From: Radiomics in the evaluation of ovarian masses — a systematic review
Limitation | Opportunity |
---|---|
1. Multiple scanners and imaging protocols | Phantom studies Image harmonisation techniques IBSI-compliant radiomic packages Validate reproducibility using publicly available data sets |
2. Comparison with human readers | Undertake blinded review with human interpreters with varying experience levels |
3. 2D segmentation and manual segmentation | Undertake all segmentation using 3D Develop automated/semi-automated segmentation techniques |
4. Lack of external validation | Use publicly available data sets as external validation |
5. Open source code | Upload code to public repositories such as GitHub |