MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Optimization in Imaging and Therapy

A Wang1*, H Enderling2*, K Sheng3*, (1) Stanford University, Stanford, CA, (2) H. Lee Moffitt Cancer Center & Research Institute, Tampa, FL,(3) UCLA School of Medicine, Los Angeles, CA




Presentations

(Monday, 7/15/2019) 4:30 PM - 6:00 PM

Room: 225BCD

Better radiation therapy can be achieved through better image guidance, better modeling of treatment response and better radiation dose conformity. Optimization plays important roles in all three aspects.
For cone beam CT (CBCT), the conventional filtered back projection method suffers from image artifacts and low signal to noise ratio that limit its utility. On the other hand, the CBCT reconstruction problem can be formulated as an iterative inverse optimization problem and solved accordingly. The methods for iterative CBCT reconstruction will be presented as well its benefits and limitations.
Mathematical, biological, and clinical sciences can be integrated to conduct 'virtual trials' to ultimately personalize cancer treatment. Based on established logistic tumor growth dynamics concepts, a patient-specific proliferation saturation index (PSI) to derive the fraction of radiation-sensitive proliferating population of cancer cells within an imaging-derived tumor volume was introduced. PSI describes the tumor-extrinsic tissue-environmental properties of the patient that influence tumor growth dynamics. PSI can be calculated prior to therapy from routine images, and can inform about the optimized radiation protocol to maximally shrink the tumor. With the ability to predict responses to therapy comes the opportunity to evaluate clinically observed responses beyond RECIST, and to adapt radiation dose and dose fractionation – to escalate dose when necessary and to de-escalate dose in selected patients without sacrificing control. The PSI concept and the current prospective clinical evaluation, as well as future opportunities to integrate mathematical oncology concepts towards radiation treatment personalization, will be presented.
There is a constant need to deliver better and more efficient radiotherapy to patients. Recent advances in optimization methods, computer vision, and robotics afford the potential for markedly improved volumetric modulated arc therapy (VMAT) by exploring new dimensions in optimization. The new dimensions include non-coplanar trajectories, dynamic collimator rotation, dual-layer multileaf collimators and energy modulations. The ability to explore these additional dimensions are established based on the non-progressive sampling VMAT framework that has been shown superior to the progressive sampling method. The presentation will describe the mathematical tools to model and solve these large scale optimization problems, as well as the dosimetric benefits and delivery efficiency on current radiation therapy delivery platforms.

Learning objectives:

1. Understand iterative CBCT reconstruction as an optimization problem.
2. Understand the mathematical modeling of patient-specific tumor growth dynamics and its applications in radiotherapy outcome prediction.
3. Understand the methods and benefits of including additional degrees of freedom into VMAT optimziation.

Handouts

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email