Artificial intelligence for dysphagia screening: A machine learning approach
6 affiliations
- 1Graduate Program in Dentistry, Faculty of Dentistry, Federal University of Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil
- 2Graduate Program in Information Technology and Healthcare Management, Federal University of Health Sciences of Porto Alegre, Porto Alegre, Rio Grande do Sul, Brazil
- 3Department of Speech-Language Pathology, Federal University of Health Sciences of Porto Alegre, Porto Alegre, Rio Grande do Sul, Brazil
- 4Department of Exact and Applied Social Sciences, Federal University of Health Sciences of Porto Alegre, Porto Alegre, Rio Grande do Sul, Brazil
- 5Department of Preventive and Social Dentistry, Faculty of Dentistry, Federal University of Rio Grande do Sul, Porto Alegre, Rio Grande do Sul, Brazil
- 6Department of Epidemiology and Health Promotion, College of Dentistry, New York University, New York, New York, United States of America
Abstract
Early and reliable screening for oropharyngeal dysphagia symptoms in at-risk populations is a crucial first stage in effective management. This study aims to propose a machine learning model that can predict whether individuals are at risk of having confirmed dysphagia based on diagnostic accuracy. This was an exploratory cross-sectional diagnostic accuracy study. Adults and older adults were screened and assessed, and a videofluoroscopy swallowing study was performed to confirm dysphagia in participants with complaints about swallowing and signs or symptoms identified in the clinical assessment by a speech therapist. Eight algorithms were tested based on a theoretical model to evaluate their diagnostic accuracy: Naive Bayes, K-Nearest Neighbors, Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, Multilayer Perceptron, and Convolutional Neural Networks. The Pandas library in Python was used for data visualization and manipulation. In total, 465 participants were included, of whom 153 had dysphagia. A screening with 15 items related to personal, health, and oral health characteristics is proposed. The final four best-performing models were Naive Bayes (accuracy 0.76, sensitivity 0.84), Logistic Regression (accuracy 0.79, sensitivity 0.82), Support Vector Machine (accuracy 0.78, sensitivity 0.79), and Convolutional Neural Network (accuracy 0.78, sensitivity 0.74). The use of artificial intelligence based on different machine-learning approaches is promising for dysphagia screening, demonstrating adequate diagnostic accuracy. Future studies should be conducted to externally validate the proposed screening models. Author summary: Early identification of oropharyngeal dysphagia is critical for clinical management. This study developed and tested machine learning models to predict dysphagia using clinical screening data validated by videofluoroscopy (n = 465). A 15-item screening model achieved accuracies of 0.76–0.79 and sensitivities of 0.74–0.84 across top-performing algorithms. Artificial intelligence shows promise for dysphagia screening, warranting further external validation.
Related
- Evaluation of an Artificial Intelligence and Machine Learning Algorithm to Enhance Angiographic Coronary Visibility in Morbid Obesity: A Retrospective Observational Study
- Hospital market concentration and hospital adoption of artificial intelligence and machine learning.
- Artificial intelligence and machine learning for prediction of postoperative complications: A systematic review focused on anesthesiology, perioperative risk stratification, and clinical applicability.
- Machine learning models based on artificial intelligence-derived CT quantitative parameters and CT semantic features for predicting the invasive risk of subsolid pulmonary nodules: a dual-center study
- Artificial intelligence and machine learning for organizational and system-level functions in primary health care: A scoping review protocol
BibTeX
@article{rech2026artificial,
title = {{Artificial intelligence for dysphagia screening: A machine learning approach}},
author = {Rafaela Soares Rech and Carla Diniz Lopes Becker and Dieine Estela Bernieri Schiavon and Danielli Pires Vieira and Juliana Balbinot Hilgert and Fernando Neves Hugo},
journal = {PLOS Digital Health},
year = {2026},
month = sep,
doi = {10.1371/journal.pdig.0001755},
url = {https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001755},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Artificial intelligence for dysphagia screening: A machine learning approach
AU - Rafaela Soares Rech
AU - Carla Diniz Lopes Becker
AU - Dieine Estela Bernieri Schiavon
AU - Danielli Pires Vieira
AU - Juliana Balbinot Hilgert
AU - Fernando Neves Hugo
PY - 2026
DA - 2026/09/29
JO - PLOS Digital Health
DO - 10.1371/journal.pdig.0001755
UR - https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0001755
AB - Early and reliable screening for oropharyngeal dysphagia symptoms in at-risk populations is a crucial first stage in effective management. This study aims to propose a machine learning model that can predict whether individuals are at risk of having confirmed dysphagia based on diagnostic accuracy. This was an exploratory cross-sectional diagnostic accuracy study. Adults and older adults were screened and assessed, and a videofluoroscopy swallowing study was performed to confirm dysphagia in participants with complaints about swallowing and signs or symptoms identified in the clinical assessment by a speech therapist. Eight algorithms were tested based on a theoretical model to evaluate their diagnostic accuracy: Naive Bayes, K-Nearest Neighbors, Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, Multilayer Perceptron, and Convolutional Neural Networks. The Pandas library in Python was used for data visualization and manipulation. In total, 465 participants were included, of whom 153 had dysphagia. A screening with 15 items related to personal, health, and oral health characteristics is proposed. The final four best-performing models were Naive Bayes (accuracy 0.76, sensitivity 0.84), Logistic Regression (accuracy 0.79, sensitivity 0.82), Support Vector Machine (accuracy 0.78, sensitivity 0.79), and Convolutional Neural Network (accuracy 0.78, sensitivity 0.74). The use of artificial intelligence based on different machine-learning approaches is promising for dysphagia screening, demonstrating adequate diagnostic accuracy. Future studies should be conducted to externally validate the proposed screening models. Author summary: Early identification of oropharyngeal dysphagia is critical for clinical management. This study developed and tested machine learning models to predict dysphagia using clinical screening data validated by videofluoroscopy (n = 465). A 15-item screening model achieved accuracies of 0.76–0.79 and sensitivities of 0.74–0.84 across top-performing algorithms. Artificial intelligence shows promise for dysphagia screening, warranting further external validation.
SN - 2767-3170
ER -