HomeNewsRadiomics extracted from dual-energy computed tomography and integrated in a trained machine-learning classifier shows promise in diagnosing and predicting acute pulmonary embolism

Radiomics extracted from dual-energy computed tomography and integrated in a trained machine-learning classifier shows promise in diagnosing and predicting acute pulmonary embolism

    A single-center retrospective study analyzed the clinical and imaging data of 131 patients who underwent contrast-enhanced pulmonary artery dual-energy CT (DECT) angiography for suspected pulmonary embolism (PE). A total of 107 radiomic features were extracted, selected, and integrated into a trained machine-learning classifier. This innovative system achieved 94% accuracy in identifying patients with acute PE and also demonstrated a strong predictive capability for patient outcomes. These findings suggest the potential utility of this approach in optimizing the diagnosis and prognosis of PE in clinical practice.