University of Twente Student Theses
Accuracy and stability of radiomic features for characterising tumour heterogeneity using multimodality imaging: a phantom study
Meijer, K.M. (2019) Accuracy and stability of radiomic features for characterising tumour heterogeneity using multimodality imaging: a phantom study.
|
PDF
2MB |
| Abstract: | Introduction: The aim of this study was to establish the accuracy and variability of radiomic features on a multimodality imaging platform using a multi-compartment heterogeneous phantom. Materials and Methods: A 3D printed multi-compartment phantom, was developed to simulate heterogeneous composition patterns in positron emitting tomography (PET), computed tomography (CT) and magnetic resonance (MR) imaging. Image acquisition was repeated 5 times for variability analysis. Effect of scalability was assessed by comparison of the three scaled compartments. Furthermore, high resolution CT imaging was performed (ground truth) to determine the accuracy of radiomic features. A total of 54 radiomic features were extracted using PyRadiomics. Results: Regarding the variability analysis, features extracted from MR had the highest reproducibility (N=1 with a variation 20%), followed by CT (N=17 with a variation 20%) and PET (N=28 with a variation 20%). Scalability analysis showed that the size of the phantom structures significantly affected the absolute values of the majority of the radiomic features. Finally, the absolute value of most first order class and GLCM features showed less than 20% deviation from the ground truth reference values. Conclusion: This study showed that heterogeneous patterns can be simulated using a multi-compartment imaging phantom and the results showed different, but promising results regarding variability, scalability and accuracy. |
| Item Type: | Essay (Master) |
| Faculty: | TNW: Science and Technology |
| Subject: | 44 medicine, 50 technical science in general |
| Programme: | Technical Medicine MSc (60033) |
| Link to this item: | https://purl.utwente.nl/essays/78085 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page