Scientists at Osaka Metropolitan University have developed an innovative AI model that accurately classifies cardiac functions and identifies Valvular Heart Diseases using chest radiographs. This breakthrough research could enhance traditional diagnostic methods like echocardiography, improve diagnostic efficiency, and be particularly beneficial in settings with limited access to specialized technicians.

Valvular heart disease is a leading cause of heart failure, typically diagnosed using echocardiography. However, the shortage of qualified technicians with specialized skills in this technique poses a challenge. On the other hand, chest radiography, or chest X-rays, is a common diagnostic test used to detect lung diseases. While the heart is also visible on chest radiographs, its ability to identify cardiac function or disease was not well-known until now.

Conducted in many hospitals, chest radiographs are accessible and reproducible, requiring minimal time. Dr. Daiju Ueda and his team from Osaka Metropolitan University’s Department of Diagnostic and Interventional Radiology saw the potential for chest radiographs to supplement echocardiography if they could accurately determine cardiac function and disease.

The research team successfully developed an AI model that utilizes machine learning to classify cardiac functions and valvular heart diseases accurately. To avoid potential bias and increase accuracy, the model was trained on a diverse dataset from multiple institutions. They collected 22,551 chest radiographs and associated echocardiograms from 16,946 patients at four facilities between 2013 and 2021. By training the AI model on these datasets, they were able to establish connections between the features extracted from the chest radiographs and the corresponding echocardiograms.

The AI model demonstrated precise classification of six selected types of valvular heart disease, with an Area Under the Curve (AUC) ranging from 0.83 to 0.92. The AUC is a rating index that measures the performance of an AI model, with values closer to 1 indicating better performance. Notably, the AI model achieved an AUC of 0.92 for detecting left ventricular ejection fraction, a crucial measure for monitoring cardiac function.

Dr. Ueda emphasized the significance of this research: “In addition to improving doctors’ diagnostic efficiency, this system could be used in areas without specialists, during nighttime emergencies, and for patients who have difficulty undergoing echocardiography.”

This groundbreaking AI model has the potential to revolutionize cardiac diagnostics by providing accurate and efficient assessments of cardiac functions and valvular heart diseases using readily available chest radiographs.

Reference:
– Ueda, D., Matsumoto, T., Ehara, S., Yamamoto, A., Walston, S.L., Ito, A., Shimono, T., Shiba, M., Takeshita, T., Fukuda, D., Miki, Y. “Artificial intelligence-based model to classify cardiac functions from chest radiographs: a multi-institutional, retrospective model development and validation study.” The Lancet Digital Health. DOI: 10.1016/S2589-7500(23)00107-3

The post New AI Model Uses Chest Radiographs to Identify Cardiac Functions and Valvular Heart Diseases appeared first on TS2 SPACE.