Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Faculty of Health and Social Sciences | en_US |
| dc.contributor.advisor | Cai, Jing (HTI) | en_US |
| dc.creator | Wong, Tsz Lung | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/12851 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Biochemical Prostate Specific Antigen (PSA) prediction for prostate cancer patients receiving MR-guided radiotherapy using multi-modality radiomics | en_US |
| dcterms.abstract | Introduction | en_US |
| dcterms.abstract | Prediction of early treatment response in prostate cancer is highly acclaimed and important for treatment decision making. However, there is a lack of prediction strategy applied in current clinical practice. This retrospective study will predict one-month post-treatment serum PSA level as a surrogate of early treatment response for prostate cancer patient treated with SBRT via pre-treatment CT and MR images using radiomics approaches. | en_US |
| dcterms.abstract | Methods | en_US |
| dcterms.abstract | One hundred and nine prostate cancer patients receiving 5 fractions of Stereotactic Body Radiotherapy (SBRT) in Elekta Unity MR-Linac at Hong Kong Sanatorium Hospital Eastern Cancer Centre from Jan 2020 to May 2021 were retrospectively recruited into the study. Pre-treatment CT and T2-w MR images were processed using RADAR (software developed in-house by Department of Health, Technology and Information at Hong Kong Polytechnic University). 1130 radiomics features were extracted from the prostate CTV (clinical target volume) in CT and MR images. The radiomic model was trained using CT and MR-based radiomic features to predict the serum PSA status one month after the completion of treatment. The maximum relevance and minimum redundancy (mRMR) feature selection algorithm was implemented to select the final features for the radiomic model building. The performance of the radiomic model was validated using a 3-fold cross-validation. Area under the receiver operating characteristic curve (AUC of ROC) for the cross-validation were used to evaluate the predictive power of the radiomic model. | en_US |
| dcterms.abstract | Results | en_US |
| dcterms.abstract | Highest 8 features from CT were selected to build the CT-model and highest 8 features from MR to build the MR-model. Averaged AUC in training and testing over cross-validations were 0.83 (SD=0.04) and 0.77 (SD=0.08) for CT-based radiomic model. For MR-based model, the AUC for training and testing were 0.85 (SD=0.04) and 0.78 (SD=0.09) respectively. The statistical test showed there is no significant AUC difference in CT vs. MR-based radiomic model (p > 0.05). The combined model (CT + MR) without re-prioritizing both radiomics features using CT (n=6) and MR (n=6) showed AUC of 0.89 (SD=0.02) and 0.82 (SD=0.09) for training and testing respectively. The statistical test demonstrated a statistically significant improvement (p<0.05) when compared to CT alone model but not with MR alone model in response prediction. The model (CTMR) using a hybrid and mRMR re-prioritized features from both modalities with 11 shortlisted features (CT=6, MR=5) demonstrated AUC of 0.90 (SD=0.03) and 0.83 (SD=0.09) in training and testing respectively. When comparing the AUC with CT or MR alone model, the statistical test showed the result is significant (p>0.05) in both tests. | en_US |
| dcterms.abstract | Conclusion | en_US |
| dcterms.abstract | The results indicate radiomics features from CT & MR have the potential to predict the change of one-month post-treatment serum PSA level for prostate cancer. And the combined model (CTMR) using mRMR to re-prioritizes the combined features achieve the best AUC of ROC curve when compared with other models (p<0.05). Our results demonstrated multi-modality CT & MR radiomics can successfully predict early PSA results after SBRT in prostate cancer patients. And the prediction power in hybrid model is more superior than single modality models. The potential application of the results in clinical environment could assist the clinician in practicing precision medicine based on multi-modality radiomics model prediction. | en_US |
| dcterms.extent | xx, 106 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2023 | en_US |
| dcterms.educationalLevel | DHSc | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.LCSH | Prostate -- Cancer -- Diagnosis | en_US |
| dcterms.LCSH | Prostate -- Cancer -- Prognosis | en_US |
| dcterms.LCSH | Prostate -- Cancer -- Treatment | en_US |
| dcterms.LCSH | Diagnostic imaging | en_US |
| dcterms.LCSH | Hong Kong Polytechnic University -- Dissertations | en_US |
| dcterms.accessRights | restricted access | en_US |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 7300.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 2.27 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/12851