Asthma and COPD are common lung diseases, causing health issues and higher healthcare costs. A recent study reveals that deep learning, a type of AI using large data sets, can enhance the identification of patients at higher risk for multiple hospitalizations due to these conditions.
The study, published on Dec. 13, 2023, in Respiratory Research, analyzed electronic health records and compared different machine learning models with the deep learning method, multilayer perceptron, showing the most effective performance in predicting hospital readmissions.
According to Dr. Jose Gomez-Villalobos, the study indicates that AI can aid pulmonologists in creating new classifications for asthma, COPD, and other conditions. This can help identify patient groups benefiting from specific treatments, which is only sometimes apparent to clinicians. Knowing which patients need targeted therapies can decrease their chances of returning to the hospital. The strength of AI lies in its ability to enhance and transform clinical practices for better patient outcomes.
“The study revealed significant racial and ethnic disparities in disease exacerbations, affecting minority groups more. Using deep learning and AI in clinical practice can prioritize care for vulnerable individuals,” said Dr. Naftali Kaminski. Dr. Jose Gomez-Villalobos sees applying computational methods and AI to customize interventions, improve outcomes for high-risk patients, and reduce costly emergency room visits. The power of AI allows clinicians to transform practices for better outcomes for all patients.
The study highlights racial disparities in disease exacerbations. Dr. Kaminski emphasizes using AI to prioritize care for vulnerable individuals. Dr. Gomez-Villalobos envisions AI customizing interventions for high-risk patients, reducing costly ER visits. Overall, AI empowers clinicians to transform practices for better patient outcomes.
Journal reference:
- Lopez, K., Li, H., Lipkin-Moore, Z. et al. Deep learning prediction of hospital readmissions for asthma and COPD. Respiratory Research. DOI: 10.1186/s12931-023-02628-7.
Note: This article have been indexed to our site. We do not claim legitimacy, ownership or copyright of any of the content above. To see the article at original source Click Here