Development and Inter-Institutional Validation of An Automatic Vertebral-Body Misalignment Error Detector for Cone-Beam CT Guided Radiotherapy
D Luximon1*, T Ritter2, E Fields2, J Neylon1, R Petragallo1, Y Abdulkadir1, J Charters1, D Low1, J Lamb1, (1) University of California, Los Angeles, Los Angeles, CA, (2) VCU Health System, Chesterfield, VA
Presentations
SU-F-BRA-2 (Sunday, 7/10/2022) 2:00 PM - 3:00 PM [Eastern Time (GMT-4)]
Ballroom A
Purpose: Off-by-one vertebral body misalignments are rare but serious errors in Cone-Beam Computed Tomography (CBCT) guided radiotherapy. We propose an automatic error detection algorithm that uses a three-branch convolutional neural network error detection model (EDM) to detect off-by-one vertebral body misalignments using planning computed tomography (CT) images and setup CBCT images.
Methods: Algorithm training and test data consisted of planning CTs and setup CBCTs from 480 patients undergoing radiotherapy treatment in the thoracoabdominal region at two radiotherapy clinics. The clinically applied registration was used to derive true-negative (no error) data. The setup and planning images were then misaligned by one vertebral body in both the superior and inferior directions, simulating the most likely misalignment scenarios. For each of the aligned and misaligned 3D image pairs, 2D slice pairs were automatically extracted in each anatomical plane about a point within the vertebral column. The three slice pairs obtained were then inputted to the EDM which returned a probability of vertebral misalignment. One model (EDM₁) was trained solely on data from institution #1. EDM₁ was further trained using a lower learning rate on a dataset from institution #2 to produce EDM(FT). These two models were validated on a randomly selected and unseen dataset composed of images from both institutions, for a total of 303 image pairs. The model performances were quantified using a receiver operating characteristic analysis.
Results: EDM₁ and EDM(FT) resulted in an area under curve of 99.5% and 99.4%, respectively. For a fixed specificity of 99%, EDM₁ and EDM(FT) resulted in a sensitivity of 93% and 95%, respectively.
Conclusion: The proposed algorithm demonstrated accuracy in identifying off-by-one vertebral body misalignments in CBCT-guided radiotherapy that was sufficiently high to allow for practical implementation. We found that fine-tuning the model on a multi-facility dataset can further enhance the robustness of the algorithm.
Funding Support, Disclosures, and Conflict of Interest: The research reported in this study was supported by the Agency for Healthcare Research and Quality (AHRQ) under award number 1R01HS026486.
Keywords
Image-guided Therapy, Cone-beam CT, Quality Assurance
Taxonomy
IM/TH- Image Analysis (Single Modality or Multi-Modality): Computer/machine vision
Contact Email