beneficial preliminary and initiative exploration for diseases that are difficult to predict the onset time such as IS.

Prediction of subjective cognitive decline after corpus callosum infarction by an interpretable machine learning-derived early warning strategy

Background and purpose

Corpus callosum (CC) infarction is an extremely rare subtype of cerebral ischemic stroke, however, the symptoms of cognitive impairment often fail to attract early attention of patients, which seriously affects the long-term prognosis, such as high mortality, personality changes, mood disorders, psychotic reactions, financial burden and so on. This study seeks to develop and validate models for early predicting the risk of subjective cognitive decline (SCD) after CC infarction by machine learning (ML) algorithms.

Methods

This is a prospective study that enrolled 213 (only 3.7%) CC infarction patients from a nine-year cohort comprising 8,555 patients with acute ischemic stroke. Telephone follow-up surveys were carried out for the patients with definite diagnosis of CC infarction one-year after disease onset, and SCD was identified by Behavioral Risk Factor Surveillance System (BRFSS) questionnaire. Based on the significant features selected by the least absolute shrinkage and selection operator (LASSO), seven ML models including Extreme Gradient Boosting (XGBoost), Logistic Regression (LR), Light Gradient Boosting Machine (LightGBM), Adaptive Boosting (AdaBoost), Gaussian Naïve Bayes (GNB), Complement Naïve Bayes (CNB), and Support vector machine (SVM) were established and their predictive performances were compared by different metrics. Importantly, the SHapley Additive exPlanations (SHAP) was also utilized to examine internal behavior of the highest-performance ML classifier.

Results

The Logistic Regression (LR)-model performed better than other six ML-models in SCD predictability after the CC infarction, with the area under the receiver characteristic operator curve (AUC) of 77.1% in the validation set. Using LASSO and SHAP analysis, we found that infarction subregions of CC infarction, female, 3-month modified Rankin Scale (mRS) score, age, homocysteine, location of angiostenosis, neutrophil to lymphocyte ratio, pure CC infarction, and number of angiostenosis were the top-nine significant predictors in the order of importance for the output of LR-model. Meanwhile, we identified that infarction subregion of CC, female, 3-month mRS score and pure CC infarction were the factors which independently associated with the cognitive outcome.

Conclusion

Our study firstly demonstrated that the LR-model with 9 common variables has the best-performance to predict the risk of post-stroke SCD due to CC infarcton. Particularly, the combination of LR-model and SHAP-explainer could aid in achieving personalized risk prediction and be served as a decision-making tool for early intervention since its poor long-term outcome.

Differences of survival benefits brought by various treatments in ovarian cancer patients with different tumor stages

PURPOSE

The current study aimed to explore the prognosis of ovarian cancer patients in different subgroup using three prognostic research indexes. The current study aimed to build a prognostic model for ovarian cancer patients.

METHODS

The study dataset was downloaded from Surveillance Epidemiology and End Results database. Accelerated Failure Time algorithm was used to construct a prognostic model for ovary cancer.

RESULTS

The mortality rate in the model group was 51.6% (9,314/18,056), while the mortality rate in the validation group was 52.1% (6,358/12,199). The current study constructed a prognostic model for ovarian cancer patients. The C indexes were 0.741 (95% confidence interval: 0.731-0.751) in model dataset and 0.738 (95% confidence interval: 0.726-0.750) in validation dataset. Brier score was 0.179 for model dataset and validation dataset. The C indexes were 0.741 (95% confidence interval: 0.733-0.749) in bootstrap internal validation dataset. Brier score was 0.178 for bootstrap internal validation dataset.

CONCLUSION

The current research indicated that there were significant differences in the survival benefits of treatments among ovarian cancer patients with different stages. The current research developed an individual mortality risk predictive system that could provide valuable predictive information for ovarian cancer patients.

Diagnostic accuracy of autoverification and guidance system for COVID-19 RT-PCR results

BACKGROUND

To date, most countries worldwide have declared that the pandemic of COVID-19 is over, while the WHO has not officially ended the COVID-19 pandemic, and China still insists on the personalized dynamic COVID-free policy. Large-scale nucleic acid testing in Chinese communities and the manual interpretation for SARS-CoV-2 nucleic acid detection results pose a huge challenge for labour, quality and turnaround time (TAT) requirements. To solve this specific issue while increase the efficiency and accuracy of interpretation, we created an autoverification and guidance system (AGS) that can automatically interpret and report the COVID-19 reverse transcriptase-polymerase chain reaction (RT-PCR) results relaying on computer-based autoverification procedure and then validated its performance in real-world environments. This would be conductive to transmission risk prediction, COVID-19 prevention and control and timely medical treatment for positive patients in the context of the predictive, preventive and personalized medicine (PPPM).

METHODS

A diagnostic accuracy test was conducted with 380,693 participants from two COVID-19 test sites in China, the Hong Kong Hybribio Medical Laboratory (n = 266,035) and the mobile medical shelter at a Shanghai airport (n = 114,658). These participants underwent SARS-CoV-2 RT-PCR from March 28 to April 10, 2022. All RT-PCR results were interpreted by laboratorians and by using AGS simultaneously. Considering the manual interpretation as gold standard, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were applied to evaluate the diagnostic value of the AGS on the interpretation of RT-PCR results.

RESULTS

Among the 266,035 samples in Hong Kong, there were 16,356 (6.15%) positive, 231,073 (86.86%) negative, 18,606 (6.99%) indefinite, 231,073 (86.86%, negative) no retest required and 34,962 (13.14%, positive and indefinite) retest required; the 114,658 samples in Shanghai consisted of 76 (0.07%) positive, 109,956 (95.90%) negative, 4626 (4.03%) indefinite, 109,956 (95.90%, negative) no retest required and 4702 (4.10%, positive and indefinite) retest required. Compared to the fashioned manual interpretation, the AGS is a procedure of high accuracy [99.96% (95%CI, 99.95-99.97%) in Hong Kong and 100% (95%CI, 100-100%) in Shanghai] with perfect sensitivity [99.98% (95%CI, 99.97-99.98%) in Hong Kong and 100% (95%CI, 100-100%) in Shanghai], specificity [99.87% (95%CI, 99.82-99.90%) in Hong Kong and 100% (95%CI, 99.92-100%) in Shanghai], PPV [99.98% (95%CI, 99.97-99.99%) in Hong Kong and 100% (95%CI, 99.99-100%) in Shanghai] and NPV [99.85% (95%CI, 99.80-99.88%) in Hong Kong and 100% (95%CI, 99.90-100%) in Shanghai]. The need for manual interpretation of total samples was dramatically reduced from 100% to 13.1% and the interpretation time fell from 53 h to 26 min in Hong Kong; while the manual interpretation of total samples was decreased from 100% to 4.1% and the interpretation time dropped from 20 h to 16 min at Shanghai.

CONCLUSIONS

The AGS is a procedure of high accuracy and significantly relieves both labour and time from the challenge of large-scale screening of SARS-CoV-2 using RT-PCR. It should be recommended as a powerful screening, diagnostic and predictive system for SARS-CoV-2 to contribute timely the ending of the COVID-19 pandemic following the concept of PPPM.

Thirty-six months recurrence after acute ischemic stroke among patients with comorbid type 2 diabetes: A nested case-control study

Background

Stroke patients have to face a high risk of recurrence, especially for those with comorbid T2DM, which usually lead to much more serious neurologic damage and an increased likelihood of death. This study aimed to explore determinants of stroke relapse among patients with comorbid T2DM.

Materials and methods

We conducted this case-control study nested a prospective cohort of ischemic stroke (IS) with comorbid T2DM. During 36-month follow-up, the second stroke occurred in 84 diabetic IS patients who were allocated into the case group, while 613 patients without recurrence were the controls. We collected the demographic data, behaviors and habits, therapies, and family history at baseline, and measured the variables during follow-up. LASSO and Logistic regression analyses were carried out to develop a prediction model of stroke recurrence. The receiver operator characteristic (ROC) curve was employed to evaluate the performance of the prediction model.

Results

Compared to participants without recurrence, the higher levels of pulse rate (78.29 ± 12.79 vs. 74.88 ± 10.93) and hypertension (72.6 vs. 61.2%) were recorded at baseline. Moreover, a lower level of physical activity (77.4 vs. 90.4%), as well as a higher proportion of hypoglycemic therapy (36.9 vs. 23.3%) was also observed during 36-month follow-up. Multivariate logistic regression revealed that higher pulse rate at admission (OR = 1.027, 95 %CI = 1.005–1.049), lacking physical activity (OR = 2.838, 95% CI = 1.418–5.620) and not receiving hypoglycemic therapy (OR = 1.697, 95% CI = 1.013–2.843) during follow-up increased the risk of stroke recurrence. We developed a prediction model using baseline pulse rate, hypoglycemic therapy, and physical activity, which produced an area under ROC curve (AUC) of 0.689.

Conclusion

Physical activity and hypoglycemic therapy play a protective role for IS patients with comorbid diabetes. In addition to targeted therapeutics, the improvement of daily-life habit contributes to slowing the progress of the IS.

Mutual effect of homocysteine and uric acid on arterial stiffness and cardiovascular risk in the context of predictive, preventive, and personalized medicine

Background

Arterial stiffness is a major risk factor and effective predictor of cardiovascular diseases and a common pathway of pathological vascular impairments. Homocysteine (Hcy) and uric acid (UA) own the shared metabolic pathways to affect vascular function. Serum uric acid (UA) has a great impact on arterial stiffness and cardiovascular risk, while the mutual effect with Hcy remains unknown yet. This study aimed to evaluate the mutual effect of serum Hcy and UA on arterial stiffness and 10-year cardiovascular risk in the general population. From the perspective of predictive, preventive, and personalized medicine (PPPM/3PM), we assumed that combined assessment of Hcy and UA provides a better tool for targeted prevention and personalized intervention of cardiovascular diseases via suppressing arterial stiffness.

Methods

This study consisted of 17,697 participants from Beijing Health Management Cohort, who underwent health examination between January 2012 and December 2019. Brachial-ankle pulse wave velocity (baPWV) was used as an index of arterial stiffness.

Results

Individuals with both high Hcy and UA had the highest baPWV, compared with those with low Hcy and low UA (β: 30.76, 95% CI: 18.36-43.16 in males; β: 53.53, 95% CI: 38.46-68.60 in females). In addition, these individuals owned the highest 10-year cardiovascular risk (OR: 1.49, 95% CI: 1.26-1.76 in males; OR: 7.61, 95% CI: 4.63-12.68 in females). Of note, males with high homocysteine and low uric acid were significantly associated with increased cardiovascular risk (OR: 1.30, 95% CI: 1.15-1.47), but not the high uric acid and low homocysteine group (OR: 1.02, 95% CI: 0.90-1.16).

Conclusions

This study found the significantly mutual effect of Hcy and UA on arterial stiffness and cardiovascular risk using a large population and suggested the clinical importance of combined evaluation and control of Hcy and UA for promoting cardiovascular health. The adverse effect of homocysteine on arteriosclerosis should be addressed beyond uric acid, especially for males. Monitoring of the level of both Hcy and UA provides a window opportunity for PPPM/3PM in the progression of arterial stiffness and prevention of CVD. Hcy provides a novel predictor beyond UA of cardiovascular health to identify individuals at high risk of arterial stiffness for the primary prevention and early treatment of CVD. In the progressive stage of arterial stiffness, active control of Hcy and UA levels from the aspects of dietary behavior and medication treatment is conducive to alleviating the level of arterial stiffness and reducing the risk of CVD. Further studies are needed to evaluate the clinical effect of Hcy and UA targeted intervention on arterial stiffness and cardiovascular health.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-022-00298-x.

Oculomics for sarcopenia prediction: a machine learning approach toward predictive, preventive, and personalized medicine

Aims

Sarcopenia is characterized by a gradual loss of skeletal muscle mass and strength with increased adverse outcomes. Recently, large-scale epidemiological studies have demonstrated a relationship between several chronic disorders and ocular pathological conditions using an oculomics approach. We hypothesized that sarcopenia can be predicted through eye examinations, without invasive tests or radiologic evaluations in the context of predictive, preventive, and personalized medicine (PPPM/3PM).

Methods

We analyzed data from the Korean National Health and Nutrition Examination Survey (KNHANES). The training set (80%, randomly selected from 2008 to 2010) data were used to construct the machine learning models. Internal (20%, randomly selected from 2008 to 2010) and external (from the KNHANES 2011) validation sets were used to assess the ability to predict sarcopenia. We included 8092 participants in the final dataset. Machine learning models (XGBoost) were trained on ophthalmological examinations and demographic factors to detect sarcopenia.

Results

In the exploratory analysis, decreased levator function (odds ratio [OR], 1.41; P value <0.001), cataracts (OR, 1.31; P value = 0.013), and age-related macular degeneration (OR, 1.38; P value = 0.026) were associated with an increased risk of sarcopenia in men. In women, an increased risk of sarcopenia was associated with blepharoptosis (OR, 1.23; P value = 0.038) and cataracts (OR, 1.29; P value = 0.010). The XGBoost technique showed areas under the receiver operating characteristic curves (AUCs) of 0.746 and 0.762 in men and women, respectively. The external validation achieved AUCs of 0.751 and 0.785 for men and women, respectively. For practical and fast hands-on experience with the predictive model for practitioners who may be willing to test the whole idea of sarcopenia prediction based on oculomics data, we developed a simple web-based calculator application (https://knhanesoculomics.github.io/sarcopenia) to predict the risk of sarcopenia and facilitate screening, based on the model established in this study.

Conclusion

Sarcopenia is treatable before the vicious cycle of sarcopenia-related deterioration begins. Therefore, early identification of individuals at a high risk of sarcopenia is essential in the context of PPPM. Our oculomics-based approach provides an effective strategy for sarcopenia prediction. The proposed method shows promise in significantly increasing the number of patients diagnosed with sarcopenia, potentially facilitating earlier intervention. Through patient oculometric monitoring, various pathological factors related to sarcopenia can be simultaneously analyzed, and doctors can provide personalized medical services according to each cause. Further studies are needed to confirm whether such a prediction algorithm can be used in real-world clinical settings to improve the diagnosis of sarcopenia.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-022-00292-3.