Guang Li

Introduction: Although long short-term memory (LSTM) networks have been tested to predict short-term respiratory motion, their performance in long-term forecasting under breathing irregularities must be assessed. We aim to evaluate and enhance the long-term prediction of internal motion from an external surrogate using subject-specific LSTM models through a hybrid, adaptive approach. Methods: Concurrent internal navigator and external bellows respiratory-motion waveforms were acquired for ten volunteers during two four-dimensional magnetic resonance imaging (4DMRI) scans lasting 3–10 min each. Approximately 20 min intervened between the first (mid-term) and second (long-term) scan. After training on the first half of the mid-term data, subject-specific LSTM models were applied to the remaining mid-term and entire long-term datasets to predict internal waveforms. The accuracy of a model's prediction was assessed with Pearson's correlation ( C ), referenced to the native waveforms, maximized through the time-domain cross-correlation (TCC), and enhanced by correcting residual phase shifts in the LSTM models using a hybrid (LSTM-TCC) approach. Hyperparameter selection by minimizing the root mean square error (RMSE) to identify high-performance ( C ≥ 0.8) LSTM models was evaluated by the area under the receiver operating characteristic curve (AUROC). The temporal accuracy of inspiratory-peak predictions was characterized. Results: Compared to the native waveforms ( C = 0.42 ± 0.28) and TCC method ( C = 0.77 ± 0.09), the LSTM models yielded more accurate predictions ( C = 0.89 ± 0.07) in the mid-term scans. Over 20–30 min, LSTM predictions faltered ( C < 0.80) in two subjects but were rescued by LSTM-TCC ( C = 0.90 ± 0.09). The temporal error in predicting inspiratory peaks was smaller for LSTM-TCC (Δ t = 0.15 ± 0.11sec) than LSTM (Δ t = 0.18 ± 0.15sec). RMSE reliably identified high-performance models: A U R O C L S T M m i d − t e r m = 0.82 , A U R O C L S T M l o n g − t e r m = 0.74 , and A U R O C h y b r i d l o n g − t e r m = 0.83 . Conclusion: The feasibility of a novel adaptive subject-specific LSTM-TCC modeling was tested in 10 subjects, demonstrating that high accuracy of external-to-internal motion predictions in 3–10 min can be extended to 30 min overcoming breathing irregularities without remodeling. Further investigations of the adaptive LSTM-TCC model are warranted as a potential clinical solution.

Purpose: This study aimed to commission the AlignRT inBore system in Ethos/Halcyon linear accelerator (Linac) systems for surface‐guided radiotherapy (SGRT), including developing a soft collision quality assurance (QA) to pass daily AlignRT QA without daily calibration, ensuring compatibility between SGRT and image‐guided radiotherapy (IGRT) setups, establishing a baseline performance in AlignRT motion monitoring, and enabling breast DIBH SGRT using manual beam gating with timing accuracy assessment. Methods: The AlignRT inBore system contains three ceiling‐mounted camera pods for SGRT setup at the simulation isocenter (sim‐ISO), and two ring‐mounted camera sets on the spring‐supported bore cover for motion monitoring at treatment ISO (txt‐ISO), after iterative cone‐beam computed tomography (iCBCT) setup. The commissioning includes confirming the ring cameras are outside the radiation fields, verifying IEC coordinate consistency between SGRT and IGRT, and ensuring the compatibility of AlignRT daily QA with collision QA. A dot‐array plate was used for SGRT calibration and daily QA at both ISO positions. A soft collision check method was developed to minimize the bore‐cover displacement by (1) triggering the collision alarm from the rear side, distant from the front ring cameras, and (2) reproducing the rear bore‐cover position by marking before the collision check. Clinical staff were trained for the new daily QA procedure, and the average period between calibrations was calculated over the past 12 months. Additionally, the application‐specific breast DIBH SGRT procedure was commissioned through an end‐to‐end test with manual gating, which was retrospectively evaluated for timing accuracy. Results: The AlignRT inBore camera ring was moved 2 cm away from the iCBCT field toward the bore front, and the IEC 61217 coordinate system yields consistent SGRT and IGRT setups. AlignRT inBore daily QA can pass using the soft collision procedure without daily calibration, with an average calibration period of 2 weeks, as ad‐hoc calibrations are required after open‐bore services. The manual gating accuracy is found to be 0.6s ± 0.2s, which can be further reduced if the beam can be proactively turned off before coaching patients to relax from a DIBH. Conclusion: The commissioning of the AlignRT inBore systems was successfully implemented, including developing and implementing the soft collision QA without daily calibration for 2 weeks, updating inBore camera position (2 cm to the bore front), ensuring SGRT‐IGRT couch‐shift consistency, establishing a baseline performance (< 0.2 mm drift), and allowing SGRT applications on Ethos/Halcyon systems. The application‐specific breast DIBH SGRT procedure with manual beam gating is also commissioned with a new retrospective timing analysis for assessing gating accuracy (0.6s ± 0.2s).

Purpose: To improve the setup reproducibility of neck curvature using real-time optical surface imaging (OSI) guidance on 2 regions of interest (ROIs) to infer cervical spine (c-spine) curvature for surface-guided radiotherapy (SGRT) of head-and-neck (HN) and c-spine cancer. Methods: A novel SGRT setup approach was designed to reproduce neck curvature with 2 ROIs: upper-chest ROI and open-face ROI. It was hypothesized that the neck curvature could be reproduced if both ROIs were aligned within ±3 mm/2˚ tolerance. This was tested prospectively in 7 volunteers using real-time 3D-OSI guidance and lateral 2D-photography verification after the 3D and 2D references were captured from the initial conventional setup. Real-time SGRT was performed to align chest-ROI and face-ROI, and the longitudinal distance between them was adjustable using a head-support slider. Verification of neck curvature anteriorly and posteriorly was achieved by overlaying edge-extracted lateral pictures. Retrospectively, the relationship between anterior surface and spinal canal alignment was checked in 11 patients using their simulation CT (simCT) and setup cone-beam CT (CBCT). After the anterior surface was rigidly aligned, the spinal canal alignment was checked and quantified using the mean-distance-to-agreement (MDA) and DICE similarity index, and surface-to-spine correlation was calculated. Results: The reproducibility of neck curvatures using the 2xROI SGRT setup is verified and the mean neck-outline-matching difference is within ±2 mm in lateral photographic overlays. The chest-ROI alignment takes 110 ± 58 s and the face-ROI takes 60 ± 35 s. When the anterior body surface is aligned (MDA = 1.1 ± 0.6 mm, DICE = 0.96 ± 0.02,) the internal spinal canal is also aligned (MDA = 1.0 ± 0.3 mm, DICE = 0.84 ± 0.04) in 11 patients. The surface-to-spine correlation is c = 0.90 (MDA) and c = 0.85 (DICE). Conclusion: This study demonstrates the feasibility of the novel 2-ROI SGRT setup technique to achieve reproducible neck and c-spine curvature regardless of neck visibility and availability as ROI. Staff training is needed to adopt this unconventional SGRT technique to improve patient setup.

To investigate automation of the preparation of the region of interest (ROI) for surface‐guided radiotherapy (SGRT) of the whole breast with two algorithms based on contour anatomies: using the body contour, and using the breast contour. The patient dataset used for modeling consisted of 39 breast cancer patients previously treated with SGRT. The patient's anatomical structures (body and ipsilateral breast) were retrieved from the planning system, and the clinical ROI (cROI) drawn by the planners was retrieved from the SGRT system for comparison. For the body‐contour‐based algorithm, a convolutional neural network (MobileNet‐v2) was utilized to train a synthetic human model dataset to predict body joint locations. With the body joint location knowledge, an automated ROI (aROI^body) can be created based on: (1) the superior‐inferior (S‐I) borders defined by the joint locations, (2) the left‐right (L‐R) borders defined with 3/4 of chest width, and (3) a curation of the ROI to avoid the ipsilateral armpit. For the breast‐contour‐based algorithm, an aROI^breast was created by first defining the ROI in the S‐I direction with the ipsilateral breast boundaries. Other steps are the same as with the body‐contour‐based algorithm. Among the 39 patients, 24 patients were used to fine‐tune the algorithm parameters, and the remaining 15 patients were used to evaluate the quality of the aROIs against the cROIs. A blinded evaluation was performed by three SGRT expert physicists to rate the acceptability and the quality (1–10 scale) of the aROIs and cROIs, and the dice similarity coefficient (DSC) was also calculated to compare the similarity between the aROIs and cROIs. The results showed that the average acceptability was 14/15 (range: 13/15–15/15) for cROIs, 13.3/15 (range: 13/15–14/15) for aROI^body, and 14.6/15 (range: 14/15–15/15) for aROI^breast. The average quality was 7.4 ± 0.8 for cROIs, 8.1 ± 1.2 for aROI^body, and 8.2 ± 0.9 for aROI^breast. The DSC with cROIs was 0.81 ± 0.06 for aROI^body, and 0.83 ± 0.04 for aROI^breast. The ROI creation time was ∼120 s for clinical, 1.3 s for aROI^body, and 1.2 s for aROI^breast. The proposed automated algorithms can improve the ROI compliance with the SGRT protocol, with a shortened preparation time. It is ready to be integrated into the clinical workflow for automated ROI preparation.

To compare the setup accuracy of optical surface image (OSI) versus orthogonal x‐ray images (2DkV) using cone beam computed tomography (CBCT) as ground truth for radiotherapy of left breast cancer in deep‐inspiration breath‐hold (DIBH). Ten left breast DIBH patients treated with volumetric modulated arc therapy (VMAT) were studied retrospectively. OSI, 2DkV, and CBCT were acquired weekly at treatment setup. OSI, 2DkV, and CBCT were registered to planning CT or planning DRR based on a breast surface region of interest (ROI), bony anatomy (chestwall and sternum), and both bony anatomy and breast surface, respectively. These registrations provided couch shifts for each imaging system. The setup errors, or the difference in couch shifts between OSI and CBCT were compared to those between 2DkV and CBCT. A second OSI was acquired during last beam delivery to evaluate intrafraction motion. The median absolute setup errors were (0.21, 0.27, 0.23 cm, 0.6°, 1.3°, 1.0°) for OSI, and (0.26, 0.24, 0.18 cm, 0.9°, 1.0°, 0.6°) for 2DkV in vertical, longitudinal and lateral translations, and in rotation, roll and pitch, respectively. None of the setup errors was significantly different between OSI and 2DkV. For both systems, the systematic and random setup errors were ≤0.6 cm and ≤1.5° in all directions. Nevertheless, larger setup errors were observed in some sessions in both systems. There was no correlation between OSI and CBCT whereas there was modest correlation between 2DkV and CBCT. The intrafraction motion in DIBH detected by OSI was small with median absolute translations <0.2 cm, and rotations ≤0.4°. Though OSI showed comparable and small setup errors as 2DkV, it showed no correlation with CBCT. We concluded that to achieve accurate setup for both bony anatomy and breast surface, daily 2DkV can't be omitted following OSI for left breast patients treated with DIBH VMAT.

As patient breathing irregularities can introduce a large uncertainty in targeting the internal tumor volume (ITV) of lung cancer patients, and thereby affect treatment quality, this study evaluates dose tolerance of tumor motion amplitude variations in ITV-based volumetric modulated arc therapy (VMAT). A motion-incorporated planning technique was employed to simulate treatment delivery of 10 lung cancer patients’ clinical VMAT plans using original and three scaling-up (by 0.5, 1.0, and 2.0 cm) motion waveforms from single-breath four-dimensional computed tomography (4DCT) and multi-breath time-resolved 4D magnetic resonance imaging (TR-4DMRI). The planning tumor volume (PTV = ITV + 5 mm margin) dose coverage (PTV D95%) was evaluated. The repeated waveforms were used to move the isocenter in sync with the clinical leaf motion and gantry rotation. The continuous VMAT arcs were broken down into many static beam fields at the control points (2°-interval) and the composite plan represented the motion-incorporated VMAT plan. Eight motion-incorporated plans per patient were simulated and the plan with the native 4DCT waveform was used as a control. The first (D95% ≤ 95%) and second (D95% ≤ 90%) plan breaching points due to motion amplitude increase were identified and analyzed. The PTV D95% in the motion-incorporated plans was 99.4 ± 1.0% using 4DCT, closely agreeing with the corresponding ITV-based VMAT plan (PTV D95% = 100%). Tumor motion irregularities were observed in TR-4DMRI and triggered D95% ≤ 95% in one case. For small tumors, 4 mm extra motion triggered D95% ≤ 95%, and 6–8 mm triggered D95% ≤ 90%. For large tumors, 14 mm and 21 mm extra motions triggered the first and second breaching points, respectively. This study has demonstrated that PTV D95% breaching points may occur for small tumors during treatment delivery. Clinically, it is important to monitor and avoid systematic motion increase, including baseline drift, and large random motion spikes through threshold-based beam gating.

To define and evaluate the appropriate abdominal region of interest (ROI) as a surrogate of diaphragm positioning in deep‐inspiration breath‐hold (DIBH) for surface‐guided radiotherapy (SGRT) of abdominal cancers using 3D optical surface imaging (OSI). Six potential abdominal ROIs were evaluated to calculate their correlations with the diaphragm position using 4DCT images of 20 abdominal patients. Twelve points of interest (POIs) were defined (six on the central soft tissue and six on the bilateral ribs) at three superior–inferior levels, and different sub‐groups represented different ROIs. ROI‐1 was the largest, containing all 12 POIs from the xiphoid to the umbilicus and between the lateral body midlines while ROI‐2 had only eight inferior POIs, ROI‐3 had six lateral POIs, and ROI‐4 had four superior‐lateral POIs over the ribs, ROI‐5 contained six central and two most inferior‐lateral POIs and ROI‐6 contained six central and four inferior‐lateral POIs. Internally, the right diaphragm dome was used to represent its positions in 4DCT (0% and 50% within the cycle). The Pearson correlation coefficients were calculated between the diaphragm dome and all 12 external POIs individually or grouped as six ROIs. The quality of the abdominal ROIs was evaluated as potential internal surrogates and, therefore, potential ROIs for SGRT DIBH setup. The four most inferior POIs show the highest mean correlation (r = 0.75) with diaphragmatic motion, and the correlation decreases as POIs move superiorly. The mean correlations are the highest for ROIs with little or no rib support: r = 0.67 for ROI‐2, r = 0.64 for ROI‐5, and r = 0.63 for ROI‐6, while lower for ROIs with rib support: ROI‐1 has r = 0.60, ROI‐3 has r = 0.50, and ROI‐4 has only r = 0.28. This study demonstrates that the rectangular/triangular soft‐tissue ROI (with little rib support) is an optimal surrogate for body positioning and diaphragmatic motion, even when treating tumors under the rib cage. This evidence‐based ROI definition should be utilized when treating abdominal cancers with free‐breathing (FB) and/or DIBH setup.

This article reviews the recent advancements and future potential of optical surface imaging (OSI) in clinical applications as a four-dimensional (4D) imaging modality for surface-guided radiotherapy (SGRT), including OSI systems, clinical SGRT applications, and OSI-based clinical research. The OSI is a non-ionizing radiation imaging modality, offering real-time 3D surface imaging with a large field of view (FOV), suitable for in-room interactive patient setup, and real-time motion monitoring at any couch rotation during radiotherapy. So far, most clinical SGRT applications have focused on treating superficial breast cancer or deep-seated brain cancer in rigid anatomy, because the skin surface can serve as tumor surrogates in these two clinical scenarios, and the procedures for breast treatments in free-breathing (FB) or at deep-inspiration breath-hold (DIBH), and for cranial stereotactic radiosurgery (SRS) and radiotherapy (SRT) are well developed. When using the skin surface as a body-position surrogate, SGRT promises to replace the traditional tattoo/laser-based setup. However, this requires new SGRT procedures for all anatomical sites and new workflows from treatment simulation to delivery. SGRT studies in other anatomical sites have shown slightly higher accuracy and better performance than a tattoo/laser-based setup. In addition, radiographical image-guided radiotherapy (IGRT) is still necessary, especially for stereotactic body radiotherapy (SBRT). To go beyond the external body surface and infer an internal tumor motion, recent studies have shown the clinical potential of OSI-based spirometry to measure dynamic tidal volume as a tumor motion surrogate, and Cherenkov surface imaging to guide and assess treatment delivery. As OSI provides complete datasets of body position, deformation, and motion, it offers an opportunity to replace fiducial-based optical tracking systems. After all, SGRT has great potential for further clinical applications. In this review, OSI technology, applications, and potential are discussed since its first introduction to radiotherapy in 2005, including technical characterization, different commercial systems, and major clinical applications, including conventional SGRT on top of tattoo/laser-based alignment and new SGRT techniques attempting to replace tattoo/laser-based setup. The clinical research for OSI-based tumor tracking is reviewed, including OSI-based spirometry and OSI-guided tumor tracking models. Ongoing clinical research has created more SGRT opportunities for clinical applications beyond the current scope.