Knowledge-based Planning Model Research Articles

A Knowledge-Based Planning model for IMRT in breast and lung cancer

The advent of Knowledge-Based Planning (KBP) models has introduced a transformative approach to Intensity-Modulated Radiation Therapy (IMRT) treatment planning in breast cancer and lung cancer cases. This paper explores the application of KBP models to these specific cancer types, highlighting their potential to enhance treatment accuracy, efficiency, and patient outcomes. By leveraging historical treatment data and machine learning techniques, KBP-IMRT offers a data-driven framework for optimizing dose distributions, minimizing radiation exposure to healthy tissues, and improving overall treatment plan quality. Through a comprehensive review of the literature and clinical case studies, this paper underscores the advantages of KBP-IMRT, such as streamlined planning processes and improved plan consistency, while acknowledging the challenges associated with model development and implementation. As the field of radiotherapy continues to evolve, KBP models hold the promise of shaping the future of personalized and precise cancer treatment strategies.

HN-SIB-BPI: A single click, sub-site specific, dosimetric scorecard tuned RapidPlan model created from a foundation model for treating head and neck with bilateral neck

To develop a Knowledge Based Planning (KBP) model for creating quantifiably high quality VMAT treatment plans in a single click for head and neck cases treated Simultaneous Integrated Boost (SIB) with bilateral parotid involvement (BPI) where both parotids are near, abutting or partially overlapping target volume. Eclipse RapidPlan and the publicly available PlanScorecard tool were used to assess existing Head and Neck RapidPlan models on two representative cases. The best performer was used as a foundation model to assist in creating new initial training set doses from previously treated cases. Those initial 27 cases were first replanned using only the selected foundation model, then further improved based on manual replanning, informed by dosimetric scorecard assessment. A new, initial model was trained from those 27 foundation model created cases that had been manually improved. Then, that initial model was used to replan those cases again, resulting in higher scores. Additional cases were also replanned using the initial model along with some manual changes to the optimization objectives to increase the score. This resulted in a total of 66 cases from which the final, released, HN-SIB-BPI was trained. A 27 case subset of the full training set was replanned and rescored at each phase of the process with a 260 total point 3-target scorecard. The average score increased: 210.5 foundation model; 226.96 manually improved plans; 230.1 initial model; 231.7 HN-SIB-BPI. On the same 27 case subset, mean ipsilateral and contralateral parotid dose decreased by 1.05Gy and 1.58Gy respectively from the foundation model to HN-SIB-BP. Eight external cases were created from HN-SIB-BPI with dosimetric scorecard validation on Halcyon (3-PTV:221.38/260; 2-PTV:196.1/228.5) and TrueBeam (3-PTV:222.01/260; 2-PTV:202.24/228.5). A specific clinical intent (ie: max parotid sparing) can be articulated in a comprehensive and precise manner by creating a dosimetric scorecard with individual metrics points assigned to each OAR and target metric reflecting their relative importance. This process improved the KBP model (HN-SIB-BPI) in several quantifiable ways including further sparing of parotid dose. All results and tools in this work are shared publicly.

Evaluation of knowledge-based planning models for male pelvic CBCT-based online adaptive radiotherapy on conventional linear accelerators.

To assess the practicality of employing a commercial knowledge-based planning tool (RapidPlan) to generate adapted intact prostate and prostate bed volumetric modulated arc therapy (VMAT) plans on iterative cone-beam computed tomography (iCBCT) datasets. Intact prostate and prostate bed RapidPlan models were trained utilizing planning data from 50 and 44 clinical cases, respectively. To ensure that refined models were capable of producing adequate clinical plans with a single optimization, models were tested with 50 clinical planning CT datasets by comparing dose-volume histogram (DVH) and plan quality metric (PQM) values between clinical and RapidPlan-generated plans. The RapidPlan tool was then used to retrospectively generate adapted VMAT plans on daily iCBCT images for 20 intact prostate and 15 prostate bed cases. As before, DVH and PQM metrics were utilized to dosimetrically compare scheduled (iCBCT Verify) and adapted (iCBCT RapidPlan) plans. Timing data was collected to further evaluate the feasibility of integrating this approach within an online adaptive radiotherapy workflow. Model testing results confirmed the models were capable of producing VMAT plans within a single optimization that were overall improved upon or dosimetrically comparable to original clinical plans. Direct application of RapidPlan on iCBCT datasets produced satisfactory intact prostate and prostate bed plans with generally improved target volume coverage/conformality and rectal sparing relative to iCBCT Verify plans as indicated by DVH values, though bladder metrics were marginally increased on average. Average PQM values for iCBCT RapidPlans were significantly improved compared to iCBCT Verify plans. The average time required [in mm:ss] to generate adapted plans was 06:09±02:06 (intact) and 07:12±01:04 (bed). This study demonstrated the feasibility of leveraging RapidPlan to expeditiously generate adapted VMAT intact prostate and prostate bed plans on iCBCT datasets. In general, adapted plans were dosimetrically improved relative to scheduled plans, emphasizing the practicality of the proposed approach.

Optimizing volumetric modulated arc therapy prostate planning using an automated Fine-Tuning process through dynamic adjustment of optimization parameters

In radiotherapy treatment planning, optimization is essential for achieving the most favorable plan by adjusting optimization criteria. This study introduced an innovative approach to automatically fine-tune optimization parameters for volumetric modulated arc therapy prostate planning, ensuring all constraints were met. A knowledge-based planning model was invoked, and the fine-tuning process was applied through an in-house developed script. Among 25 prostate plans, this fine-tuning increased the number of plans meeting all constraints from 10/25 to 22/25, with a reduction in mean monitor units per gray without increasing plan’s complexity. This automation improved efficiency by saving time and resources in treatment planning.

Clinical Impact of Contouring Variability for Prostate Cancer Tumor Boost

PurposeThe focal radiotherapy (RT) boost technique was shown in a phase III randomized controlled trial (RCT) to improve prostate cancer outcomes without increasing toxicity. This technique relies on the accurate delineation of prostate tumors on MRI. A recent prospective study evaluated radiation oncologists’ accuracy when asked to delineate prostate tumors on MRI and demonstrated high variability in tumor contours. We sought to evaluate the impact of contour variability and inaccuracy on predicted clinical outcomes. We hypothesized that radiation oncologists' contour inaccuracies would yield meaningfully worse clinical outcomes. Materials & Methods45 radiation oncologists and 2 expert radiologists contoured prostate tumors on 30 patient cases. Of these cases, those with CT simulation or diagnostic CT available were selected for analysis. A knowledge-based planning model was developed to generate focal RT boost plans for each contour per the RCT protocol. Probability of biochemical failure (BF) was determined using a model from the RCT. The primary metric evaluated was delta BF (ΔBF = Participant BF – Expert BF). An absolute increase in BF ≥5% was considered clinically meaningful. Results8 patient cases and 394 target volumes for focal RT boost planning were included in this analysis. In general, participant plans were associated with worse predicted clinical outcomes compared to the expert plan, with an average absolute increase in BF of 4.3%. 37% of participant plans were noted to have an absolute increase in BF of 5% or more. ConclusionRadiation oncologists’ attempts to contour tumor targets for focal RT boost are frequently inaccurate enough to yield meaningfully inferior clinical outcomes for patients.

Knowledge-based versus deep learning based treatment planning for breast radiotherapy

Background and PurposeTo improve radiotherapy (RT) planning efficiency and plan quality, knowledge-based planning (KBP) and deep learning (DL) solutions have been developed. We aimed to make a direct comparison of these models for breast cancer planning using the same training, validation, and testing sets. Materials and MethodsTwo KBP models were trained and validated with 90 RT plans for left-sided breast cancer with 15 fractions of 2.6 Gy. The versions either used the full dataset (non-clean model) or a cleaned dataset (clean model), thus eliminating geometric and dosimetric outliers. Results were compared with a DL U-net model (previously trained and validated with the same 90 RT plans) and manually produced RT plans, for the same independent dataset of 15 patients. Clinically relevant dose volume histogram parameters were evaluated according to established consensus criteria. ResultsBoth KBP models underestimated the mean heart and lung dose equally 0.4 Gy (0.3–1.1 Gy) and 1.4 Gy (1.1–2.8 Gy) compared to the clinical plans 0.8 Gy (0.5–1.8 Gy) and 1.7 Gy (1.3–3.2 Gy) while in the final calculations the mean lung dose was higher 1.9–2.0 Gy (1.5–3.5 Gy) for both KPB models. The U-Net model resulted in a mean planning target volume dose of 40.7 Gy (40.4–41.3 Gy), slightly higher than the clinical plans 40.5 Gy (40.1–41.0 Gy). ConclusionsOnly small differences were observed between the estimated and final dose calculation and the clinical results for both KPB models and the DL model. With a good set of breast plans, the data cleaning module is not needed and both KPB and DL models lead to clinically acceptable results.

Effects of model size and composition on quality of head-and-neck knowledge-based plans.

Knowledge-based planning (KBP) aims to automate and standardize treatment planning. New KBP users are faced with many questions: How much does model size matter, and are multiple models needed to accommodate specific physician preferences? In this study, six head-and-neck KBP models were trained to address these questions. The six models differed in training size and plan composition: The KBPFull (n=203 plans), KBP101 (n=101), KBP50 (n=50), and KBP25 (n=25) were trained with plans from two head-and-neck physicians. KBPA and KBPB each contained n=101 plans from only one physician, respectively. An independent set of 39 patients treated to 6000-7000cGy by a third physician was re-planned with all KBP models for validation. Standard head-and-neck dosimetric parameters were used to compare resulting plans. KBPFull plans were compared to the clinical plans to evaluate overall model quality. Additionally, clinical and KBPFull plans were presented to another physician for blind review. Dosimetric comparison of KBPFull against KBP101 , KBP50 , and KBP25 investigated the effect of model size. Finally, KBPA versus KBPB tested whether training KBP models on plans from one physician only influences the resulting output. Dosimetric differences were tested for significance using a paired t-test (p<0.05). Compared to manual plans, KBPFull significantly increased PTV Low D95% and left parotid mean dose but decreased dose cochlea, constrictors, and larynx. The physician preferred the KBPFull plan over the manual plan in 20/39 cases. Dosimetric differences between KBPFull , KBP101 , KBP50 , and KBP25 plans did not exceed 187cGy on aggregate, except for the cochlea. Further, average differences between KBPA and KBPB were below 110cGy. Overall, all models were shown to produce high-quality plans. Differences between model outputs were small compared to the prescription. This indicates only small improvements when increasing model size and minimal influence of the physician when choosing treatment plans for training head-and-neck KBP models.

Knowledge Based Planning for Lung SBRT: Model Transferability between Treatment Systems and Calculation Algorithms

Knowledge based planning (KBP) is known to be effective and clinically feasible in achieving high quality treatment plans with less variability between planners and reduced planning time. A KBP model is highly dependent on characteristics of the training data. Our institution developed and validated a lung SBRT KBP model from plans treated with a C-arm linear accelerator (Linac). This study aims to evaluate the quality of lung SBRT plans with a KBP model when this model is used for an O-ring Linac with different MLC design, beam quality, and calculation algorithm. Two academic hospitals in the same healthcare network were included in this study. One hospital trained a KBP model for five-fraction lung SBRT using 43 plans calculated using the Eclipse AcurosXB v15.6 algorithm for a C-arm Linac. The model was developed to prioritize dose gradient outside the PTV while satisfying institutional critical organ limits and validated using an additional set of 12 plans. A cohort of 10 patients (11 plans) previously treated with lung SBRT using the O-ring Linac at the second hospital was selected. These plans were calculated using the Eclipse AAA v15.6 algorithm with prescription doses of 30-50 Gy in five fractions. Each plan was re-optimized using the KBP model, without manual adjustment of the optimization objectives, using the same beam geometry as in the clinical plans. KBP plans were normalized to achieve the same PTV coverage as the clinical plans. Clinical and KBP plans were evaluated for ITV D98%, PTV conformity index (CI), R50 and maximum dose at 2 cm from the PTV (D2cm). All clinical and KBP plans met critical organ dose constraints. The range of CI was 1-1.13 vs. 0.96-1.05 for clinical plans and KBP plans, respectively. R50 and D2cm were lower with KBP plans (p = 0.007 and p = 0.05, respectively). ITV D98% range was 105-120% vs. 117-122% of the PTV prescription dose for clinical plans and KBP plans, respectively, and higher with KBP plans (p<0.001). For plans where the PTV overlaps the chest wall (n = 7), the maximum chest wall doses (D0.03cc) from KBP plans were higher than the clinical plans (p = 0.16), with an absolute chest wall D0.03cc difference of 9 Gy (54 Gy vs 63 Gy). The KBP model includes an ITV SIB of 120% of the PTV prescription, whereas the clinical plans reduced the ITV SIB when the PTV overlapped the chest wall, highlighting a difference in clinical practice between the two hospitals in this study. This study supports that a lung SBRT KBP model trained using plans for one Linac system and the AcurosXB algorithm can be used to generate clinically acceptable plans without manual planner intervention for a different Linac system and the AAA calculation algorithm, indicating the potential of a KBP model use for centralized planning process across different Linac systems and calculation algorithms.

Knowledge Based Planning Evolution: Multi-Criteria Optimization of Prostates. Is There Always a Better Model?

The current paradigm of radiotherapy Knowledge Based Planning (KBP) utilizes dose volume relationships on previously treated plans and may be insensitive to clinician plan preferences. To improve KBP models, incorporation of plan quality metrics and optimization weightings can be achieved using Clinical Scoring Metrics (CSM) and Multi-Criteria Optimization (MCO) respectively. A combination of KBP, MCO and CSM may result in higher quality plan output with potential to improve all prospective clinical plans. This study assesses the benefit of MCO, using CSM, and its application in the evolution of the KBP model. A retrospective Dose Volume Histogram (DVH) parameter review was undertaken for 40 intact prostate radiotherapy cases. Following analysis, these cases were used to develop the CSM system to reflect clinician preferences and departmental protocol. Hypofractionated plans containing simultaneous integrated boost (SIB) and urethral sparing were originally planned using a clinically approved KBP model (Model A). Model A was then refined using MCO, improving the 40 individual model plans to form Model B. A further 20 prostate patients, who were excluded from KBP development, were planned using both models to compare plan quality using the CSM system. Using a single optimization, Model A resulted in a median plan score of 134.1 (92.4-151.5), compared to the median Model B score of 145.1 (114.5-175.1) out of a possible score of 200 points. A Wilcoxon rank sum test was conducted and found the model score difference was significant (p = 0.014). Organ At Risk (OAR) doses, including Rectum V30 Gy, were found to significantly (p = 0.019) decrease, with Model A resulting in a median of 22.1 Gy (9.3-39.0), compared to the Model B median of 16.3 Gy (6.8-35.7). PTV coverage metric V57 Gy, resulted in no significant differences (p = 0.47) with a median score of 98.4% (97.2% - 99.9%) and 98.4% (96.9% - 99.2%) for model A and B, respectively. The application of MCO was used to influence and produce higher quality KBP models, as supported by a CSM system. OAR doses were found to significantly decrease, with no effect on target coverage. As a result, existing plan objectives were tightened through the utilization of KBP models, even with a single optimization resulting in higher scoring plans. The future development of this work may increase the efficiency of planning operations and allow the clinician to accurately anticipate planning goals for not only prostate, but all disease sites.

Standardizing LT Chest Wall Radiotherapy Treatment Planning in a Low- or Middle- Income Country Radiotherapy Clinic Using Knowledge Based Planning

Radiotherapy departments in low- or middle-income countries (LMICs) tend to lag behind introducing emerging technologies like intensity-modulated radiotherapy (IMRT). IMRT has become the standard of care in high-income countries (HIC) due to reduced toxicity and improved outcomes in a wide variety of cancers. The purpose of this work is showing the results of left Chest-Wall knowledge-based planning (KBP) standardization and implementation in a LMIC setting. A Halcyon Linac was installed at our clinic in Guatemala in 2019 and currently used to treat ∼90 IMRT patients daily. The standardization of IMRT procedures has been difficult for complex sites like chest-wall. The steps for standardization included: AAPM TG-263 nomenclature implementation, and planning workflows within the TPS, creation of optimization structures, and plan quality evaluation following RTOG1005 protocol hypofractionation arm. 25 plans were created manually achieving all RTOG1005 protocol constraints. The statistics were analyzed trough the model analytics tool provided by KPB manufacturer. The results show that more plans are needed to improve the KBP model. This initial model was used to create a standardized clinical protocol in the TPS in order to continue adding plans to the KBP model database. This approach ensures that we obtain consistent plan quality and standardize our planning. The manual planning objectives achieved: CONCLUSION: The experience using the TPS to standardize our treatment planning process achieved good consistency in our planning objectives. This approach will help create KBP models according to our own clinic-specific requirements. Future work will be made to compare our LMIC KBP models with those made at a HIC academic radiotherapy center.

Identifying and Re-Planning Challenging Retrospective SRS Patients to Evaluate the Potential for a White Matter Tract Sparing Knowledge-Based Planning Model

Diffusion tensor imaging (DTI) with automated tractography, along with auto-parcellation of volumetric MRI (vMRI), allows for reliable parcellation of critical white matter (WM) and hippocampal structures in the brain. Knowledge-based planning (KBP) may facilitate the optimization of new brain SRS plans that highly prioritize WM tracts and hippocampal sparing, allowing for cognitive preservation. However, building a high-performing model is challenging as the geometry of WM tracts and their proximity to planning target volumes (PTVs) can vary substantially. In this work we propose an iterative step for creating a KBP model where we used a preliminary model to identify challenging plans and worked on improving their dosimetry to refine the model predictions. Tractography of eloquent deep WM was obtained from high resolution DTI, along with auto-segmentation of bilateral hippocampi from vMRI and an open-source software for processing and analyzing brain MRI images. These ROIs were imported to the treatment planning system as organs-at-risk (OARsDTI). A preliminary KBP model was built from an IRB-approved cohort of 30 consecutive patients (75 plans total) treated for brain metastasis with non-coplanar SRS with total dose from 20 to 30 Gy in one to five fractions using a LINAC. This preliminary model was used to identify the 7 plans with the highest doses to the OARsDTI. These plans were re-optimized with new objectives on the OARsDTI that were partially or totally overlapped with the isodose line representing 30% of the prescription dose (as a surrogate for proximity to the PTV). The OARsDTI that were overlapping with the PTV were excluded from this analysis. We evaluated the re-optimized plans (PlanNew) by comparing maximum and mean dose (Dmax and Dmean, respectively) of OARsDTI to the initial values from the preliminary plans (PlanPre). In order to verify that the PlanNew were not compromising other structures or treatment goals, we also evaluated conformity index (CI), PTV coverage and the dosimetry of OARs included in the PlanPre. Additionally, the PlanNew were normalized to our clinical standard used for the PlanPre (100% of prescription dose to 98% of the PTV). In the 7 plans, there were 30 OARsDTI overlapping with the 30% isodose line. Refining the dosimetric objectives for the OARsDTI decreased the Dmean on average 31% when comparing PlanNew to Planpre (p<0.05, Wilcoxon rank-sum test). The CI of the PlanNew were not statistically different from the PlanPre. Although Dmax of PlanNew was 12% lower than Dmax of Planpre, the difference was not statistically different. The addition of a plan refinement step was shown to improve dose to the OARsDTI in 7 challenging SRS plans. The generation of these plans are a necessary step to creating a refined KBP model capable of accounting for the high variability in WM tracts geometry and their proximity with PTV.

Left Anterior Descending Coronary Artery-Optimized Knowledge-Based Radiotherapy Planning in Patients with Locally Advanced Non-Small Cell Lung Cancer

The left anterior descending (LAD) coronary artery volume receiving 15 Gy (V15 Gy) has been associated with major adverse cardiac events and mortality in patients with locally advanced non-small cell lung cancer (LA-NSCLC) treated with radiotherapy (RT). However, coronary arteries are not routinely contoured as standard-of-care and there is limited data describing coronary-sparing planning techniques. Our objective was to evaluate whether LAD-sparing could be improved compared to real-world plans using a commercial knowledge-based planning (KBP) software without clinically unacceptable tradeoffs to other critical organs at risk (OAR). Retrospective replanning analysis was conducted on 31 RT plans developed for patients with LA-NSCLC. Each case was manually replanned with volumetric modulated arc radiotherapy (VMAT) optimized for LAD V15 Gy sparing with a prescription dose of 60 Gy. The resulting plans were used to create an LAD-optimized KBP model. The model included both KBP-generated objectives and additional point objectives to meet LAD (V15 Gy < 10%), lung (V5 Gy < 60%, V20 < 30%, mean < 17 Gy), heart (mean < 20 Gy), spinal cord (D0.03cc < 45 Gy), esophagus (V60 Gy < 17%, mean < 34 Gy), and planning target volume (PTV; D95% = 100%). The LAD-optimized model was deployed to replan 33 additional VMAT plans for LA-NSCLC with prescription dose 60 Gy, delivered between 2018 and 2022. Dose constraints or objectives were only unmet with KBP when not met in the original plan. Descriptive statistics were used to compare dosimetric parameters between original and LAD-optimized KBP plans. For the 33 plans identified for LAD-optimized KBP replanning, the original plans had a mean LAD V15 Gy of 17.9% (standard error [SE], 3.9%), while the LAD-optimized KBP plans had a significantly lower mean LAD V15 Gy of 5.9% (SE, 1.9%; p = 0.031). There was no significant difference in mean heart dose between the original and LAD-optimized KBP plans (8.4 Gy [SE 6.6 Gy] vs. 7.3 Gy [SE 5.8 Gy]; p = 0.48). Of these 33 KBP plans, 42% (n = 14) had mean reductions in LAD V15 Gy from 40.5% (SE 4.3%) to 10.5% (SE 3.5%; p = 0.0002), with 9/14 now able to meet < 10%, without affecting other critical OAR objectives (p > 0.5 for lungs, esophagus, spinal cord). 45% (n = 15) originally had an LAD V15 Gy < 1%, which were unchanged using KBP (p > 0.3). Four plans (12%) had higher LAD V15 Gy in the KBP vs. original plan (2/4 still meeting < 10%), due to the model prioritizing other OARs to a greater degree. Improvements in LAD-sparing was achievable in more than 40% of plans without apparent clinically significant tradeoffs in other critical OARs and without corresponding change in MHD. This result underscores the importance of standard implementation of LAD contouring and dose constraints, particularly given the discordance between whole heart and LAD dose exposure.

Automated evaluation for rapid implementation of knowledge-based radiotherapy planning models.

Knowledge-based planning (KBP) offers the ability to predict dose-volume metrics based on information extracted from previous plans, reducing plan variability and improving plan quality. As clinical integration of KBP is increasing there is a growing need for quantitative evaluation of KBP models. A .NET-based application, RapidCompare, was created for automated plan creation and analysis of Varian RapidPlan models. RapidCompare was designed to read calculation parameters and a list of reference plans. The tool copies the reference plan field geometry and structure set, applies the RapidPlan model, optimizes the KBP plan, and generates data for quantitative evaluation of dose-volume metrics. A cohort of 85 patients, divided into training (50), testing (10), and validation (25) groups, was used to demonstrate the utility of RapidCompare. After training and tuning, the KBP model was paired with three different optimization templates to compare various planning strategies in the validation cohort. All templates used the same set of constraints for the planning target volume (PTV). For organs-at-risk, the optimization template provided constraints using the whole dose-volume histogram (DVH), fixed-dose/volume points, or generalized equivalent uniform dose (gEUD). The resulting plans from each optimization approach were compared using DVH metrics. RapidCompare allowed for the automated generation of 75 total plans for comparison with limited manual intervention. In comparing optimization techniques, the Dose/Volume and Lines optimization templates generated plans with similar DVH metrics, with a slight preference for the Lines technique with reductions in heart V30Gy and spinal cord max dose. The gEUD model produced high target heterogeneity. Automated evaluation allowed for the exploration of multiple optimization templates in a larger validation cohort than would have been feasible using a manual approach. A final KBP model using line optimization objectives produced the highest quality plans without human intervention.

Study on the transferability of the knowledge-based VMAT model to predict IMRT plans in prostate cancer radiotherapy

ObjectiveThe aim of this study was to investigate the feasibility of VMAT library-derived model transfer in the prediction of IMRT plans by dosimetry comparison among with three groups of IMRT plans: two groups of automatic IMRT plans generated by the knowledge-based the volumetric modulated arc therapy (VMAT) model and intensity-modulated radiation therapy (IMRT) model and one group of manual IMRT plans.Methods52 prostate cancer patients who had completed radiotherapy were selected and randomly divided into 2 groups with 40 and 12 separately. Then both VMAT and IMRT plans were manually designed for all patients. The total plans in the group with 40 cases as training datasets were added to the knowledge-based planning (KBP) models for learning and finally obtained VMAT and IMRT training models. Another 12 cases were selected as the validation group to be used to generated auto IMRT plans by KBP VMAT and IMRT models. At last, the radiotherapy plans from three groups were obtained: the automated IMRT plan (V-IMRT) predicted by the VMAT model, the automated IMRT plan (I-IMRT) predicted by the IMRT model and the manual IMRT plan (M-IMRT) designed before. The dosimetric parameters of planning target volume (PTV) and organ at risks (OARs) as well as the time parameters (monitor unit, MU) were statistically analyzed.ResultsThe dose limit of all plans in the training datasets met the clinical requirements. Compared with the training plans added to VMAT model, the dosimetry parameters have no statistical differences in PTV (P > 0.05); the dose of X% volume (Dx%) with D25% and D35% in rectal and the maximum dose (Dmax) in the right femoral head were lower (P = 0.04, P = 0.01, P = 0.00) while D50% in rectal was higher (< 0.05) in the IMRT model plans. In the 12 validation cases, both automated plans showed better dose distribution compared with the M-IMRT plan: the Dmax of PTV in the I-IMRT plans and the dose in volume of interesting (VOI) of bladder and bilateral femoral heads were lower with a statistically significant difference (P < 0.05). Compared with the I-IMRT plans, dosimetric parameters in PTV and VOI of all OARs had no statistically significant differences (P > 0.05), but the Dmax in left femoral heard and D15% in the right femoral head were lower and have significant differences (P < 0.05). Furthermore, the low-dose regions, which was defined as all volumes outside of the PTV (RV) with the statistical parameters of mean dose (Dmean), the volume of covering more than 5 Gy dose (V5Gy), and also the time parameter (MU) required to perform the plan were considered. The results showed that Dmean in V-IMRT was smaller than that in the I-IMRT plan (P = 0.02) and there was no significant difference in V5Gy and MU (P > 0.05).ConclusionCompared with the manual plan, the IMRT plans generated by the KBP models had a significant advantage in dose control of both OARs and PTV. Compared to the I-IMRT plans, the V-IMRT plans was not only without significant disadvantages, but it also achieved slightly better control of the low-dose region, which meet the clinical requirements and can used in the clinical treatment. This study demonstrates that it is feasible to transfer the KBP VMAT model in the prediction of IMRT plans.

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.

Development and Validation of Single-Optimization Knowledge-Based Volumetric Modulated Arc Therapy Model Plan in Nasopharyngeal Carcinomas

Knowledge-based planning (KBP) has evolved to standardize and expedite the complex process of radiation therapy planning for nasopharyngeal cancer (NPC). Herein, we aim to develop and validate the suitability of a single-optimization KBP for NPC. Volumetric modulated arc therapy plans of 103 patients with NPC treated between 2016 and 2020 were reviewed and used to generate a KBP model. A validation set of 15 patients was employed to compare the quality of single optimization KBP and clinical plans using the paired t test and the Wilcoxon signed rank test. The time required for either planning was also analyzed. Most patients (86.7%) were of locally advanced stage (III/IV). The median dose received by 95% of the high-risk planning target volume was significantly higher for the KBP (97.1% vs 96.4%; P=.017). The median homogeneity (0.09 vs 0.1) and conformity (0.98 vs 0.97) indices for high-risk planning target volume and sparing of the normal tissues like optic structures, spinal cord, and uninvolved dysphagia and aspiration-related structures were better with the KBP (P < .05). In the blinded evaluation, the physician preferred the KBP plan in 13 out of 15 patients. The median time required to generate the KBP and manual plans was 53 and 77 minutes, respectively. KBP with a single optimization is an efficient and time saving alternative for manual planning in NPC.

A comparative study of deep learning-based knowledge-based planning methods for 3D dose distribution prediction of head and neck.

In this paper, we compare four novel knowledge-based planning (KBP) algorithms using deep learning to predict three-dimensional (3D) dose distributions of head and neck plans using the same patients' dataset and quantitative assessment metrics. A dataset of 340 oropharyngeal cancer patients treated with intensity-modulated radiation therapy was used in this study, which represents the AAPM OpenKBP - 2020 Grand Challenge dataset. Four 3D convolutional neural network architectures were built. The models were trained on 64% of the data set and validated on 16% for voxel-wise dose predictions: U-Net, attention U-Net, residual U-Net (Res U-Net), and attention Res U-Net. The trained models were then evaluated for their performance on a test data set (20% of the data) by comparing the predicted dose distributions against the ground-truth using dose statistics and dose-volume indices. The four KBP dose prediction models exhibited promising performance with an averaged mean absolute dose error within the body contour <3Gy on 68 plans in the test set. The average difference in predicting the D99 index for all targets was 0.92Gy (p=0.51) for attention Res U-Net, 0.94Gy (p=0.40) for Res U-Net, 2.94Gy (p=0.09) for attention U-Net, and 3.51Gy (p=0.08) for U-Net. For the OARs, the values for the and indices were 2.72Gy (p<0.01) for attention Res U-Net, 2.94Gy (p<0.01) for Res U-Net, 1.10Gy (p<0.01) for attention U-Net, 0.84Gy (p<0.29) for U-Net. All models demonstrated almost comparable performance for voxel-wise dose prediction. KBP models that employ 3D U-Net architecture as a base could be deployed for clinical use to improve cancer patient treatment by creating plans with consistent quality and making the radiotherapy workflow more efficient.

PD-0981 Use of RapidPlan knowledge-based planning models for planning of rectum cases with Ethos TPS

PD-0981 Use of RapidPlan knowledge-based planning models for planning of rectum cases with Ethos TPS

Can knowledge based treatment planning of VMAT for post-mastectomy locoregional radiotherapy involving internal mammary chain and supraclavicular fossa improve performance efficiency?

To validate and evaluate the performance of knowledge-based treatment planning for Volumetric Modulated Arc Radiotherapy for post-mastectomy loco-regional radiotherapy. Two knowledge-based planning (KBP) models for different dose prescriptions were built using the Eclipse RapidPlanTM v 16.1 (Varian Medical Systems, Palo Alto, USA) utilising the plans of previously treated patients with left-sided breast cancer who had undergone irradiation of the left chest wall, internal mammary nodal (IMN) region and supra-clavicular fossa (SCF). Plans of 60 and 73 patients were used to generate the KBP models for the prescriptions of 40 Gy in 15 fractions and 26 Gy in 5 fractions, respectively. A blinded review of all the clinical plans (CLI) and KBPs was done by two experienced radiation oncology consultants. Statistical analysis of the two groups was also done using the standard two-tailed paired t-test or Wilcoxon signed rank test, and p<0.05 was considered significant. A total of 20 metrics were compared. The KBPs were found to be either better (6/20) or comparable (10/20) to the CLIs for both the regimens. Dose to heart, contralateral breast,contralateral lung were either better or comparable in the KBP plans except of ipsilateral lung. Mean dose (Gy) for the ipsilateral lung are significantly (p˂0.001) higher in KBP though the values were acceptable clinically. Plans were of similar quality as per the result of the blinded review which was conducted by slice-by-slice evaluation of dose distribution for target coverage, overdose volume and dose to the OARs. However, it was also observed that treatment times in terms of monitoring units (MUs) and complexity indices are more in CLIs as compared with KBPs (p<0.001). KBP models for left-sided post-mastectomy loco-regional radiotherapy were developed and validated for clinical use. These models improved the efficiency of treatment delivery as well as work flow for VMAT planning involving both moderately hypo fractionated and ultra-hypo fractionated radiotherapy regimens.

Implementation, Dosimetric Assessment, and Treatment Validation of Knowledge-Based Planning (KBP) Models in VMAT Head and Neck Radiation Oncology

The aim of this study was to evaluate knowledge-based treatment planning (KBP) models in terms of their dosimetry and deliverability and to investigate their clinical benefits. Three H&N KBP models were built utilizing RapidPlan™, based on the dose prescription, which is given according to the planning target volume (PTV). The training set for each model consisted of 43 clinically acceptable volumetric modulated arc therapy (VMAT) plans. Model quality was assessed and compared to the delivered treatment plans using the homogeneity index (HI), conformity index (CI), structure dose difference (PTV, organ at risk-OAR), monitor units, MU factor, and complexity index. Model deliverability was assessed through a patient-specific quality assurance (PSQA) gamma index-based analysis. The dosimetric assessment showed better OAR sparing for the RapidPlan™ plans and for the low- and high-risk PTV, and the HI, and CI were comparable between the clinical and RapidPlan™ plans, while for the intermediate-risk PTV, CI was better for clinical plans. The 2D gamma passing rates for RapidPlan™ plans were similar or better than the clinical ones using the 3%/3 mm gamma-index criterion. Monitor units, the MU factors, and complexity indices were found to be comparable between RapidPlan™ and the clinical plans. Knowledge-based treatment plans can be safely adapted into clinical routines, providing improved plan quality in a time efficient way while minimizing user variability.