Improving Treatment Of Local Liver Ablation Therapy With Deep Learning And Biomechanical Modeling

Author

Brian Anderson, The University of Texas M D Anderson Cancer Center UTHealth Graduate School of Biomedical SciencesFollow
Kristy Brock, The University of Texas M D Anderson Cancer Center UTHealth Graduate School of Biomedical SciencesFollow
Laurence Court, The University of Texas M D Anderson Cancer Center UTHealth Graduate School of Biomedical SciencesFollow
Carlos Eduardo Cardenas, The University of Texas M D Anderson Cancer Center UTHealth Graduate School of Biomedical Sciences
Erik Cressman, The University of Texas MD Anderson Cancer Center UTHealth Graduate School of Biomedical Sciences
Ankit Patel, Rice UniversityFollow

Author ORCID Identifier

0000-0002-2748-1444

Date of Graduation

5-2021

Document Type

Dissertation (PhD)

Program Affiliation

Medical Physics

Degree Name

Doctor of Philosophy (PhD)

Advisor/Committee Chair

Kristy K Brock

Committee Member

Laurence E Court

Committee Member

Carlos E Cardenas

Committee Member

Richard E Wendt

Committee Member

Erik N K Cressman

Committee Member

Ankit Patel

Abstract

In the United States, colorectal cancer is the third most diagnosed cancer, and 60-70% of patients will develop liver metastasis. While surgical liver resection of metastasis is the standard of care for treatment with curative intent, it is only avai lable to about 20% of patients. For patients who are not surgical candidates, local percutaneous ablation therapy (PTA) has been shown to have a similar 5-year overall survival rate. However, PTA can be a challenging procedure, largely due to spatial uncertainties in the localization of the ablation probe, and in measuring the delivered ablation margin.

For this work, we hypothesized that biomechanical modeling could be used to reduce the spatial uncertainties inherent in PTA, furthermore, that deep learning could create segmentations that are qualitatively preferred to manual contours of the liver , delineate the liver structures vii rapidly, and predict local progression based on intra-procedural imaging. Firstly, our study with biomechanical modeling to reduce spatial uncertainties and measure the minimum distance to agreement found a signif icant dif ference (p

Our work resulted in the validation of biomechanical modeling in ablation assessment, creation of automatic segmentation models for the liver, disease, and ablation volume within our treatment planning system, and an outcome prediction model. The liv er model has been used to segment over 1,800 exams in our clinic since 3/23/2021, and our outcome prediction model provides visual interpretations of model decisions. The culmination of this work has enabled our on-going Phase 2 Clinical Trial (NCT04083378). Future studies will improve upon autosegmentation models, and further investigate outcome-prediction modeling.

Keywords

Deep learning, liver, interventional radiology, percutaneous ablation therapy