PURPOSE: Automation and computer assistance can support quality assurance tasks in radiotherapy. Retrospective image review requires significant human resources, and automation of image review remains a noteworthy missing element in previous work. Here, we present initial findings from a proof-of-concept clinical implementation of an AI-assisted review of CBCT registrations used for patient setup. METHODS: An automated pipeline was developed and executed nightly, utilizing python scripts to interact with the clinical database through DICOM networking protocol and automate data retrieval and analysis. A previously developed artificial intelligence (AI) algorithm scored CBCT setup registrations based on misalignment likelihood, using a scale from 0 (most unlikely) through 1 (most likely). Over a 45-day period, 1357 pre-treatment CBCT registrations from 197 patients were retrieved and analyzed by the pipeline. Daily summary reports of the previous days registrations were produced. Initial action levels targeted 10% of cases to highlight for in-depth physics review. A validation subset of 100 cases was scored by three independent observers to characterize AI-model performance. RESULTS: Following an ROC analysis, a global threshold for model predictions of 0.87 was determined, with a sensitivity of 100% and specificity of 82%. Inspecting the observer scores for the stratified validation dataset showed a statistically significant correlation between observer scores and model predictions. CONCLUSION: In this work, we describe the implementation of an automated AI-analysis pipeline for daily quantitative analysis of CBCT-guided patient setup registrations. The AI-model was validated against independent expert observers, and appropriate action levels were determined to minimize false positives without sacrificing sensitivity. Case studies demonstrate the potential benefits of such a pipeline to bolster quality and safety programs in radiotherapy. To the authors knowledge, there are no previous works performing AI-assisted assessment of pre-treatment CBCT-based patient alignment.

Background

This study utilizes interviews of clinical medical physicists to investigate self-reported shortcomings of the current weekly chart check workflow and opportunities for improvement.

Methods

Nineteen medical physicists were recruited for a 30-minute semi-structured interview, with a particular focus placed on image review and the use of automated tools for image review in weekly checks. Survey-type questions were used to gather quantitative information about chart check practices and importance placed on reducing chart check workloads versus increasing chart check effectiveness. Open-ended questions were used to probe respondents about their current weekly chart check workflow, opinions of the value of weekly chart checks and perceived shortcomings, and barriers and facilitators to the implementation of automated chart check tools. Thematic analysis was used to develop common themes across the interviews.

Results

Physicists ranked highly the value of reducing the time spent on weekly chart checks (average 6.3 on a scale from 1 to 10), but placed more value on increasing the effectiveness of checks with an average of 9.2 on a 1-10 scale. Four major themes were identified: (1) weekly chart checks need to adapt to an electronic record-and-verify chart environment, (2) physicists could add value to patient care by analyzing images without duplicating the work done by physicians, (3) greater support for trending analysis is needed in weekly checks, and (4) automation has the potential to increase the value of physics checks.

Conclusion

This study identified several key shortcomings of the current weekly chart check process from the perspective of the clinical medical physicist. Our results show strong support for automating components of the weekly check workflow in order to allow for more effective checks that emphasize follow-up, trending, failure modes and effects analysis, and allow time to be spent on other higher value tasks that improve patient safety.

Objective. In image-guided radiotherapy (IGRT), off-by-one vertebral body misalignments are rare but potentially catastrophic. In this study, a novel detection method for such misalignments in IGRT was investigated using densely-connected convolutional networks (DenseNets) for applications towards real-time error prevention and retrospective error auditing.Approach. A total of 4213 images acquired from 527 radiotherapy patients aligned with planar kV or MV radiographs were used to develop and test error-detection software modules. Digitally reconstructed radiographs (DRRs) and setup images were retrieved and co-registered according to the clinically applied alignment contained in the DICOM REG files. A semi-automated algorithm was developed to simulate patient positioning errors on the anterior-posterior (AP) and lateral (LAT) images shifted by one vertebral body. A DenseNet architecture was designed to classify either AP images individually or AP and LAT image pairs. Receiver-operator characteristic curves (ROC) and areas under the curves (AUC) were computed to evaluate the classifiers on test subsets. Subsequently, the algorithm was applied to the entire dataset in order to retrospectively determine the absolute off-by-one vertebral body error rate for planar radiograph guided RT at our institution from 2011-2021.Main results. The AUCs for the kV models were 0.98 for unpaired AP and 0.99 for paired AP-LAT. The AUC for the MV AP model was 0.92. For a specificity of 95%, the paired kV model achieved a sensitivity of 99%. Application of the model to the entire dataset yielded a per-fraction off-by-one vertebral body error rate of 0.044% [0.0022%, 0.21%] for paired kV IGRT including one previously unreported error.Significance. Our error detection algorithm was successful in classifying vertebral body positioning errors with sufficient accuracy for retrospective quality control and real-time error prevention. The reported positioning error rate for planar radiograph IGRT is unique in being determined independently of an error reporting system.

Objective

A recently proposed "5DCT" protocol uses deformable registration of free-breathing fast-helical CT scans to generate a breathing motion model. In order to allow accurate registration, free-breathing images are required to be free of doubling-artefacts, which arise when tissue motion is greater than scan speed.

Methods

Using a unique set of digital phantoms based on patient data and verified with a motion phantom, this work identifies the minimum scanner parameters required to successfully generate free-breathing artefact-free fast-helical scans. A motion phantom and 5 patients were imaged 25 times under free-breathing conditions in alternating directions with a 64-slice CT scanner employing a low-dose fast-helical protocol. A series of high temporal resolution (0.1 s) 5DCT scan data sets was generated in each case. A simulated CT scanner was used to "image" each free-breathing data set. Various CT scanner detector widths and rotation times were simulated, and verified using the motion phantom results. Motion-induced artefacts were quantified in patient images using structural similarity maps to determine the similarity between axial slices.

Results

Increasing amounts of motion-induced artefacts were observed with increasing rotation times >0.2 s for 16 mm detector configuration.

Conclusion

The current generation of 16-slice CT scanners, which are present in the majority of Radiation Oncology departments, are not capable of generating free-breathing sorting artefact-free images required for 5DCT. Advances in knowledge: A recently proposed "5DCT" protocol uses deformable registration of free-breathing fast-helical CT scans to generate a breathing motion model. In order to allow accurate registration, free-breathing images are required to be free of doubling-artefacts, which arise when tissue motion is greater than scan speed. The results suggest that the current generation of 16-slice CT scanners, present in the majority of Radiation Oncology departments, are not capable of generating the free-breathing images required for 5DCT.

Purpose

Present knowledge of patient setup and alignment errors in image guided radiation therapy (IGRT) relies on voluntary reporting, which is thought to underestimate error frequencies. A manual retrospective patient-setup misalignment error search is infeasible owing to the bulk of cases to be reviewed. We applied a deep learning-based misalignment error detection algorithm (EDA) to perform a fully automated retrospective error search of clinical IGRT databases and determine an absolute gross patient misalignment error rate.

Methods and materials

The EDA was developed to analyze the registration between planning scans and pretreatment cone beam computed tomography scans, outputting a misalignment score ranging from 0 (most unlikely) to 1 (most likely). The algorithm was trained using simulated translational errors on a data set obtained from 680 patients treated at 2 radiation therapy clinics between 2017 and 2022. A receiver operating characteristic analysis was performed to obtain target thresholds. DICOM Query and Retrieval software was integrated with the EDA to interact with the clinical database and fully automate data retrieval and analysis during a retrospective error search from 2016 to 2017 and from 2021 to 2022 for the 2 institutions, respectively. Registrations were flagged for human review using both a hard-thresholding method and a prediction trending analysis over each individual patient's treatment course. Flagged registrations were manually reviewed and categorized as errors (>1 cm misalignment at the target) or nonerrors.

Results

A total of 17,612 registrations were analyzed by the EDA, resulting in 7.7% flagged events. Three previously reported errors were successfully flagged by the EDA, and 4 previously unreported vertebral body misalignment errors were discovered during case reviews. False positive cases often displayed substantial image artifacts, patient rotation, and soft tissue anatomy changes.

Conclusions

Our results validated the clinical utility of the EDA for bulk image reviews and highlighted the reliability and safety of IGRT, with an absolute gross patient misalignment error rate of 0.04% ± 0.02% per delivered fraction.

Purpose

The purpose of this work is to develop a cardiac-induced lung motion model to be integrated into an existing breathing motion model.

Methods

The authors' proposed cardiac-induced lung motion model represents the lung tissue's specific response to the subject's cardiac cycle. The model is mathematically defined as a product of a converging polynomial function h of the cardiac phase (c) and the maximum displacement y(X0) of each voxel (X0) among all the cardiac phases. The function h(c) was estimated from cardiac-gated MR imaging of ten healthy volunteers using an Akaike Information Criteria optimization algorithm. For each volunteer, a total of 24 short-axis and 18 radial planar views were acquired on a 1.5 T MR scanner during a series of 12-15 s breath-hold maneuvers. Each view contained 30 temporal frames of equal time-duration beginning with the end-diastolic cardiac phase. The frames in each of the planar views were resampled to create a set of three-dimensional (3D) anatomical volumes representing thoracic anatomy at different cardiac phases. A 3D multiresolution optical flow deformable image registration algorithm was used to quantify the difference in tissue position between the end-diastolic cardiac phase and the remaining cardiac phases. To account for image noise, voxel displacements whose maximum values were less than 0.3 mm, were excluded. In addition, the blood vessels were segmented and excluded in order to eliminate registration artifacts caused by blood-flow.

Results

The average cardiac-induced lung motions for displacements greater than 0.3 mm were found to be 0.86 ± 0.74 and 0.97 ± 0.93 mm in the left and right lungs, respectively. The average model residual error for the ten healthy volunteers was found to be 0.29 ± 0.08 mm in the left lung and 0.38 ± 0.14 mm in the right lung for tissue displacements greater than 0.3 mm. The relative error decreased with increasing cardiac-induced lung tissue motion. While the relative error was > 60% for submillimeter cardiac-induced lung tissue motion, the relative error decreased to < 5% for cardiac-induced lung tissue motion that exceeded 10 mm in displacement.

Conclusions

The authors' studies implied that modeling and including cardiac-induced lung motion would improve breathing motion model accuracy for tissues with cardiac-induced motion greater than 0.3 mm.