Clinical Scorecard: Evaluating the Prognostic Value of Radiomic Analysis of Pulmonary Artery Thrombus in Patients with Pulmonary Embolism
At a Glance
Category
Detail
Condition
Acute pulmonary embolism (APE)
Key Mechanisms
Radiomic feature extraction from CT pulmonary angiogram (CTPA) images of pulmonary artery thrombus to predict 30-day mortality and troponin levels
Target Population
Patients with acute pulmonary embolism undergoing CTPA imaging
Care Setting
Hospital setting with access to CT imaging and radiological expertise
Key Highlights
CTPA is the standard imaging modality for diagnosing APE and assessing parameters like right ventricle enlargement and IVC reflux linked to mortality risk.
Radiomic texture features of the pulmonary embolus area correlate with clinical outcomes including 30-day mortality and troponin levels.
Manual segmentation of thrombus area followed by radiomics analysis enables classification and prediction of patient prognosis in APE.
Guideline-Based Recommendations
Diagnosis
Use CT pulmonary angiogram (CTPA) as the standard imaging technique for acute pulmonary embolism diagnosis.
Apply manual segmentation of thrombus area on CTPA images for detailed radiomic feature extraction.
Management
Consider radiomic analysis of thrombus texture features to aid in risk stratification and prognosis prediction.
Use clinical scores such as PESI, sPESI, and Geneva score alongside imaging and biomarker data for comprehensive assessment.
Monitoring & Follow-up
Monitor 30-day mortality outcomes and troponin levels as key prognostic indicators in APE patients.
Employ follow-up imaging and clinical evaluation to assess treatment response and patient status.
Risks
Exclude patients with chronic pulmonary embolism from acute radiomic prognostic assessments.
Ensure accurate manual segmentation to avoid errors in radiomic feature extraction.
Patient & Prescribing Data
58 male and 28 female patients with acute pulmonary embolism, mean age 64.7 ± 14.8 years
Radiomic features extracted prior to thrombolytic treatment can predict mortality and troponin-related severity; thrombolytic treatment was not administered during imaging acquisition.
Clinical Best Practices
Perform manual segmentation of pulmonary artery thrombus on CTPA images verified by a radiologist.
Use radiomics software tools such as PyRadiomics and MeVisLab for feature extraction and analysis.
Balance data classes using over- and undersampling techniques when analyzing biomarkers like troponin for predictive modeling.
Integrate radiomic features with clinical scores and biomarkers for improved prognostic accuracy.
