Adaptive Radiation Therapy in the Era of Big Data with Artificial Intelligence
X Qi1*, X Gu2*, Y Wang3*, S Bentzen4*, X Li5*, I El Naqa6*, (1) UCLA School of Medicine, Los Angeles, CA, (2) UT Southwestern Medical Center, Dallas, TX, (3) Massachusetts General Hospital, Boston, MA, (4) University of Maryland, Baltimore, MD, (5) Medical College of Wisconsin, Milwaukee, WI, (6) Moffitt Cancer Center, Tampa, FL
Presentations
10:50 AM AI/DL Based Auto-Segmentation for ART: Technical Development and Clinical Translation - X Gu, Presenting Author
11:10 AM Machine Learning-Based Auto Planning and Dose Prediction for ART - Y Wang, Presenting Author
11:30 AM Learning from Experience: AI and Data Mining for Outcome-Driven Radiation Treatment Planning - S Bentzen, Presenting Author
11:50 AM Treatment Response Based on Delta Radiomics for ART - X Li, Presenting Author
12:10 PM Automated Adaptive Decision Making with Deep Leaning Neural Network (DLNN) - I El Naqa, Presenting Author
TH-AB-TRACK 5-0 (Thursday, 7/29/2021) 10:30 AM - 12:30 PM [Eastern Time (GMT-4)]
Abstract:
Adaptive radiation therapy (ART) aims to adjust the treatment plan based on anatomical and/or functional changes assessed based on frequent images acquired during the treatment course. Despite the appealing advantages, clinical implementations of ART, especially on-line ART, have been limited due to its complexity and lack of efficient and robust software tools in almost every step of the process. The rapid advances of artificial intelligence (AI) and AI-based technologies hold great promise for ART. This session will start with the current clinical ART workflows and its clinical challenges, then will discuss the latest AI-based development & clinical applications in each aspect of the ART workflow, including AI/deep learning (DL) based auto-segmentation; AI-based tools for auto-planning, secondary dose verification, plan check and
patient-specific QA. Advanced development accounting for treatment response (as a feedback), including AI and data mining for outcome-driven radiation treatment planning; treatment response-based on delta radiomics for ART, and automated ART decision making frameworks built from deep learning neural networks will be also presented.
Learning objectives:
1. Understand the current clinical practice and challenges of ART and the role of AI
2. Present the latest developments & clinical applications of AI-based automatic segmentation
3. Present AI-based tools for auto-planning, secondary dose verification, plan check and
patient-specific QA
4. Learn AI and data mining for outcome-driven radiation treatment planning
5. Learn treatment response based on delta radiomics for ART
6. Present AI methods for automated ART decision making using deep learning neural networks for clinical implementation of ART
Funding Support, Disclosures, and Conflict of Interest: Received a grant from Varian Inc.
Handouts
- 166-57696-15651646-171094-1593263898.pdf (Issam El Naqa)
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email