CT ventilation images produced by a 3D neural network show improvement over the Jacobian and HU DIR‐based methods to predict quantized lung function
Our 3D neural network produces a quantized CTVI with higher similarity to the ground truth than the 2D U-Net and DIR-based Jacobian and HU methods. As it produces a quantized CTVI directly, CTVI<sub>nnU-Net</sub> avoids the need for thresholding to identify high-function lung regions. With faster evaluation and improved accuracy, neural networks show promise for the clinical implementation of functional lung avoidance.
Authors
- Jeremy Lim (ORCID: https://orcid.org/0000-0003-4220-2404)
- Ricky O’Brien (ORCID: https://orcid.org/0000-0002-6586-7356)
- Hilary L. Byrne (ORCID: https://orcid.org/0000-0001-7023-7976)
- Daryl Wilding‐McBride (ORCID: https://orcid.org/0000-0003-3796-1297)
Institutions
- The University of Sydney (AU)
- RMIT University (AU)
Publication Details
- Journal
- Medical Physics
- Published
- 2024-11-23
- DOI
- https://doi.org/10.1002/mp.17532
- Citations
- 2
- Primary Topic
- Advanced Radiotherapy Techniques
- Type
- article
- Field-Weighted Citation Impact
- 1.22