Influence of multi-leaf collimator leaf transmission on head and neck intensity-modulated radiation therapy and volumetric-modulated arc therapy planning

Revolutionizing radiation therapy: the role of AI in clinical practice

This review provides an overview of the application of artificial intelligence (AI) in radiation therapy (RT) from a radiation oncologist's perspective. Over the years, advances in diagnostic imaging have significantly improved the efficiency and effectiveness of radiotherapy. The introduction of AI has further optimized the segmentation of tumors and organs at risk, thereby saving considerable time for radiation oncologists. AI has also been utilized in treatment planning and optimization, reducing the planning time from several days to minutes or even seconds. Knowledge-based treatment planning and deep learning techniques have been employed to produce treatment plans comparable to those generated by humans. Additionally, AI has potential applications in quality control and assurance of treatment plans, optimization of image-guided RT and monitoring of mobile tumors during treatment. Prognostic evaluation and prediction using AI have been increasingly explored, with radiomics being a prominent area of research. The future of AI in radiation oncology offers the potential to establish treatment standardization by minimizing inter-observer differences in segmentation and improving dose adequacy evaluation. RT standardization through AI may have global implications, providing world-standard treatment even in resource-limited settings. However, there are challenges in accumulating big data, including patient background information and correlating treatment plans with disease outcomes. Although challenges remain, ongoing research and the integration of AI technology hold promise for further advancements in radiation oncology.

Predicting the complexity of head-and-neck volumetric-modulated arc therapy planning using a radiation therapy planning quality assurance software

Background/Aim

The more complex the treatment plan, the higher the possibility of errors in dose verification. Recently, a treatment planning quality assurance (QA) software (PlanIQ) with a function to objectively evaluate the quality of volumetric-modulated arc therapy (VMAT) treatment plans by scoring and calculating the ideal dose-volume histogram has been marketed. This study aimed to assess the association between the scores of ideal treatment plans identified using PlanIQ and the results of dose verification and to investigate whether the results of dose verification can be predicted based on the complexity of treatment plans.

Materials and methods

Dose verification was performed using an ionization chamber dosimeter, a radiochromic film, and a three-dimensional dose verification system, Delta4 PT. Correlations between the ideal treatment plan scores obtained by PlanIQ and the results of the absolute dose verification and dose distribution verification were obtained, and it was examined whether dose verifications could be predicted from the complexity of the treatment plans.

Results

Even when the score from the ideal treatment plan was high, the results of absolute dose verification and dose distribution verification were sometimes poor. However, even when the score from the ideal treatment plan was low, the absolute volume verification and dose distribution verification sometimes yielded good results.

Conclusions

Treatment plan complexity can be determined in advance from the ideal treatment plan score calculated by PlanIQ. However, it is difficult to predict the results of dose verification using an ideal treatment plan.

Evaluation of prediction and classification performances in different machine learning models for patient-specific quality assurance of head-and-neck VMAT plans

PURPOSE

The purpose of this study is to evaluate the prediction and classification performances of the gamma passing rate (GPR) for different machine learning models and to select the best model for achieving machine learning-based patient-specific quality assurance (PSQA).

METHODS

The measurement verification of 356 head-and-neck volumetric modulated arc therapy plans was performed using a diode array phantom (Delta4 Phantom), and GPR values at 2%/2 mm with global normalization and 3%/2 mm with local normalization were calculated. Machine learning models, including ridge regression (RIDGE), random forest (RF), support vector regression (SVR), and stacked generalization (STACKING), were used to predict the GPR. Each machine learning model was trained using 260 plans, and the prediction accuracy was evaluated using the remaining 96 plans. The prediction error between the measured and predicted GPR was evaluated. For the classification evaluation, the lower control limit for the measured GPR and lower control limit for predicted GPR (LCL_p ) was defined to identify whether the GPR values represent a "pass" or a "fail." LCL_p values with 99% and 99.9% confidence levels were calculated as the upper prediction limits for the GPR estimated from the linear regression between the measured and predicted GPR.

RESULTS

There was an overestimation trend of the low measured GPR. The maximum prediction errors for RIDGE, RF, SVR, and STACKING were 3.2%, 2.9%, 2.3%, and 2.2% at the global 2%/2 mm and 6.3%, 6.6%, 6.1%, and 5.5% at the local 3%/2 mm, respectively. In the global 2%/2 mm, the sensitivity was 100% for all the machine learning models except RIDGE when using 99% LCL_p . The specificity was 76.1% for RIDGE, RF, and SVR and 66.3% for STACKING; however, the specificity decreased dramatically when 99.9% LCL_p was used. In the local 3%/2 mm, however, only STACKING showed 100% sensitivity when using 99% LCL_p . The decrease in the specificity using 99.9% LCL_p was smaller than that in the global 2%/2 mm, and the specificity for RIDGE, RF, SVR, and STACKING was 61.3%, 61.3%, 72.0%, and 66.8%, respectively.

CONCLUSIONS

STACKING had better prediction accuracy for low GPR values than other machine learning models. Applying LCL_p to a regression model enabled the consistent evaluation of quantitative and qualitative GPR predictions. Adjusting the confidence level of the LCL_p helped improve the balance between the sensitivity and specificity. We suggest that STACKING can assist the safe and efficient operation of PSQA.

Impact of treatment planning quality assurance software on volumetric-modulated arc therapy plans for prostate cancer patients

INTRODUCTION

Software that evaluates the quality of treatment plans (PlanIQ^TM) has become commercially available in recent years. It includes a feasibility assessment tool that provides the ideal dose volume histogram (DVH) for each organ at risk, based on the ideal dose falloff from the prescribed dose at the target boundary. It is important to investigate whether the PlanIQ^TM assessment tool (Feasibility DVH^TM) can assist treatment planners who have limited to no experience in treatment planning. Therefore, the present study aimed to evaluate this tool's usefulness for improving the quality of treatment plans.

MATERIALS & METHODS

This study included 5 patients with prostate cancer. The treatment planners were 2 graduate students, 2 undergraduate students, and one clinical planner. All students were radiological technology and medical physics students with no clinical experience. Two different volumetric-modulated arc therapy (VMAT) plans were developed before and after Feasibility DVH^TM. The quality of each treatment plan was evaluated based on a scoring system implemented in PlanIQ^TM.

RESULTS

Of 5 patients included, 4 received improved treatment plans when Feasibility DVH^TM was used. Moreover, 4 of 5 treatment planners showed improvement in treatment planning using Feasibility DVH^TM.

CONCLUSIONS

The findings suggest that using the Feasibility DVH^TM tool may improve treatment plans for different planners and patients. However, planners at any level of experience should be trained to check the dose distribution in addition to checking the DVH, which depends on the adequacy of the contours.

Analysis of prostate intensity- and volumetric-modulated arc radiation therapy planning quality with PlanIQTM

Purpose

The purpose of this study was to assess the quality of treatment planning using the PlanIQ^TM software and to investigate whether it is possible to improve the quality of treatment planning using the “Feasibility dose‐volume histogram (DVH)^TM” implemented in the PlanIQ^TM software.

Methods

Using the PlanIQ^TM software, we retrospectively analyzed the learning curve regarding the quality of the treatment plans for 148 patients of prostate intensity‐modulated radiation therapy and volumetric‐modulated radiation therapy performed at our institution over the past eight years. We also sought to examine the possibility of improving treatment planning quality by re‐planning in 47 patients where the quality of the target dose and the dose limits for organs at risk (OARs) were inadequate. The re‐planning treatment plans referred to the Feasibility DVH^TM implemented in the PlanIQ^TM software and modified the treatment planning system based on the target dose and OAR constraints.

Results

Analysis of the learning curve of the treatment plans quality using PlanIQ^TM software retrospectively showed a trend of improvement in the treatment plan quality from year to year. The improvement in the treatment plans quality was more influenced by dose reduction in the OARs than by target coverage.

In all cases where re‐planning was performed, the improvement in the treatment plan's quality resulted in a better treatment plan than the one adopted for delivery to patients in the clinical plan.

Conclusions

The PlanIQ^TM provided insights into the quality of the treatment plans at our institution and identified problems and areas for improvement in the treatment plans, allowing for the development of appropriate treatment plans for specific patients.