Can the Student Outperform the Master? A Plan Comparison Between Pinnacle Auto-Planning and Eclipse knowledge-Based RapidPlan Following a Prostate-Bed Plan Competition

Dosimetric evaluation of different planning strategies for hypofractionated whole-breast irradiation technique

Purpose.The dose hotspot areas in hypofractionated whole-breast irradiation (WBI) greatly increase the risk of acute skin toxicity because of the anatomical peculiarities of the breast. In this study, we presented several novel planning strategies that integrate multiple sub-planning target volumes (sub-PTVs), field secondary placement, and RapidPlan models for right-sided hypofractionated WBI.Methods.A total of 35 cases of WBI with a dose of 42.5 Gy for PTVs using tangential intensity-modulated radiotherapy (IMRT) were selected. Both PTVs were planned for simultaneous treatment using the original manual multiple sub-PTV plan (OMMP) and the original manual single-PTV plan (OMSP). The manual field secondary placement multiple sub-PTV plan (m-FSMP) with multiple objects on the original PTV and the manual field secondary placement single-objective plan (m-FSSP) were initially planned, which were distribution-based of V105 (volume receiving 105% of the prescription dose). In addition, two RapidPlan-based plans were developed, including the RapidPlan-based multiple sub-PTVs plan (r-FSMP) and the RapidPlan-based single-PTV plan (r-FSSP). Dosimetric parameters of the plans were compared, and V105 was evaluated using multivariate analysis to determine how it was related to the volume of PTV and the interval of lateral beam angles (ILBA).Results.The lowest mean V105 (5.64 ± 6.5%) of PTV was observed in m-FSMP compared to other manual plans. Upon validation, r-FSSP demonstrated superior dosimetric quality for OAR compared to the two other manual planning methods, except for V5(the volume of ipsilateral lung receiving 5 Gy) of the ipsilateral lung. While r-FSMP showed no significant difference (p = 0.06) compared to r-FSSP, it achieved the lowest V105 value (4.3 ± 4.5%), albeit with a slight increase in the dose to some OARs. Multivariate GEE linear regression showed that V105 is significantly correlated with target volume and ILBA.Conclusions.m-FSMP and r-FSMP can substantially enhance the homogeneity index (HI) and reduce V105, thereby minimizing the risk of acute skin toxicities, even though there may be a slight dose compromise for certain OARs.

A geometry and dose-volume based performance monitoring of artificial intelligence models in radiotherapy treatment planning for prostate cancer

Background and Purpose

Clinical Artificial Intelligence (AI) implementations lack ground-truth when applied on real-world data. This study investigated how combined geometrical and dose-volume metrics can be used as performance monitoring tools to detect clinically relevant candidates for model retraining.

Materials and Methods

Fifty patients were analyzed for both AI-segmentation and planning. For AI-segmentation, geometrical (Standard Surface Dice 3 mm and Local Surface Dice 3 mm) and dose-volume based parameters were calculated for two organs (bladder and anorectum) to compare AI output against the clinically corrected structure. A Local Surface Dice was introduced to detect geometrical changes in the vicinity of the target volumes, while an Absolute Dose Difference (ADD) evaluation increased focus on dose-volume related changes. AI-planning performance was evaluated using clinical goal analysis in combination with volume and target overlap metrics.

Results

The Local Surface Dice reported equal or lower values compared to the Standard Surface Dice (anorectum: (0.93 ± 0.11) vs (0.98 ± 0.04); bladder: (0.97 ± 0.06) vs (0.98 ± 0.04)). The ADD metric showed a difference of (0.9 ± 0.8)Gy for the anorectum D 1 cm 3 . The bladder D 5cm 3 reported a difference of (0.7 ± 1.5)Gy. Mandatory clinical goals were fulfilled in 90 % of the DLP plans.

Conclusions

Combining dose-volume and geometrical metrics allowed detection of clinically relevant changes, applied to both auto-segmentation and auto-planning output and the Local Surface Dice was more sensitive to local changes compared to the Standard Surface Dice. This monitoring is able to evaluate AI behavior in clinical practice and allows candidate selection for active learning.

Improving knowledge-based treatment planning for lung cancer radiotherapy with automatic multi-criteria optimized training plans

BACKGROUND

Knowledge-based planning (KBP) is a method for automated radiotherapy treatment planning where appropriate optimization objectives for new patients are predicted based on a library of training plans. KBP can save time and improve organ at-risk sparing and inter-patient consistency compared to manual planning, but its performance depends on the quality of the training plans. We used another system for automated planning, which generates multi-criteria optimized (MCO) plans based on a wish list, to create training plans for the KBP model, to allow seamless integration of knowledge from a new system into clinical routine. Model performance was compared for KBP models trained with manually created and automatic MCO treatment plans.

MATERIAL AND METHODS

Two RapidPlan models with the same 30 locally advanced non-small cell lung cancer patients included were created, one containing manually created clinical plans (RP_CLIN) and one containing fully automatic multi-criteria optimized plans (RP_MCO). For 15 validation patients, model performance was compared in terms of dose-volume parameters and normal tissue complication probabilities, and an oncologist performed a blind comparison of the clinical (CLIN), RP_CLIN, and RP_MCO plans.

RESULTS

The heart and esophagus doses were lower for RP_MCO compared to RP_CLIN, resulting in an average reduction in the risk of 2-year mortality by 0.9 percentage points and the risk of acute esophageal toxicity by 1.6 percentage points with RP_MCO. The oncologist preferred the RP_MCO plan for 8 patients and the CLIN plan for 7 patients, while the RP_CLIN plan was not preferred for any patients.

CONCLUSION

RP_MCO improved OAR sparing compared to RP_CLIN and was selected for implementation in the clinic. Training a KBP model with clinical plans may lead to suboptimal output plans, and making an extra effort to optimize the library plans in the KBP model creation phase can improve the plan quality for many future patients.

Feasibility and dosimetric evaluation of single- and multi-isocentre stereotactic body radiation therapy for multiple liver metastases

Objectives

Single-isocentre volumetric-modulated arc therapy (VMAT) stereotactic body radiation therapy (SBRT) improves treatment efficiency and patient compliance for patients with multiple liver metastases (MLM). However, the potential increase in dose spillage to normal liver tissue using a single-isocentre technique has not yet been studied. We comprehensively evaluated the quality of single- and multi-isocentre VMAT-SBRT for MLM and propose a RapidPlan-based automatic planning (AP) approach for MLM SBRT.

Methods

A total of 30 patients with MLM (two or three lesions) were selected for this retrospective study. We manually replanned all patients treated with MLM SBRT by using the single-isocentre (MUS) and multi-isocentre (MUM) techniques. Then, we randomly selected 20 MUS and MUM plans for training to generate the single-isocentre RapidPlan model (RPS) and the multi-isocentre RapidPlan model (RPM). Finally, we used data from the remaining 10 patients to validate RPS and RPM.

Results

Compared with MUS, MUM reduced the mean dose delivered to the right kidney by 0.3 Gy. The mean liver dose (MLD) was 2.3 Gy higher for MUS compared with MUM. However, the monitor units, delivery time, and V20Gy of normal liver (liver-gross tumour volume) for MUM were significantly higher than for MUS. Based on validation, RPS and RPM slightly improved the MLD, V20Gy, normal tissue complications, and dose sparing to the right and left kidneys and spinal cord compared with manual plans (MUS vs RPS and MUM vs RPM), but RPS and RPM significantly increased monitor units and delivery time.

Conclusions

The single-isocentre VMAT-SBRT approach could be used for MLM to reduce treatment time and patient comfort at the cost of a small increase in the MLD. Compared with the manual plans, RapidPlan-based plans, especially RPS, have slightly improved quality.

Artificial Intelligence in Radiation Therapy

Artificial intelligence (AI) has great potential to transform the clinical workflow of radiotherapy. Since the introduction of deep neural networks, many AI-based methods have been proposed to address challenges in different aspects of radiotherapy. Commercial vendors have started to release AI-based tools that can be readily integrated to the established clinical workflow. To show the recent progress in AI-aided radiotherapy, we have reviewed AI-based studies in five major aspects of radiotherapy including image reconstruction, image registration, image segmentation, image synthesis, and automatic treatment planning. In each section, we summarized and categorized the recently published methods, followed by a discussion of the challenges, concerns, and future development. Given the rapid development of AI-aided radiotherapy, the efficiency and effectiveness of radiotherapy in the future could be substantially improved through intelligent automation of various aspects of radiotherapy.

Assessment of efficacy in automated plan generation for Varian Ethos intelligent optimization engine

Varian Ethos, a new treatment platform, is capable of automatically generating treatment plans for initial planning and for online, adaptive planning, using an intelligent optimization engine (IOE). The primary purpose of this study is to assess the efficacy of Varian Ethos IOE for auto‐planning and intercompare different treatment modalities within the Ethos treatment planning system (TPS). A total of 36 retrospective prostate and proximal seminal vesicles cases were selected for this study. The prescription dose was 50.4 Gy in 28 fractions to the proximal seminal vesicles, with a simultaneous integrated boost of 70 Gy to the prostate gland. Based on RT intent, three treatment plans were auto‐generated in the Ethos TPS and were exported to the Eclipse TPS for intercomparison with the Eclipse treatment plan. When normalized for the same planning target volume (PTV) coverage, Ethos plans D_max% were 108.1 ± 1.2%, 108.4 ± 1.6%, and 109.6 ± 2.0%, for the 9‐field IMRT, 12‐field IMRT, and 2‐full arc VMAT plans, respectively. This compared well with Eclipse plan D_max% values, which was 108.8 ± 1.4%. OAR indices were also evaluated for Ethos plans using Radiation Therapy Oncology Group report 0415 as a guide and were found to be comparable to each other and the Eclipse plans. While all Ethos plans were comparable, we found that, in general, the Ethos 12‐field IMRT plans met most of the dosimetric goals for treatment. Also, Ethos IOE consistently generated dosimetrically hotter VMAT plans versus IMRT plans. On average, Ethos TPS took 13 min to generate 2‐full arc VMAT plans, compared to 5 min for 12‐field IMRT plans. Varian Ethos TPS can generate multiple treatment plans in an efficient time frame and the quality of the plans could be deemed clinically acceptable when compared to manually generated treatment plans.

Automatic one-click planning for hippocampal-avoidance whole-brain irradiation in RayStation

This study aimed to verify the accuracy of auto-contouring and auto-dose optimization for hippocampal-avoidance whole-brain radiation therapy (HA-WBRT). Head computed tomography (CT) images of 15 patients were selected. The regions of interest, containing the brain, hippocampus, eyes, and lacrimal glands, were contoured manually and automatically on CT images. They were compared and evaluated for concordance rates using the Simpson coefficient. To verify the performance of dose optimization, auto-dose planning was compared with manual planning for 15 cases. All optimization plans were performed using the volumetric modulated arc therapy technique. The automatically contoured brain showed a very high concordance rate with the manually contoured brain; the Simpson coefficient was 0.990 ± 0.01. Contrastingly, the concordance rate of the hippocampal contour was low at 0.642 ± 0.15 (right) and 0.500 ± 0.16 (left); however, the rate improved to 0.871 ± 0.09 (right) and 0.852 ± 0.11 (left) with an additional 3-mm margin. For 2% of each planning target volume with the prescribed dose (D2%) and Dmean, there was no significant difference between the automatic and manual plans (35.50 Gy vs 35.23 Gy; p = 0.233 and 33.09 Gy vs 32.84 Gy; p = 0.073, respectively). The D98% was significantly better for the manual plan than for the automatic plan (25.49 Gy vs 26.11 Gy; p < 0.01). Dmax and D100% for the hippocampus did not show any significant difference between the automatic and manual plans (15.65, 16.09 Gy (right, left) vs 15.51, 15.80 Gy; p = 0.804, 0.233 and 8.08, 8.03 Gy vs 8.13, 8.01 Gy; p = 0.495, 1 respectively). The accuracy of auto-contouring for HA-WBRT can be guaranteed by providing an appropriate margin, and the precision of the auto-dose optimization was comparable to that of the manual plan.

Knowledge-based radiation treatment planning: A data-driven method survey

This paper surveys the data-driven dose prediction methods investigated for knowledge-based planning (KBP) in the last decade. These methods were classified into two major categories-traditional KBP methods and deep-learning (DL) methods-according to their techniques of utilizing previous knowledge. Traditional KBP methods include studies that require geometric or anatomical features to either find the best-matched case(s) from a repository of prior treatment plans or to build dose prediction models. DL methods include studies that train neural networks to make dose predictions. A comprehensive review of each category is presented, highlighting key features, methods, and their advancements over the years. We separated the cited works according to the framework and cancer site in each category. Finally, we briefly discuss the performance of both traditional KBP methods and DL methods, then discuss future trends of both data-driven KBP methods to dose prediction.

Evaluation of a generalized knowledge-based planning performance for VMAT irradiation of breast and locoregional lymph nodes-Internal mammary and/or supraclavicular regions

PURPOSE

To evaluate the performance of eleven Knowledge-Based (KB) models for planning optimization (RapidPlantm (RP), Varian) of Volumetric Modulated Arc Therapy (VMAT) applied to whole breast comprehensive of nodal stations, internal mammary and/or supraclavicular regions.

METHODS AND MATERIALS

Six RP models have been generated and trained based on 120 VMAT plans data set with different criteria. Two extra-structures were delineated: a PTV for the optimization and a ring structure. Five more models, twins of the previous models, have been created without the need of these structures.

RESULTS

All models were successfully validated on an independent cohort of 40 patients, 30 from the same institute that provided the training patients and 10 from an additional institute, with the resulting plans being of equal or better quality compared with the clinical plans. The internal validation shows that the models reduce the heart maximum dose of about 2 Gy, the mean dose of about 1 Gy and the V20Gy of 1.5 Gy on average. Model R and L together with model B without optimization structures ensured the best outcomes in the 20% of the values compared to other models. The external validation observed an average improvement of at least 16% for the V5Gy of lungs in RP plans. The mean heart dose and for the V20Gy for lung IPSI were almost halved. The models reduce the maximum dose for the spinal canal of more than 2 Gy on average.

CONCLUSIONS

All KB models allow a homogeneous plan quality and some dosimetric gains, as we saw in both internal and external validation. Sub-KB models, developed by splitting right and left breast cases or including only whole breast with locoregional lymph nodes, have shown good performances, comparable but slightly worse than the general model. Finally, models generated without the optimization structures, performed better than the original ones.