Cardio-oncology & otherDiagnostic or prognostic study
18F-FDG-Based Radiomics and Machine Learning
- Diagnostic accuracy study
- retrospective
- single-centre
- radiomics and machine learning
- independent test cohort
Authors' conclusion
The use of ML for analyzing 18F-FDG-PET/CT images in PVE diagnosis was feasible and beneficial, particularly when ML was included in the ESC 2015 criteria.
Study context
- Population
- Patients with suspected aortic prosthetic valve endocarditis undergoing 18F-FDG positron emission tomography (PET)/computed tomography (CT), single centre, 2015–2021test cohort n = 40
- Participants
- 108 patients
- Comparison
- ESC-ML vs PET-ML
- Outcome
- Performance of radiomics and machine learning-based analysis of 18F-FDG PET/CT (PET-ML) as a major criterion for the ESC score (ESC-ML) in PVE diagnosis
Performance
As reported in the abstract
PVE diagnosis, test cohort
| Test or model | Sensitivity | Specificity |
|---|---|---|
| ESC-ML | 72% | 83% |
Secondary outcomes
- Prosthetic valve endocarditis (PVE) diagnosis with PET-ML standalone, test cohort
- 13 of 22 · Positive in definite PVE · Positive in rejected PVE 3 of 18 · Sensitivity 59% · Specificity 83%
Original publication
18F-FDG-Based Radiomics and Machine Learning. JACC Cardiovasc Imaging. Published 12 April 2023. doi:10.1016/j.jcmg.2023.01.020
Summarised from the published abstract; every number is checked against it