Obstructive Sleep Apnea: A Prediction Model Using Supervised Machine Learning Method.

Stud Health Technol Inform

Department of Community Medicine, School of Medicine, Iran University of Medical Sciences, Tehran, Iran.

Published: June 2020

AI Article Synopsis

  • Obstructive Sleep Apnea (OSA) is a common sleep disorder that can lead to various health issues, prompting researchers to explore machine learning for predicting its occurrence.
  • The study utilized a dataset of 231 records, applying the CRISP-DM methodology to preprocess the data and implement popular machine learning algorithms, ultimately identifying the Naïve Bayes and Logistic Regression classifiers as the most effective models.
  • With the ability to screen high-risk individuals, these machine learning methods showed strong performance, featuring a sensitivity of 93.42% for SVM and a specificity of 59.49% for Naïve Bayes, making them valuable tools for physicians.

Article Abstract

Obstructive Sleep Apnea (OSA) is the most common breathing-related sleep disorder, leading to increased risk of health problems. In this study, we investigated and evaluated the supervised machine learning methods to predict OSA. We used popular machine learning algorithms to develop the prediction models, using a dataset with non-invasive features containing 231 records. Based on the methodology, the CRISP-DM, the dataset was checked and the blanked data were replaced with average/most frequented items. Then, the popular machine learning algorithms were applied for modeling and the 10-fold cross-validation method was used for performance comparison purposes. The dataset has 231 records, of which 152 (65.8%) were diagnosed with OSA. The majority was male (143, 61.9%). The results showed that the best prediction model with an overall AUC reached the Naïve Bayes and Logistic Regression classifier with 0.768 and 0.761, respectively. The SVM with 93.42% sensitivity and the Naïve Bayes of 59.49% specificity can be suitable for screening high-risk people with OSA. The machine learning methods with easily available features had adequate power of discrimination, and physicians can screen high-risk OSA as a supplementary tool.

Download full-text PDF

Source
http://dx.doi.org/10.3233/SHTI200576DOI Listing

Publication Analysis

Top Keywords

machine learning
20
obstructive sleep
8
sleep apnea
8
prediction model
8
supervised machine
8
learning methods
8
popular machine
8
learning algorithms
8
231 records
8
naïve bayes
8

Similar Publications

The study of somatic mutations in single cells provides insights into aging and carcinogenesis, which is complicated by the dependency on whole-genome amplification (WGA). Here, we describe a detailed workflow starting from single-cell isolation to WGA by primary template-directed amplification (PTA), sequencing, quality control, and downstream analyses. A machine learning approach, the PTA Analysis Toolkit (PTATO), is used to filter the hundreds to thousands of artificial variants induced by WGA from true mutations at high sensitivity and accuracy.

View Article and Find Full Text PDF

The gut microbiota is a crucial link between diet and cardiovascular disease (CVD). Using fecal metaproteomics, a method that concurrently captures human gut and microbiome proteins, we determined the crosstalk between gut microbiome, diet, gut health, and CVD. Traditional CVD risk factors (age, BMI, sex, blood pressure) explained < 10% of the proteome variance.

View Article and Find Full Text PDF

Background: Atherosclerosis (AS) is a major cause of cardiovascular diseases and neutrophil extracellular traps (NETs) may be actively involved in the development of atherosclerosis. Identifying key biomarkers in this process is essential for developing targeted treatments for AS.

Methods: We performed bioinformatics analysis using a NETosis-related gene (NRGs) set and three AS datasets (GSE100927, GSE21545, and GSE159677).

View Article and Find Full Text PDF

Predicting preterm birth using electronic medical records from multiple prenatal visits.

BMC Pregnancy Childbirth

December 2024

Department of Mathematics and Computer Science, Eindhoven University of Technology, Eindhoven, 5612 AZ, North Brabant, The Netherlands.

This study aimed to predict preterm birth in nulliparous women using machine learning and easily accessible variables from prenatal visits. Elastic net regularized logistic regression models were developed and evaluated using 5-fold cross-validation on data from 8,830 women in the Nulliparous Pregnancy Outcomes Study: New Mothers-to-Be (nuMoM2b) dataset at three prenatal visits: - , - , and - weeks of gestational age (GA). The models' performance, assessed using Area Under the Curve (AUC), sensitivity, specificity, and accuracy, consistently improved with the incorporation of data from later prenatal visits.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!