Dr Philip Whybra
Research Associate
- whybrap@cardiff.ac.uk
- N/1.51, Adeiladau'r Frenhines - Adeilad y Gogledd, 5 The Parade, Heol Casnewydd, Caerdydd, CF24 3AA
Trosolwg
I have strong links to Cardiff University, having first graduated here with a Masters in Physics (MPhys) in 2015. After time in industry, I re-joined the School of Engineering at Cardiff as a PhD candidate in October 2016. I now continue to work within the Cancer Imaging and Data Analytics (CIDA) team and the Medical Engineering Research Group as a Research Associate.
My research interests are broadly in the topics of Medical Image Analysis, Imaging Biomarkers, and Radiomics. I am curious about all aspects of image analysis that could be used to further personalise cancer treatment.
A key aspect of my work concerns the standardisation of radiomic techniques. This is a necessity to enable transition of any promising research models to use within a clinical setting. Notably, I am a core member of an international effort known as the Image Biomarker Standardisation Initiative (IBSI) (https://theibsi.github.io).
Bywgraffiad
Education and Work
- Research Associate, Cardiff University, UK (2021-current)
- Research Assistant, Cardiff University, UK (2021)
- PhD - School of Engineering, Cardiff University, UK (2016-2021)
- Lab Demonstrator (Computational Physics, PX1224 and PX2134), Cardiff University, UK (2017-2018)
- Intern Research Scientist, Mirada Medical, Oxford (2016)
- MPhys - School of Physics and Astronomy, Cardiff University, UK (2011-2015)
Pwyllgorau ac adolygu
- Journal Reviewer, Medical Physics
Cyhoeddiadau
2024
- Whybra, P. et al. 2024. The image biomarker standardization initiative: Standardized convolutional filters for reproducible radiomics and enhanced clinical insights. Radiology 310(2) (10.1148/radiol.231319)
2023
- Whybra, P. and Spezi, E. 2023. Sensitivity of standardised radiomics algorithms to mask generation across different software platforms. Scientific Reports 13, article number: 14419. (10.1038/s41598-023-41475-w)
- Duman, A., Whybra, P., Powell, J., Thomas, S., Sun, X. and Spezi, E. 2023. PO-1620 Transferability of deep learning models to the segmentation of gross tumour volume in brain cancer. Radiotherapy & Oncology 182(S1), pp. S1315-S1316. (10.1016/S0167-8140(23)66535-1)
2021
- Whybra, P. 2021. Standardisation and optimisation of radiomic techniques for the identification of robust imaging biomarkers in oncology. PhD Thesis, Cardiff University.
- Shi, Z. et al. 2021. Prediction of lymph node metastases using pre-treatment PET radiomics of the primary tumour in esophageal adenocarcinoma: an external validation study. British Journal of Radiology 94(1118), article number: 20201042. (10.1259/bjr.20201042)
2020
- Mori, M. et al. 2020. Training and validation of a robust PET radiomic-based index to predict distant-relapse-free-survival after radio-chemotherapy for locally advanced pancreatic cancer. Radiotherapy and Oncology 153, pp. 258-264. (10.1016/j.radonc.2020.07.003)
- Zwanenburg, A. et al. 2020. The Image Biomarker Standardization Initiative: standardized quantitative radiomics for high throughput image-based phenotyping. Radiology 295(2), pp. 328-338. (10.1148/radiol.2020191145)
2019
- Shi, Z. et al. 2019. External validation of radiation-induced dyspnea models on esophageal cancer radiotherapy patients. Frontiers in Oncology 9, article number: 1411. (10.3389/fonc.2019.01411)
- Piazzese, C., Foley, K., Whybra, P., Hurt, C., Crosby, T. and Spezi, E. 2019. Discovery of stable and prognostic CT-based radiomic features independent of contrast administration and dimensionality in oesophageal cancer. PLoS ONE 14(11), article number: e0225550. (10.1371/journal.pone.0225550)
- Whybra, P., Parkinson, C., Foley, K., Staffurth, J. and Spezi, E. 2019. Assessing radiomic feature robustness to interpolation in 18F-FDG PET imaging. Scientific Reports 9(1), article number: 9649. (10.1038/s41598-019-46030-0)
- Foley, K. G. et al. 2019. External validation of a prognostic model incorporating quantitative PET image features in esophageal cancer. Radiotherapy and Oncology 133, pp. 205-212. (10.1016/j.radonc.2018.10.033)
- Whybra, P., Parkinson, C., Foley, K., Staffurth, J. and Spezi, E. 2019. A novel normalisation technique for voxel size dependent radiomic features in oesophageal cancer [Abstract]. Radiotherapy and Oncology 133, pp. S523-S524. (10.1016/S0167-8140(19)31383-0)
- Piazzese, C., Whybra, P., Qasem, E., Harris, D., Gtaes, R., Foley, K. and Spezi, E. 2019. Radiomics in rectal cancer: prognostic significance of 3D features extracted from diagnostic MRI [Abstract]. Radiotherapy and Oncology 133, pp. S1048-S1048. (10.1016/S0167-8140(19)32346-1)
- Whybra, P., Parkinson, C., Foley, K., Staffurth, J. and Spezi, E. 2019. PO-0963 A novel normalisation technique for voxel size dependent radiomic features in oesophageal cancer. Radiotherapy and Oncology 133(S1), pp. S523-S524. (10.1016/S0167-8140(19)31383-0)
- Piazzese, C., Whybra, P., Carrington, R., Crosby, T., Staffurth, J., Foley, K. and Spezi, E. 2019. PO-0964 Stability and prognostic significance of CT radiomic features from oesophageal cancer patients. Radiotherapy and Oncology 133(S1), pp. S524-S525. (10.1016/S0167-8140(19)31384-2)
- Piazzese, C., Whybra, P., Qasem, E., Harris, D., Gtaes, R., Foley, K. and Spezi, E. 2019. EP-1926 Radiomics in rectal cancer: prognostic significance of 3D features extracted from diagnostic MRI. Radiotherapy and Oncology 133(S1), pp. S1048. (10.1016/S0167-8140(19)32346-1)
2018
- Parkinson, C. et al. 2018. Dependency of patient risk stratification on PET target volume definition in oesophageal cancer. Presented at: ESTRO, Barcelona, Spain, 20-24 April 2018.
- Parkinson, C., Whybra, P., Staffurth, J., Marshall, C. and Spezi, E. 2018. ATLAAS - Investigation into the incorporation of morphological data on automated segmentation. Presented at: EANM Congress 2018: European Association of Nuclear Medicine., Dusseldorf, Germany, 12-18 October 2018.
- Parkinson, C. et al. 2018. Evaluation of prognostic models developed using standardised image features from different PET automated segmentation methods. EJNMMI Research 8, article number: 29. (10.1186/s13550-018-0379-3)
- Piazzese, C., Whybra, P., Carrington, R., Crosby, T., Staffurth, J., Foley, K. and Spezi, E. 2018. Evaluation of 2D and 3D radiomics features extracted from CT images of oesophageal cancer patients. Radiotherapy and Oncology 127, pp. S1180-S1181. (10.1016/S0167-8140(18)32450-2)
- Zwanenburg, A. et al. 2018. Results from the image biomarker standardisation initiative [Poster]. Radiotherapy and Oncology 127(S1), pp. S543-S544. (10.1016/S0167-8140(18)31291-X)
- Shi, Z. et al. 2018. External validation of radiation-induced dyspnea models on esophageal cancer radiotherapy patients. Radiotherapy and Oncology 127, pp. S168-S168. (10.1016/S0167-8140(18)30628-5)
- Parkinson, C. et al. 2018. Dependency of patient risk stratification on PET target volume definition in Oesophageal cancer. Radiotherapy and Oncology 127(S1), pp. S503-S504. (10.1016/S0167-8140(18)31241-6)
- Whybra, P., Foley, K., Parkinson, C., Staffurth, J. and Spezi, E. 2018. Effect of interpolation on 3D texture analysis of PET imaging in oesophageal cancer. Radiotherapy and Oncology 127(S1), pp. S1164-S1165. (10.1016/S0167-8140(18)32426-5)
Radiomics and quantitative medical imaging analysis
- Radiomic techniques transform medical imaging into quantitative, descriptive features.
- Quantitative features could be used as imaging biomarkers that guide patient treatment decisions.
- Cancer diagnosis, treatment and monitoring could significantly benefit from radiomic analysis due to the abundance of routine imaging already acquired as part of treatment.
Some examples of potential radiomics applications in cancer management:
- Diagnosis (e.g. benign or malignancy status)
- Treatment response (i.e. likelihood a given patient will benefit from a given treatment)
- Tumour aggression (e.g. chance of relapse or disease progression after treatment)
Quantitative radiomic features are typically aggregated from regions of interest defined in the image. In cancer research, this region is usually a segmentation of the primary tumour.
The features collected broadly describe concepts including tumour shape, intensity statistics, and image texture. Additionally, specially designed filters can emphasise image characteristics prior to feature aggregation.
Radiomics standardisation
Standardisation and reproducibility poses a significant challenge to the adoption of radiomics techniques in a clinical setting. Large scale adoption will require appropriate validation of models.
A key part of my research is in the standardisation of radiomics algorithms and reporting guidelines necessary to replicate studies. To achieve this aim, I am a core member of an international effort (Image Biomarker Standardisation Initiative (IBSI) - https://theibsi.github.io) to standardise image processing through the development of consensus based benchmarks.
