A team from the University of East Anglia, the University of Sheffield, and the University of Leeds in the UK has jointly developed an intelligent computer model that can use artificial intelligence (AI) to analyze the results of cardiac magnetic resonance imaging (MRI) examinations. This automation technology not only provides patients with fast and reliable heart health assessments, but also significantly saves time and resources. The paper was published in the recently published European Journal of Experimental Radiology. Researchers say that AI models can accurately measure the size and function of heart chambers, and their results are comparable to those analyzed by doctors, but much faster. Standard manual MRI analysis may take 45 minutes or longer, while the new AI model can complete the analysis in just a few seconds. Researchers trained AI models using retrospective observational data from 814 patients. To ensure the accuracy of the model results, the research team subsequently used scans and data from an additional 101 patients for testing. Although other studies have explored the application of AI in interpreting MRI scans, the latest AI models have been trained using data from multiple hospitals and different types of scanners, and tested on different patient groups from different hospitals. In addition, AI models achieve a complete analysis of the entire heart by providing a view that displays all four chambers, whereas most early studies typically focused only on examining the two main chambers of the heart. The process of automating the assessment of cardiac function and structure will greatly save time and resources, and ensure that doctors obtain consistent results. This innovation is expected to bring more efficient diagnosis, better treatment decisions, and ultimately improve outcomes for heart disease patients. In addition, the potential of AI to predict mortality based on cardiac scan results also indicates its enormous prospects in fundamentally changing the field of cardiac care and improving patient outcomes. Researchers suggest that future studies should expand the sample size, include more patients from different hospitals, use various types of MRI scanners to test the model, and consider other common diseases that may be encountered in medical practice to validate its effectiveness in a wider range of real-world situations. (New Society)