RETURN TO ABSTRACT LISTING


Objective:
Technological progress in the acquisition of medical images and the extraction of underlying quantitative imaging data has introduced exciting prospects for the diagnostic assessment of a wide range of conditions. This study aims to investigate the diagnostic utility of a machine learning classifier based on dual-energy CT (DECT) radiomics for classifying pulmonary embolism (PE) severity and assessing the risk for early death.

Materials and Methods:
Patients who underwent CT pulmonary angiogram (CTPA) between January 2015 and March 2022 were considered for inclusion in this retrospective study. An independent reading of two radiologists provided an independent reference standard for the presence or absence of PE. Risk stratification was performed according to current ESC guidelines. Based on DECT imaging, 107 radiomic features were extracted for each patient using standardized image processing. After dividing the dataset into training and test sets, stepwise feature reduction based on reproducibility, variable importance and correlation analyses were performed to select the most relevant features, which were then used to train and validate a gradient-boosted tree model.

Results:
The trained machine learning classifier achieved a classification accuracy of 0.94 for classifying high risk PE patients with an AUC of 0.86. This CT-based radiomics signature showed good diagnostic accuracy to distinguish between the different risk categories in patients with central PE, especially in the elevated risk classes.

Conclusion:
Models utilizing DECT-derived radiomics features can accurately stratify patients with pulmonary embolism into established clinical risk scores. This approach holds the potential to enhance patient management and optimize patient flow by assisting in the clinical decision-making process. It also offers the advantage of saving both time and resources by leveraging existing DECT imaging, eliminating the necessity for an additional scoring system.