Acute pulmonary embolism (APE) is a cardiovascular disease characterized by high incidence and mortality rates. A timely and accurate diagnosis is critical for the prognosis. Multiple parameters are acquired to assess acute and long-term effects, including the protein troponin. Furthermore, several studies have examined the influence of the patients‘ sex in APE in regard to mortality and pathology symptoms. However, conflicting findings have emerged from these results. A clear correlation between sex-specific disease progression, disease characteristics and for example mortality cannot be clearly established.
In our previous study, we discovered that the imaging characteristics of the arterial thrombus can be utilized for the prediction of mortality and the biomarker troponin. Our new publication, “Influence of patients’ sex on radiomics-based predictions of lung artery thrombus” (DOI: 10.1007/s11548-026-03704-1), focuses in this regard on the differences between the patient sexes and their potential influence on machine learning processes in this radiomics analysis.
Our results showed that the classification process using combined sex data present lower predictive power compared to the sex-specific analysis. In general, calculations involving only female patients yielded the best results with a weighted F1 score of up to 0.9626 and a balanced accuracy of up to 0.8245. This indicates that the imaging characteristics of the thrombus are distinct between the patient sexes, which must be investigated more closely in the future. A clear differentiation of the clinical presentation of APE between the sexes could aid in the diagnosis and treatment of APE patients.
Published in the International Journal of Computer Assisted Radiology and Surgery.
Contact person: Prof. Dr.-Ing. Patrique Fiedler, M. Sc. Lea Ehrhardt
