Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Faculty of Health and Social Sciences | en_US |
| dc.contributor.advisor | Chan, Wing Chi Lawrence (HTI) | en_US |
| dc.creator | Ng, Andrew Tik Ho | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/14200 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Utilization of artificial intelligence in predicting mortality prior to endovascular thrombectomy | en_US |
| dcterms.abstract | Introduction: Endovascular thrombectomy (EVT) is regarded as the standard of care for acute ischemic stroke (AIS) patients with large vessel occlusion (LVO). However, the mortality rates for these patients remain alarmingly high and access to EVT remains limited in many regions worldwide. Dependable mortality prediction based on timely clinical information is of great importance. Several prediction scores have been developed to predict the functional outcomes but exhibited notable limitations in patient benefit and variable readiness for predicting mortality. | en_US |
| dcterms.abstract | Methodology: The study retrospectively reviewed 151 patients who underwent EVT at Pamela Youde Nethersole Eastern Hospital between April 1, 2017, and October 31, 2023. The primary outcome of the study was 90-day mortality after AIS. The models were developed using two feature selection approaches (model I: sequential forward feature selection, model II: sequential forward feature selection after identifying variables through univariate logistic regression) and six algorithms (logistic regression (LR), random forest (RF), extreme gradient boosting (XGB), k-nearest neighbor (KNN), support vector machine (SVM), and neural network (NN)). Model performance was evaluated by using external validation data of 312 cases and compared with the traditional prediction scores, including Houston Intra-Arterial recanalization Therapy 2 (HIAT2), Totaled Health Risks in Vascular Events (THRIVE), and Predicting 90-days mortality of AIS with MT (PRACTICE) scores. Additionally, the benefit of clinical decisions, which was defined as whether EVT would be withheld from patients who would not experience mortality within 90 days, was compared between the most effective algorithm and prediction score. | en_US |
| dcterms.abstract | Results: In model I, none of the algorithms achieved an area under the receiver operating characteristic curve (AUC) exceeding 0.7, with LR achieving the highest AUC of 0.691. In contrast, model II demonstrated a significant improvement, with three algorithms exceeding an AUC of 0.7: LR at 0.712, SVM at 0.705, and NN at 0.702. The HIAT2 score surpassed all others with an AUC of 0.717, making it the only prediction score to exceed the 0.7 threshold. Meanwhile, the SVM using model II achieved an area under the precision-recall curve (AUPRC) of 0.421, balanced accuracy of 0.618, F1 score of 0.375, Matthews Correlation Coefficient (MCC) of 0.270, and Brier score of 0.143. The HIAT2 score achieved an AUPRC of 0.402, balanced accuracy of 0.704, F1 score of 0.474, MCC of 0.340, and Brier score of 0.143. | en_US |
| dcterms.abstract | Discussion: This study identified the SVM using model II as the best algorithm among the various options. Meanwhile, the HIAT2 score surpassed all algorithms using models I and II with an AUC of 0.717. However, most algorithms provided a greater net benefit than the traditional prediction scores. When assessing the impact of clinical decision-making concerning patient selection for EVT, the SVM algorithm using model II resulted in 38 more patients (38/256, 14.8%) benefiting from EVT than the HIAT2 score. | en_US |
| dcterms.abstract | Conclusion: Machine learning (ML) algorithms developed by routinely ready variables could offer beneficial insights for predicting mortality in AIS patients undergoing EVT. | en_US |
| dcterms.extent | xiii, 166 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2025 | en_US |
| dcterms.educationalLevel | DHSc | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.accessRights | restricted access | en_US |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 8653.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 1.83 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/200/14200