Comparison of Ethos template-based planning and AI-based dose prediction: General performance, patient optimality, and limitations

Anticipating potential bottlenecks in adaptive proton FLASH therapy: a ridge filter reuse strategy

Objective.Achieving FLASH dose rate with pencil beam scanning intensity modulated proton therapy is challenging. However, utilizing a single energy layer with a ridge filter (RF) can maintain dose rate and conformality. Yet, changes in patient anatomy over the treatment course can render the RF obsolete. Unfortunately, creating a new RF is time-consuming, thus, incompatible with online adaptation. To address this, we propose to re-optimize the spot weights while keeping the same initial RF.Approach.Data from six head and neck cancer patients with a repeated computed tomography (CT2) were used. FLASH treatment plans were generated with three methods on CT2: 'full-adaptation' (FA), optimized from scratch with a new RF; 'spot-adaptation only' (SAO), re-using initial RF but adjusting plan spot weights; and 'no adaptation' (NoA) where the dose from initial plans on initial CT (CT1) was recomputed on CT2. The prescribed dose per fraction was 9 Gy. Different beam angles were tested for each CT2(1 beam per fraction). The FA, SAO and NoA plans were then compared on CT2.Main results.Fractions with SAO showed a median decrease of 0.05 Gy forD98% and a median increase of 0.03 Gy forD2% of CTV when compared to their homologous FA plans on nominal case. Median conformity number decreased by 0.03. Median max dose to spinal cord increased by 0.09 Gy. The largest median increase in mean dose to organs was 0.03 Gy to the mandible. The largest observed median difference in organs receiving a minimal dose rate of 40 Gy s-1was 0.5% for the mandible. Up to 16 of the 20 evaluated SAO fractions were thus deemed clinically acceptable, with up to 8 NoA plans already acceptable before adaptation.Significance.Proposed SAO workflow showed that for most of our evaluated plans, daily reprinting of RF was not necessary.

Deep learning dose prediction to approach Erasmus-iCycle dosimetric plan quality within seconds for instantaneous treatment planning

BACKGROUND AND PURPOSE

Fast, high-quality deep learning (DL) prediction of patient-specific 3D dose distributions can enable instantaneous treatment planning (IP), in which the treating physician can evaluate the dose and approve the plan immediately after contouring, rather than days later. This would greatly benefit clinical workload, patient waiting times and treatment quality. IP requires that predicted dose distributions closely match the ground truth. This study examines how training dataset size and model size affect dose prediction accuracy for Erasmus-iCycle GT plans to enable IP.

MATERIALS AND METHODS

For 1250 prostate patients, dose distributions were automatically generated using Erasmus-iCycle. Hierarchically Densely Connected U-Nets with 2/3/4/5/6 pooling layers were trained with datasets of 50/100/250/500/1000 patients, using a validation set of 100 patients. A fixed test set of 150 patients was used for evaluations.

RESULTS

For all model sizes, prediction accuracy increased with the number of training patients, without levelling off at 1000 patients. For 4-6 level models with 1000 training patients, prediction accuracies were high and comparable. For 6 levels and 1000 training patients, the median prediction errors and interquartile ranges for PTV V95%, rectum V75Gy and bladder V65Gy were 0.01 [-0.06,0.15], 0.01 [-0.20,0.29] and -0.02 [-0.27,0.27] %-point. Dose prediction times were around 1.2 s.

CONCLUSION

Although even for 1000 training patients there was no convergence in obtained prediction accuracy yet, the accuracy for the 6-level model with 1000 training patients may be adequate for the pursued instantaneous planning, which is subject of further research.

Towards faster plan adaptation for proton arc therapy using initial treatment plan information

Background and Purpose

Proton arc therapy (PAT) is an emerging modality delivering continuously rotating proton beams. Current PAT planning approaches are time-consuming, making them unsuitable for online adaptation. This study proposes an accelerated workflow for adapting PAT plans.

Materials and Methods

The proposed workflow transfers spots from initial computed tomography (CT) to the CT of the day, updates energy layers considering the initial pattern, and re-optimizes selected transferred spots based on their initial weights and impact on the objective function.A retrospective study was conducted on five head and neck patients who underwent plan adaptation on a repeated CT. PAT plans were generated with two different methods on the repeated CT: reference, created de novo, and smart-adapted, generated with the proposed adaptive workflow. Robust optimization was performed for all plans.

Results

Smart-adapted plans achieved similar mean dose to organs at risk as the reference: the largest median increase of mean dose was 1.9 Gy to the mandible; the median of maximum dose to spinal cord was 0.5 Gy lower for the smart-adapted plans. The median target coverage, i.e. D98, to primary tumor and nodes of smart-adapted plans decreased by 0.2 and 0.4 Gy for the nominal case, and 0.4 and 0.6 Gy for the worst-case scenario; all smart-adapted plans met clinical objectives. The smart-adaptation method reduced average planning time from 19184 s to 5626 s, a 3.4-fold improvement.

Conclusions

Smart-adapted plans achieve similar plan quality to the reference method, while significantly reducing plan generation time for new patient anatomy.

Artificial Intelligence for Radiation Treatment Planning: Bridging Gaps From Retrospective Promise to Clinical Reality

Artificial intelligence (AI) radiation therapy (RT) planning holds promise for enhancing the consistency and efficiency of the RT planning process. Despite technical advancements, the widespread integration of AI into RT treatment planning faces challenges. The transition from controlled retrospective environments to real-world clinical settings introduces heightened scrutiny from clinical end users, potentially leading to decreased clinical acceptance. Key considerations for implementing AI RT planning include ensuring the AI model performance aligns with clinical standards, using high-quality training data, and incorporating sufficient data variation through meticulous curation by clinical experts. Beyond technical aspects, factors such as potential biases and the level of trust clinical end users place in AI may present unforeseen obstacles for real-world clinical use. Addressing these challenges requires bridging education and expertise gaps among clinical end users, enabling them to confidently embrace and utilize AI for routine RT planning. By fostering a better understanding of AI capabilities, building trust, and providing comprehensive training, the promises of AI RT planning can be a reality in the clinical setting. This article assesses the current clinical use of AI RT planning and explores challenges and considerations for bridging gaps in knowledge and expertise for AI operationalization, with focus on training data curation, workflow integration, explainability, bias, and domain knowledge. Remaining challenges in clinical implementation of AI RT treatment planning are examined in the context of trust building approaches.

Investigating the impact of rapidplan on ethos automated planning

Automated planning has surged in popularity within external beam radiation therapy in recent times. Leveraging insights from previous clinical knowledge could enhance auto-planning quality. In this work, we evaluated the performance of Ethos automated planning with knowledge-based guidance, specifically using Rapidplan (RP). Seventy-four patients with head-and-neck (HN) cancer and 37 patients with prostate cancer were used to construct separate RP models. Additionally, 16 patients from each group (HN and prostate) were selected to assess the performance of Ethos auto-planning results. Initially, a template-based Ethos plan (Non-RP plan) was generated, followed by integrating the corresponding RP model's DVH estimates into the optimization process to generate another plan (RP plan). We compared the target coverage, OAR doses, and total monitor units between the non-RP and RP plans. Both RP and non-RP plans achieved comparable target coverage in HN and Prostate cases, with a negligible difference of less than 0.5% (p > 0.2). RP plans consistently demonstrated lower doses of OARs in both HN and prostate cases. Specifically, the mean doses of OARs were significantly reduced by 9% (p < 0.05). RP plans required slightly higher monitor units in both HN and prostate sites (p < 0.05), however, the plan generation time was almost similar (p > 0.07). The inclusion of the RP model reduced the OAR doses, particularly reducing the mean dose to critical organs compared to non-RP plans while maintaining similar target coverage. Our findings provide valuable insights for clinics adopting Ethos planning, potentially enhancing the auto-planning to operate optimally.

RadDeploy: A framework for integrating in-house developed software and artificial intelligence models seamlessly into radiotherapy workflows

The use of and research in automation and artificial intelligence (AI) in radiotherapy is moving with incredible pace. Many innovations do, however, not make it into the clinic. One technical reason for this may be the lack of a platform to deploy such software into clinical practice. We suggest RadDeploy as a framework for integrating containerized software in clinical workflows outside of treatment planning systems. RadDeploy supports multiple DICOM as input for model containers and can run model containers asynchronously across GPUs and computers. This technical note summarizes the inner workings of RadDeploy and demonstrates three use-cases with varying complexity.

Evaluation of Ethos intelligent optimization engine for left locally advanced breast cancer

Purpose

To evaluate the feasibility to use a standard Ethos planning template to treat left-sided breast cancer with regional lymph nodes.

Material/Methods

The tuning cohort of 5 patients was used to create a planning template. The validation cohort included 15 patients treated for a locally advanced left breast cancer randomly enrolled. The Ethos planning template was tuned using standard 3 partial arc VMAT and two collimator rotation configurations: 45/285/345° and 30/60/330°. Re-planning was performed automatically using the template without editing. The study was conducted with a schedule of 42.3 Gy in 18 fractions to the breast/chestwall, internal mammary chain (IMC) and regional lymph nodes ("Nodes"). The PTV was defined as a 3D extension of the CTV with a margin of 7 mm, excluding the 5mm below the skin. The manual treatment plans were performed using Eclipse treatment planning system with AAA and PO algorithms (v15.6) and a manual arc VMAT configuration and imported in Ethos TPS (v1.1) for a dose calculation with Ethos Acuros algorithm. The automated plans were compared with the manual plans using PTV and CTV coverage, homogeneity and conformity indices (HI and CN) and doses to organs at risk (OAR) via DVH metrics. For each plan, the patient quality assurance (QA) were performed using Mobius3D and gamma index. Finally, two breast radiation oncologists performed a blinded assessment of the clinical acceptability of each of the three plans (manual and automated) for each patient.

Results

The manual and automated plans provided suitable treatment planning as regards dose constraints. The dosimetric comparison showed the CTV_breast D99% were significantly improved with both automated plans (p< 0,002) while PTV coverage was comparable. The doses to the organs at risk were equivalent for the three plans. Concerning treatment delivery, the Ethos-45° and Ethos-30° plans led to an increase in MUs compared to the manual plans, without affecting the beam on time. The average gamma index pass rates remained consistently above 98% regardless of the type of plan utilized. In the blinded evaluation, clinicians 1 and 2 assessed 13 out of 15 plans for Ethos 45° and 11 out of 15 plans for Ethos 30° as clinically acceptable.

Conclusion

Using a standard planning template for locally advanced breast cancer, the Ethos TPS provided automated plans that were clinically acceptable and comparable in quality to manually generated plans. Automated plans also dramatically reduce workflow and operator variability.

A retrospective study on the investigation of potential dosimetric benefits of online adaptive proton therapy for head and neck cancer

PURPOSE

Proton therapy is sensitive to anatomical changes, often occurring in head-and-neck (HN) cancer patients. Although multiple studies have proposed online adaptive proton therapy (APT), there is still a concern in the radiotherapy community about the necessity of online APT. We have performed a retrospective study to investigate the potential dosimetric benefits of online APT for HN patients relative to the current offline APT.

METHODS

Our retrospective study has a patient cohort of 10 cases. To mimic online APT, we re-evaluated the dose of the in-use treatment plan on patients' actual treatment anatomy captured by cone-beam CT (CBCT) for each fraction and performed a templated-based automatic replanning if needed, assuming that these were performed online before treatment delivery. Cumulative dose of the simulated online APT course was calculated and compared with that of the actual offline APT course and the designed plan dose of the initial treatment plan (referred to as nominal plan). The ProKnow scoring system was employed and adapted for our study to quantify the actual quality of both courses against our planning goals.

RESULTS

The average score of the nominal plans over the 10 cases is 41.0, while those of the actual offline APT course and our simulated online course is 25.8 and 37.5, respectively. Compared to the offline APT course, our online course improved dose quality for all cases, with the score improvement ranging from 0.4 to 26.9 and an average improvement of 11.7.

CONCLUSION

The results of our retrospective study have demonstrated that online APT can better address anatomical changes for HN cancer patients than the current offline replanning practice. The advanced artificial intelligence based automatic replanning technology presents a promising avenue for extending potential benefits of online APT.