Predicting the Risk of HIV Infection and Sexually Transmitted Diseases Among Men Who Have Sex With Men: Cross-Sectional Study Using Multiple Machine Learning Approaches (Preprint) DOI
Bing Lin, Jiaxiu Liu, K. Li

и другие.

Опубликована: Апрель 2, 2024

BACKGROUND Men who have sex with men (MSM) are at high risk for HIV infection and sexually transmitted diseases (STDs). However, there is a lack of accurate convenient tools to assess this risk. OBJECTIVE This study aimed develop machine learning models predict the STDs among MSM. METHODS We conducted cross-sectional that collected individual characteristics 1999 MSM negative or unknown serostatus in Western China from 2013 2023. self-reported their STD history were tested HIV. compared accuracy 6 methods predicting using 7 parameters comprehensive assessment, ranking according performance each parameter. selected data Sichuan external validation. RESULTS Of MSM, 72 (3.6%) positive 146 (7.3%) previous infection. After taking results intersection 3 feature screening methods, total 5 predictors screened STDs, respectively, multiple prediction constructed. Extreme gradient boost performed optimally area under curve values 0.777 (95% CI 0.639-0.915) 0.637 0.541-0.732), demonstrating stable both internal The highest combined predictive scores 33 39, respectively. Interpretability analysis showed nonadherence condom use, low knowledge, male partners, internet dating factors Low degree education, dating, female partners STDs. stratification optimal model effectively distinguished between high- low-risk classified into (predicted score &lt;0.506 ≥0.506) &lt;0.479 ≥0.479) groups. In total, 22.8% (114/500) high-risk group, 43% (215/500) group. significantly higher groups (<i>P</i>&lt;.001 <i>P</i>=.05, respectively), predicted probabilities both). displayed web applications probability estimation interactive computation. CONCLUSIONS Machine demonstrated strengths Risk can facilitate clinicians accurately assessing individuals risk, especially concealed behaviors, help them self-monitor targeted, timely diagnosis interventions reduce new infections.

Язык: Английский

Predicting the Risk of HIV Infection and Sexually Transmitted Diseases Among Men Who Have Sex With Men: Cross-Sectional Study Using Multiple Machine Learning Approaches DOI Creative Commons
Bing Lin, Jiaxiu Liu, K. Li

и другие.

Journal of Medical Internet Research, Год журнала: 2025, Номер 27, С. e59101 - e59101

Опубликована: Фев. 20, 2025

Men who have sex with men (MSM) are at high risk for HIV infection and sexually transmitted diseases (STDs). However, there is a lack of accurate convenient tools to assess this risk. This study aimed develop machine learning models predict the STDs among MSM. We conducted cross-sectional that collected individual characteristics 1999 MSM negative or unknown serostatus in Western China from 2013 2023. self-reported their STD history were tested HIV. compared accuracy 6 methods predicting using 7 parameters comprehensive assessment, ranking according performance each parameter. selected data Sichuan external validation. Of MSM, 72 (3.6%) positive 146 (7.3%) previous infection. After taking results intersection 3 feature screening methods, total 5 predictors screened STDs, respectively, multiple prediction constructed. Extreme gradient boost performed optimally area under curve values 0.777 (95% CI 0.639-0.915) 0.637 0.541-0.732), demonstrating stable both internal The highest combined predictive scores 33 39, respectively. Interpretability analysis showed nonadherence condom use, low knowledge, male partners, internet dating factors Low degree education, dating, female partners STDs. stratification optimal model effectively distinguished between high- low-risk classified into (predicted score <0.506 ≥0.506) <0.479 ≥0.479) groups. In total, 22.8% (114/500) high-risk group, 43% (215/500) group. significantly higher groups (P<.001 P=.05, respectively), predicted probabilities both). displayed web applications probability estimation interactive computation. Machine demonstrated strengths Risk can facilitate clinicians accurately assessing individuals risk, especially concealed behaviors, help them self-monitor targeted, timely diagnosis interventions reduce new infections.

Язык: Английский

Процитировано

1
Artificial intelligence in stroke risk assessment and management via retinal imaging DOI Creative Commons

Parsa Khalafi,

Soroush Morsali, Sana Hamidi

и другие.

Frontiers in Computational Neuroscience, Год журнала: 2025, Номер 19

Опубликована: Фев. 17, 2025

Retinal imaging, used for assessing stroke-related retinal changes, is a non-invasive and cost-effective method that can be enhanced by machine learning deep algorithms, showing promise in early disease detection, severity grading, prognostic evaluation stroke patients. This review explores the role of artificial intelligence (AI) patient care, focusing on imaging integration into clinical workflows. has revealed several microvascular including decrease central artery diameter an increase vein diameter, both which are associated with lacunar intracranial hemorrhage. Additionally, such as arteriovenous nicking, increased vessel tortuosity, arteriolar light reflex, decreased fractals, thinning nerve fiber layer also reported to higher risk. AI models, Xception EfficientNet, have demonstrated accuracy comparable traditional risk scoring systems predicting For diagnosis, models like Inception, ResNet, VGG, alongside classifiers, shown high efficacy distinguishing patients from healthy individuals using imaging. Moreover, random forest model effectively distinguished between ischemic hemorrhagic subtypes based features, superior predictive performance compared characteristics. support vector achieved classification pial collateral status. Despite this advancements, challenges lack standardized protocols modalities, hesitance trusting AI-generated predictions, insufficient data electronic health records, need validation across diverse populations, ethical regulatory concerns persist. Future efforts must focus validating ensuring algorithm transparency, addressing issues enable broader implementation. Overcoming these barriers will essential translating technology personalized care improving outcomes.

Язык: Английский

Процитировано

0
Dysregulated brain dynamics in the visual motor network in type 2 diabetes patients and their relationship with cognitive impairment DOI Creative Commons
Ying Yu, Bo Hu, Xiaofeng Yu

и другие.

Brain Research Bulletin, Год журнала: 2025, Номер unknown, С. 111313 - 111313

Опубликована: Март 1, 2025

Язык: Английский

Процитировано

0
Development and Validation of a Predictive Model for Intracranial Haemorrhage in Patients on Direct Oral Anticoagulants DOI Creative Commons
Yuanyuan Liu,

Linjie Li,

Jingge Li

и другие.

Clinical and Applied Thrombosis/Hemostasis, Год журнала: 2024, Номер 30

Опубликована: Янв. 1, 2024

Background Intracranial haemorrhage (ICH) poses a significant threat to patients on Direct Oral Anticoagulants (DOACs), with existing risk scores inadequately predicting ICH in these patients. We aim develop and validate predictive model for DOAC-treated Methods 24,794 treated DOAC were identified province-wide electronic medical health data platform Tianjin, China. The cohort was randomly split into 4:1 ratio development validation. utilized forward stepwise selection, Least Absolute Shrinkage Selection Operator (LASSO), eXtreme Gradient Boosting (XGBoost) select predictors. Model performance compared using the area under curve (AUC) net reclassification index (NRI). optimal stratified model. Results median age is 68.0 years, 50.4% of participants are male. XGBoost model, incorporating six independent factors (history hemorrhagic stroke, peripheral artery disease, venous thromboembolism, hypertension, age, low-density lipoprotein cholesterol levels), demonstrated superior dateset. It showed moderate discrimination (AUC: 0.68, 95% CI: 0.64–0.73), outperforming (ΔAUC = 0.063, P 0.003; NRI 0.374, < 0.001). Risk categories significantly (low risk: 0.26%, 0.74%, high 5.51%). Finally, consistent internal Conclusion In real-world Chinese population therapy, this study presents reliable risk. integrating key factors, offers valuable tool individualized assessment context oral anticoagulation therapy.

Язык: Английский

Процитировано

2
Development and validation of a machine learning-based risk prediction model for post-stroke cognitive impairment DOI Creative Commons
Xia Zhong, Jing Li,

Shunxin Lv

и другие.

Research Square (Research Square), Год журнала: 2024, Номер unknown

Опубликована: Янв. 30, 2024

Abstract Background Machine learning (ML) risk prediction models for post-stroke cognitive impairment (PSCI) are still far from optimal. This study aims to generate a reliable predictive model predicting PSCI in Chinese individuals using ML algorithms. Methods We collected data on 494 who were diagnosed with acute ischemic stroke (AIS) and hospitalized this condition January 2022 November 2023 at medical institution. All of the observed samples divided into training set (70%) validation (30%) random. Logistic regression combined least absolute shrinkage selection operator (LASSO) was utilized efficiently screen optimal features PSCI. seven different (LR, XGBoost, LightGBM, AdaBoost, GNB, MLP, SVM) compared their performance resulting variables. used five-fold cross-validation measure model's area under curve (AUC), sensitivity, specificity, accuracy, F1 score PR values. SHAP analysis provides comprehensive detailed explanation our optimized performance. Results identified 58.50% eligible AIS patients. The most HAMD-24, FBG, age, PSQI, paraventricular lesion. XGBoost model, among 7 developed based best features, demonstrates superior performance, as indicated by its AUC (0.961), sensitivity (0.931), specificity (0.889), accuracy (0.911), (0.926), AP value (0.967). Conclusion lesion is exceptional It provide clinicians tool early screening patients effective treatment decisions

Язык: Английский

Процитировано

1
The role of biomarkers of endothelial dysfunction in predicting the progression of mild cognitive impairment in patients with cardiovascular risk factors DOI
Vorob'eva Ov, V V Fateeva

S S Korsakov Journal of Neurology and Psychiatry, Год журнала: 2024, Номер 124(9), С. 88 - 88

Опубликована: Янв. 1, 2024

To develop a predictive model to assess the risk of progression mild cognitive impairment (MCI) within 12 weeks in patients with cardiovascular factors using biomarkers endothelial dysfunction.

Язык: Английский

Процитировано

1
Machine learning–based predictive model for post-stroke dementia DOI Creative Commons
Zemin Wei, Mengqi Li, Chenghui Zhang

и другие.

BMC Medical Informatics and Decision Making, Год журнала: 2024, Номер 24(1)

Опубликована: Ноя. 11, 2024

Post-stroke dementia (PSD), a common complication, diminishes rehabilitation efficacy and affects disease prognosis in stroke patients. Many factors may be related to PSD, including demographic, comorbidities, examination characteristics. However, most existing methods are qualitative evaluations of independent factors, which ignore the interaction amongst various factors. Therefore, purpose this study is explore applicability machine learning (ML) for predicting PSD.

Язык: Английский

Процитировано

1
Machine Learning–Based Predictive Model for Post-Stroke Dementia DOI Creative Commons
Zemin Wei, Mengqi Li, Chenghui Zhang

и другие.

Research Square (Research Square), Год журнала: 2024, Номер unknown

Опубликована: Март 28, 2024

Abstract Backgound: Post-stroke dementia (PSD), a common complication, diminishes rehabilitation efficacy and affects disease prognosis in stroke patients. Many factors may be related to PSD, including demographic, comorbidities, examination characteristics. However, most existing methods are qualitative evaluations of independent factors, which ignore the interaction amongst various factors. Therefore, purpose this study is explore applicability machine learning for predicting PSD. Methods: 9 acceptable features were screened out by Spearman correlation analysis Boruta algorithm. We developed evaluated 8 (ML) models: logistic regression, elastic net, k-nearest neighbors, decision tree, extreme gradient boosting, support vector machine, random forest, multilayer perceptron. Results: A total 539 patients included study. Among models used predict boosting forest showed highest area under curve (AUC), with values 0.7287 0.7285, respectively. The important PSD age, high sensitivity C-reactive protein, side location, occurrence cerebral hemorrhage. Conclusion: Our findings suggest that ML models, especially can best risk

Язык: Английский

Процитировано

0
Machine learning‐based radiomics in neurodegenerative and cerebrovascular disease DOI Creative Commons

Meilong Shi,

Feng Xiong,

He Zhi

и другие.

MedComm, Год журнала: 2024, Номер 5(11)

Опубликована: Окт. 28, 2024

Abstract Cognitive impairments, which can be caused by neurodegenerative and cerebrovascular disease, represent a growing global health crisis with far‐reaching implications for individuals, families, healthcare systems, economies worldwide. Notably, neurodegenerative‐induced cognitive impairment often presents different pattern severity compared to cerebrovascular‐induced impairment. With the development of computational technology, machine learning techniques have developed rapidly, offers powerful tool in radiomic analysis, allowing more comprehensive model that handle high‐dimensional, multivariate data traditional approach. Such models allow prediction disease development, as well accurately classify from overlapping symptoms, therefore facilitating clinical decision making. This review will focus on application learning‐based radiomics neurogenerative disease. Within category, this primarily focuses Alzheimer's while also covering other conditions such Parkinson's Lewy body dementia, Huntington's In we concentrate poststroke impairment, including ischemic hemorrhagic stroke, additional attention given small vessel moyamoya We specific challenges limitations when applying radiomics, provide our suggestion overcome those towards end, discuss what could done future use.

Язык: Английский

Процитировано

0
Prediction of preterm birth using machine learning: a comprehensive analysis based on large-scale preschool children survey data in Shenzhen of China DOI Creative Commons

Liwen Ding,

Xiaona Yin,

Guo‐Min Wen

и другие.

BMC Pregnancy and Childbirth, Год журнала: 2024, Номер 24(1)

Опубликована: Дек. 4, 2024

Preterm birth (PTB) is a significant cause of neonatal mortality and long-term health issues. Accurate prediction timely prevention PTB are essential for reducing associated child morbidity. Traditional predictive methods face challenges due to heterogeneous risk factors their interaction effects. This study aims develop evaluate six machine learning (ML) models predict using large-scale children survey data from Shenzhen, China, identify key predictors through Shapley Additive Explanations (SHAP) analysis. Data 84,050 mother–child pairs, collected in 2021 2022, were processed divided into training, validation, test sets. Six ML tested: L1-Regularised Logistic Regression, Light Gradient Boosting Machine (LightGBM), Naive Bayes, Random Forests, Support Vector Machine, Extreme (XGBoost). Model performance was evaluated based on discrimination, calibration clinical utility. SHAP analysis used interpret the importance impact individual features prediction. The XGBoost model demonstrated best overall performance, with area under receiver operating characteristic curve (AUC) scores 0.752 0.757 validation sets, respectively, along favorable Key identified multiple pregnancies, threatened abortion, maternal age conception. highlighted positive impacts pregnancies as well negative micronutrient supplementation PTB. Our found that models, particularly XGBoost, show promise accurately predicting identifying factors. These findings provide potential enhancing interventions, personalizing prenatal care, informing public initiatives.

Язык: Английский

Процитировано

0