AI-Powered Blood Test Distinguishes Deadly Cardiac Events

Yet the treatments for these conditions are fundamentally different, meaning that a misdiagnosis can have fatal consequences. Myocardial infarction occurs when a coronary artery becomes blocked and requires clot-dissolving medications to restore blood flow. Aortic dissection involves a tear in the aorta, where such medications are strictly contraindicated because they can cause severe bleeding.

Current diagnostic approaches rely heavily on contrast-enhanced CT imaging, which requires specialized equipment and can take significant time. These limitations make such tools difficult to use in ambulances or smaller medical facilities. Now, a rapid blood-based diagnostic method uses artificial intelligence (AI) and spectroscopy to distinguish between the two conditions within minutes.

The new diagnostic technique, which combines spectral blood analysis with deep-learning algorithms, to rapidly classify cardiovascular emergencies has been developed by a research team that included investigators from Xinjiang University (Urumqi, China). The method, presented in the journal Engineering Applications of Artificial Intelligence, analyzes a patient’s blood serum using Raman spectroscopy, which identifies molecular vibrations that reveal biochemical composition, and infrared spectroscopy, which detects molecular bonds and functional groups present in biological samples.

Together, these techniques capture subtle biochemical “fingerprints” left in the blood by different disease processes. Researchers then used a deep learning model to integrate the spectral data and automatically distinguish between acute myocardial infarction and aortic dissection. The diagnostic approach requires only five to ten minutes to analyze a blood sample and achieved 94.06% accuracy in differentiating the two conditions. This rapid identification is particularly important because treatments for the two diseases differ dramatically. The new spectral-AI technique could offer a portable and rapid diagnostic solution, enabling earlier and more accurate treatment decisions for patients presenting with acute chest pain.

The researchers have already developed a prototype diagnostic device, which is undergoing multicenter clinical validation. If successful, the portable system could be deployed in ambulances, emergency departments, and community clinics, where quick decision-making is critical. Such a tool could significantly reduce diagnostic delays and improve survival outcomes for patients experiencing life-threatening cardiovascular emergencies.

Related Links:
Xinjiang University