Developers and users of artificial-intelligence-based tools for automatic contouring and treatment planning in radiotherapy are expected to assess clinical acceptability of these tools. However, what is ‘clinical acceptability’? Quantitative and qualitative approaches have been used to assess this ill-defined concept, all of which have advantages and disadvantages or limitations. The approach chosen may depend on the goal of the study as well as on available resources. In this paper, we discuss various aspects of ‘clinical acceptability’ and how they can move us toward a standard for defining clinical acceptability of new autocontouring and planning tools.

Automatic review of breast plan quality for clinical trials is time-consuming and has some unique challenges due to the lack of target contours for some planning techniques. We propose using an auto-contouring model and statistical process control to independently assess planning consistency in retrospective data from a breast radiotherapy clinical trial. A deep learning auto-contouring model was created and tested quantitatively and qualitatively on 104 post-lumpectomy patients’ computed tomography images (nnUNet; train/test: 80/20). The auto-contouring model was then applied to 127 patients enrolled in a clinical trial. Statistical process control was used to assess the consistency of the mean dose to auto-contours between plans and treatment modalities by setting control limits within three standard deviations of the data’s mean. Two physicians reviewed plans outside the limits for possible planning inconsistencies. Mean Dice similarity coefficients comparing manual and auto-contours was above 0.7 for breast clinical target volume, supraclavicular and internal mammary nodes. Two radiation oncologists scored 95% of contours as clinically acceptable. The mean dose in the clinical trial plans was more variable for lymph node auto-contours than for breast, with a narrower distribution for volumetric modulated arc therapy than for 3D conformal treatment, requiring distinct control limits. Five plans (5%) were flagged and reviewed by physicians: one required editing, two had clinically acceptable variations in planning, and two had poor auto-contouring. An automated contouring model in a statistical process control framework was appropriate for assessing planning consistency in a breast radiotherapy clinical trial.

Background: Automation in radiotherapy presents a promising solution to the increasing cancer burden and workforce shortages. However, existing automated methods for breast radiotherapy lack a comprehensive, end‐to‐end solution that meets varying standards of care. Purpose: This study aims to develop a complete portfolio of automated radiotherapy treatment planning for intact breasts, tailored to individual patient factors, clinical approaches, and available resources. Methods: We developed five automated conventional treatment approaches and utilized an established RapidPlan model for volumetric arc therapy. These approaches include conventional tangents for whole breast treatment, two variants for supraclavicular nodes (SCLV) treatment with/without axillary nodes, and two options for comprehensive regional lymph nodes treatment. The latter consists of wide tangents photon fields with a SCLV field, and a photon tangents field with a matched electron field to treat the internal mammary nodes (IMNs), and a SCLV field. Each approach offers the choice of a single or two isocenter setup (with couch rotation) to accommodate a wide range of patient sizes. All algorithms start by automatically generating contours for breast clinical target volume, regional lymph nodes, and organs at risk using an in‐house nnU‐net deep learning models. Gantry angles and field shapes are then automatically generated and optimized to ensure target coverage while limiting the dose to nearby organs. The dose is optimized using field weighting for the lymph nodes fields and an automated field‐in‐field approach for the tangents. These algorithms were integrated into the RayStation treatment planning system and tested for clinical acceptability on 15 internal whole breast patients (150 plans) and 40 external patients from four different institutions in Switzerland, Argentina, Iran, and the USA (360 plans). Evaluation criteria included ensuring adequate coverage of targets and adherence to dose constraints for normal structures. A breast radiation oncologist reviewed the single institution dataset for clinical acceptability (5‐point scale) and a physicist evaluated the multi‐institutional dataset (use as is or edit). Results: The dosimetric evaluation across all datasets (510 plans) showed that 100% of the automated plans met the dose coverage requirements for the breast, 99% for the SCLV, 98% for the axillary nodes, and 91% for the IMN. As expected, hot spots were more prevalent when multiple fields were combined. For the heart, ipsilateral lung, and contralateral breast, automated plans met constraints for 95%, 92%, and 95% of the plans, respectively. Physician evaluation of the 15 internal patients indicated that all automated plans were clinically acceptable with minor edits. Notably, the use of automated contours with the RapidPlan model resulted in plans that were immediately ready for use in 73% of cases (95% confidence interval, 95% CI [51‐ 96]) of patients, with the remaining cases requiring minor stylistic edits. Similarly, the physicist's review of the 40 multi‐institution patients showed that the auto‐plans were ready for use 79% (95% CI [73,85]) of the time (95% CI [73,85]), with edits needed for the remaining cases. Conclusion: This study demonstrates the feasibility of a comprehensive automated treatment planning model for whole breast radiotherapy, effectively accommodating diverse treatment paradigms.

Background/Objectives: We assessed the influence of local patients and clinical characteristics on the performance of commercial deep learning (DL) segmentation models for head-and-neck (HN), breast, and prostate cancers. Methods: Clinical computed tomography (CT) scans and clinically approved contours of 210 patients (53 HN, 49 left breast, 55 right breast, and 53 prostate cancer) were used to train and validate segmentation models integrated within a vendor-supplied DL training toolkit and to assess the performance of both vendor-pretrained and custom-trained models. Four custom models (HN, left breast, right breast, and prostate) were trained and validated with 30 (training)/5 (validation) HN, 34/5 left breast, 39/5 right breast, and 30/5 prostate patients to auto-segment a total of 24 organs at risk (OARs). Subsequently, both vendor-pretrained and custom-trained models were tested on the remaining patients from each group. Auto-segmented contours were evaluated by comparing them with clinically approved contours via the Dice similarity coefficient (DSC) and mean surface distance (MSD). The performance of the left and right breast models was assessed jointly according to ipsilateral/contralateral locations. Results: The average DSCs for all structures in vendor-pretrained and custom-trained models were as follows: 0.81 ± 0.12 and 0.86 ± 0.11 in HN; 0.67 ± 0.16 and 0.80 ± 0.11 in the breast; and 0.87 ± 0.09 and 0.92 ± 0.06 in the prostate. The corresponding average MSDs were 0.81 ± 0.76 mm and 0.76 ± 0.56 mm (HN), 4.85 ± 2.44 mm and 2.42 ± 1.49 mm (breast), and 2.17 ± 1.39 mm and 1.21 ± 1.00 mm (prostate). Notably, custom-trained models showed significant improvements over vendor-pretrained models for 14 of 24 OARs, reflecting the influence of data/contouring variations in segmentation performance. Conclusions: These findings underscore the substantial impact of institutional preferences and clinical practices on the implementation of vendor-pretrained models. We also found that a relatively small amount of institutional data was sufficient to train customized segmentation models with sufficient accuracy.

In this study, we aimed to enhance the contouring accuracy of cardiac pacemakers by improving their visualization using deep learning models to predict MV CBCT images based on kV CT or CBCT images. Ten pacemakers and four thorax phantoms were included, creating a total of 35 combinations. Each combination was imaged on a Varian Halcyon (kV/MV CBCT images) and Siemens SOMATOM CT scanner (kV CT images). Two generative adversarial network (GAN)-based models, cycleGAN and conditional GAN (cGAN), were trained to generate synthetic MV (sMV) CBCT images from kV CT/CBCT images using twenty-eight datasets (80%). The pacemakers in the sMV CBCT images and original MV CBCT images were manually delineated and reviewed by three users. The Dice similarity coefficient (DSC), 95% Hausdorff distance (HD95), and mean surface distance (MSD) were used to compare contour accuracy. Visual inspection showed the improved visualization of pacemakers on sMV CBCT images compared to original kV CT/CBCT images. Moreover, cGAN demonstrated superior performance in enhancing pacemaker visualization compared to cycleGAN. The mean DSC, HD95, and MSD for contours on sMV CBCT images generated from kV CT/CBCT images were 0.91 ± 0.02/0.92 ± 0.01, 1.38 ± 0.31 mm/1.18 ± 0.20 mm, and 0.42 ± 0.07 mm/0.36 ± 0.06 mm using the cGAN model. Deep learning-based methods, specifically cycleGAN and cGAN, can effectively enhance the visualization of pacemakers in thorax kV CT/CBCT images, therefore improving the contouring precision of these devices.

Computed Tomography (CT) scans are the standard-of-care for the visualization and diagnosis of many clinical ailments, and are needed for the treatment planning of external beam radiotherapy. Unfortunately, the availability of CT scanners in low- and mid-resource settings is highly variable. Planar x-ray radiography units, in comparison, are far more prevalent, but can only provide limited 2D observations of the 3D anatomy. In this work we propose DIFR3CT, a 3D latent diffusion model, that can generate a distribution of plausible CT volumes from one or few (<10) planar x-ray observations. DIFR3CT works by fusing 2D features from each x-ray into a joint 3D space, and performing diffusion conditioned on these fused features in a low-dimensional latent space. We conduct extensive experiments demonstrating that DIFR3CT is better than recent sparse CT reconstruction baselines in terms of standard pixel-level (PSNR, SSIM) on both the public LIDC and in-house post-mastectomy CT datasets. We also show that DIFR3CT supports uncertainty quantification via Monte Carlo sampling, which provides an opportunity to measure reconstruction reliability. Finally, we perform a preliminary pilot study evaluating DIFR3CT for automated breast radiotherapy contouring and planning -- and demonstrate promising feasibility. Our code is available at https://github.com/yransun/DIFR3CT.