Authors: Felix Liu, Albin Fredriksson, Stefano Markidis

Published on: May 06, 2024

Impact Score: 7.4

Arxiv code: Arxiv:2405.03584

Summary

• What is new: Development of a GPU accelerated optimization solver for radiation therapy, significantly speeding up the treatment planning process.
• Why this is important: Current radiation therapy optimization problems present a computational bottleneck due to their size and complexity.
• What the research proposes: An interior point method (IPM) utilizing iterative linear algebra for GPU acceleration to solve optimization problems more efficiently.
• Results: The new GPU accelerated IPM solver achieved a 1.4 to 4.4 times faster solution time for quadratic programming sub-problems in real patient cases compared to existing solutions.

Technical Details

Technological frameworks used: C++20, CUDA for GPU acceleration

Models used: Iterative linear algebra based interior point method

Data used: Problems from RayStation treatment planning system

Potential Impact

Cancer clinics and treatment planning software markets, specifically benefiting companies like RaySearch Laboratories.

Want to implement this idea in a business?

We have generated a startup concept here: OptiTherapyTech.