Assessment of the deep learning-based gamma passing rate prediction system for 1.5 T magnetic resonance-guided linear accelerator.

Measurement-based verification is impossible for the patient-specific quality assurance (QA) of online adaptive magnetic resonance imaging-guided radiotherapy (oMRgRT) because the patient remains on the couch throughout the session. We assessed a deep learning (DL) system for oMRgRT to predict the gamma passing rate (GPR). This study collected 125 verification plans [reference plan (RP), 100; adapted plan (AP), 25] from patients with prostate cancer treated using Elekta Unity. Based on our previous study, we employed a convolutional neural network that predicted the GPRs of nine pairs of gamma criteria from 1%/1 mm to 3%/3 mm. First, we trained and tested the DL model using RPs (n = 75 and n = 25 for training and testing, respectively) for its optimization. Second, we tested the GPR prediction accuracy using APs to determine whether the DL model could be applied to APs. The mean absolute error (MAE) and correlation coefficient (r) of the RPs were 1.22 ± 0.27% and 0.29 ± 0.10 in 3%/2 mm, 1.35 ± 0.16% and 0.37 ± 0.15 in 2%/2 mm, and 3.62 ± 0.55% and 0.32 ± 0.14 in 1%/1 mm, respectively. The MAE and r of the APs were 1.13 ± 0.33% and 0.35 ± 0.22 in 3%/2 mm, 1.68 ± 0.47% and 0.30 ± 0.11 in 2%/2 mm, and 5.08 ± 0.29% and 0.15 ± 0.10 in 1%/1 mm, respectively. The time cost was within 3 s for the prediction. The results suggest the DL-based model has the potential for rapid GPR prediction in Elekta Unity.

Beginning of clinical treatment using the 1.5 Tesla MR-Linac system in Japan: a narrative review.

Background and Objective: In the field of radiation therapy, image-guided radiotherapy (IGRT) technology has been gradually improving and highly accurate radiation treatment has been possible. Research on IGRT using 1.5 Tesla magnetic resonance imaging (MRI) began in 1999, and a radiation therapy device called 1.5 Tesla magnetic resonance linear accelerator (MR-Linac), which combines a linear accelerator with 1.5 Tesla MRI, was developed in Europe. The aim of this review is to present an overview of 1.5 Tesla MR-Linac with a review of the literature and our experience. Methods: Reports related to 1.5 Tesla MR-Linac were searched for in PubMed and are discussed in relation to our experience. Key Content and Findings: The 1.5 Tesla MR-Linac enables IGRT using 1.5 Tesla MRI, further enhancing the precision of radiation therapy. Position verification by cone-beam computed tomography (CBCT) is performed in many institutions, but soft tissue contrast is often unclear in CBCT images of the abdomen and mediastinal organs. Since the 1.5 Tesla MR-Linac allows position verification using MRI, position verification can be performed using clear MRI even in regions where CBCT is unclear. With the 1.5 Tesla MR-Linac, it is possible to perform online adaptive radiotherapy (ART) using 1.5 Tesla MRI. Online ART is a method in which images are acquired while the patient is on the treatment table. The method is based on the current condition of the organs in the body on that day and an optimal treatment field is recreated. Additionally, it allows monitoring of tumor motion using cine images obtained by 1.5 Tesla MRI during the delivery of X-ray radiation. A previous report showed that patients with prostate cancer who received radiotherapy by MR-Linac had fewer side effects than those in patients who received conventional CBCT radiation therapy. Conclusions: The 1.5 Tesla MR-Linac obtained CE-mark certification in Europe in August 2018 and it has been used for clinical treatment. In Japan, clinical treatment using this device started in 2021. By using 1.5 Tesla MR-Linac, patients can be provided with higher precision radiotherapy. In this review, we provide an overview of 1.5 Tesla MR-Linac.

VMAT with DIBH in hypofractionated radiotherapy for left-sided breast cancer after breast-conserving surgery: results of a non-inferiority clinical study.

The purpose of this study was to show the safety of volumetric modulated arc therapy (VMAT) with deep inspiration breath-hold (DIBH) in hypofractionated radiotherapy for left-sided breast cancer after breast-conserving surgery in a clinical setting. Twenty-five Japanese women, aged 20-59 years, who were enrolled in this prospective non-inferiority study received VMAT under the condition of DIBH with 42.4 Gy/16 fractions for whole-breast irradiation (WBI) ± boost irradiation for the tumor bed to show the non-inferiority of VMAT with DIBH to conventional fractionated WBI with free breathing. The primary endpoint was the rate of occurrence of radiation dermatitis of Grade 3 or higher or pneumonitis of Grade 2 or higher within 6 months after the start of radiotherapy. This study was registered with UMIN00004321. All of the enrolled patients completed the planned radiotherapy without interruption. The evaluation of adverse events showed that three patients (12.0%) had Grade 2 radiation dermatitis. There was no other Grade 2 adverse event and there was no patient with an adverse event of Grade 3 or higher. Those results confirmed our hypothesis that the experimental treatment method is non-inferior compared with our historical results. There was no patient with locoregional recurrence or metastases. In conclusion, VMAT under the condition of DIBH in hypofractionated radiotherapy for left-sided breast cancer after breast-conserving surgery can be performed safely in a clinical setting.

Evaluation of the MVCT-based radiomic features as prognostic factor in patients with head and neck squamous cell carcinoma.

BACKGROUND: Megavoltage computed tomography (MVCT) images acquired during each radiotherapy session may be useful for delta radiomics. However, no studies have examined whether the MVCT-based radiomics has prognostic power. Therefore, the purpose of this study was to examine the prognostic power of the MVCT-based radiomics for head and neck squamous cell carcinoma (HNSCC) patients. METHODS: 100 HNSCC patients who received definitive radiotherapy were analyzed and divided into two groups: training (n = 70) and test (n = 30) sets. MVCT images obtained using TomoTherapy for the first fraction of radiotherapy and planning kilovoltage CT (kVCT) images obtained using Aquilion LB CT scanner were analyzed. Primary gross tumor volume (GTV) was propagated from kVCT to MVCT images using rigid registration, and 107 radiomic features were extracted from the GTV in MVCT and kVCT images. Least absolute shrinkage and selection operator (LASSO) Cox regression model was used to examine the association between overall survival (OS) and rad score calculated for each patient by weighting the feature value through the coefficient when features were selected. Then, the predictive values of MVCT-based and kVCT-based rad score and patient-, treatment-, and tumor-specific factors were evaluated. RESULTS: C-indices of the rad score for MVCT- and kVCT-based radiomics were 0.667 and 0.685, respectively. The C-indices of 6 clinical factors were 0.538-0.622. The 3-year OS was significantly different between high- and low-risk groups according to the MVCT-based rad score (50% vs. 83%; p < 0.01). CONCLUSIONS: Our results suggested that MVCT-based radiomics had stronger prognostic power than any single clinical factor and was a useful prognostic factor when predicting OS in HNSCC patients.

Evaluation of deep learning-based deliverable VMAT plan generated by prototype software for automated planning for prostate cancer patients.

This study aims to evaluate the dosimetric accuracy of a deep learning (DL)-based deliverable volumetric arc radiation therapy (VMAT) plan generated using DL-based automated planning assistant system (AIVOT, prototype version) for patients with prostate cancer. The VMAT data (cliDose) of 68 patients with prostate cancer treated with VMAT treatment (70-74 Gy/28-37 fr) at our hospital were used (n = 55 for training and n = 13 for testing). First, a HD-U-net-based 3D dose prediction model implemented in AIVOT was customized using the VMAT data. Thus, a predictive VMAT plan (preDose) comprising AIVOT that predicted the 3D doses was generated. Second, deliverable VMAT plans (deliDose) were created using AIVOT, the radiation treatment planning system Eclipse (version 15.6) and its vender-supplied objective functions. Finally, we compared these two estimated DL-based VMAT treatment plans-i.e. preDose and deliDose-with cliDose. The average absolute dose difference of all DVH parameters for the target tissue between cliDose and deliDose across all patients was 1.32 ± 1.35% (range: 0.04-6.21%), while that for all the organs at risks was 2.08 ± 2.79% (range: 0.00-15.4%). The deliDose was superior to the cliDose in all DVH parameters for bladder and rectum. The blinded plan scoring of deliDose and cliDose was 4.54 ± 0.50 and 5.0 ± 0.0, respectively (All plans scored ≥4 points, P = 0.03.) This study demonstrated that DL-based deliverable plan for prostate cancer achieved the clinically acceptable level. Thus, the AIVOT software exhibited a potential for automated planning with no intervention for patients with prostate cancer.

Evaluation of Unity 1.5 T MR-linac plan quality in patients with prostate cancer.

The Unity magnetic resonance (MR) linear accelerator (MRL) with MR-guided adaptive radiotherapy (MRgART) is capable of online MRgART where images are acquired on the treatment day and the radiation treatment plan is immediately replanned and performed. We evaluated the MRgART plan quality and plan reproducibility of the Unity MRL in patients with prostate cancer. There were five low- or moderate-risk and five high-risk patients who received 36.25 Gy or 40 Gy, respectively in five fractions. All patients underwent simulation magnetic resonance imaging (MRI) and five online adaptive MRI. We created plans for 5, 7, 9, 16, and 20 beams and for 60, 100, and 150 segments. We evaluated the target and organ doses for different number of beams and segments, respectively. Variation in dose constraint between the simulation plan and online adaptive plan was measured for each patient to assess plan reproducibility. The plan quality improved with the increasing number of beams. However, the proportion of significantly improved dose constraints decreased as the number of beams increased. For some dose parameters, there were statistically significant differences between 60 and 100 segments, and 100 and 150 segments. The plan of five beams exhibited limited reproducibility. The number of segments had minimal impact on plan reproducibility, but 60 segments sometimes failed to meet dose constraints for online adaptive plan. The optimization and delivery time increased with the number of beams and segments. We do not recommend using five or fewer beams for a reproducible and high-quality plan in the Unity MRL. In addition, many number of segments and beams may help meet dose constraints during online adaptive plan. Treatment with the Unity MRL should be performed with the appropriate number of beams and segments to achieve a good balance among plan quality, delivery time, and optimization time.

Development of a prediction model for head and neck volume reduction by clinical factors, dose-volume histogram parameters and radiomics in head and neck cancer†.

In external radiotherapy of head and neck (HN) cancers, the reduction of irradiation accuracy due to HN volume reduction often causes a problem. Adaptive radiotherapy (ART) can effectively solve this problem; however, its application to all cases is impractical because of cost and time. Therefore, finding priority cases is essential. This study aimed to predict patients with HN cancers are more likely to need ART based on a quantitative measure of large HN volume reduction and evaluate model accuracy. The study included 172 cases of patients with HN cancer who received external irradiation. The HN volume was calculated using cone-beam computed tomography (CT) for irradiation-guided radiotherapy for all treatment fractions and classified into two groups: cases with a large reduction in the HN volume and cases without a large reduction. Radiomic features were extracted from the primary gross tumor volume (GTV) and nodal GTV of the planning CT. To develop the prediction model, four feature selection methods and two machine-learning algorithms were tested. Predictive performance was evaluated by the area under the curve (AUC), accuracy, sensitivity and specificity. Predictive performance was the highest for the random forest, with an AUC of 0.662. Furthermore, its accuracy, sensitivity and specificity were 0.692, 0.700 and 0.813, respectively. Selected features included radiomic features of the primary GTV, human papillomavirus in oropharyngeal cancer and the implementation of chemotherapy; thus, these features might be related to HN volume change. Our model suggested the potential to predict ART requirements based on HN volume reduction .

Clinical impact of radiation-induced myocardial damage detected by cardiac magnetic resonance imaging and dose-volume histogram parameters of the left ventricle as prognostic factors of cardiac events after chemoradiotherapy for esophageal cancer.

This prospective study aimed to evaluate whether radiation (RT)-induced myocardial damage by cardiac magnetic resonance (CMR) imaging could be a predictor of cardiac events after chemoradiotherapy (CRT) for esophageal cancer and determine the dose-volume histogram (DVH) parameters of the left ventricle (LV) in predicting cardiac events. CMR imaging was performed before and 6 months after CRT in patients receiving definitive CRT. RT-induced myocardial damage was defined as abnormal CMR findings indicating myocardial fibrosis corresponding to an isodose line of ≥30 Gy. The cutoff values of the LV DVH parameters were calculated using the receiver operating characteristic curve based on the presence of RT-induced myocardial damage. The prognostic factors related to cardiac events of Grade 3 or higher were examined. Twenty-three patients were enrolled in the study. RT-induced myocardial damage by late gadolinium enhancement and/or an increase of 100 ms or higher in native T1 post-CRT was detected in 10 of the 23 patients. LV V45 was the best predictive factor for RT-induced myocardial damage with a cutoff value of 2.1% and an area under the curve of 0.75. The median follow-up period was 82.1 months. The 5- and 7-year cumulative incidences of cardiac events of Grade 3 or higher were 14.7 and 22.4%, respectively. RT-induced myocardial damage and LV V45 were significant risk factors (P = 0.015 and P = 0.013, respectively). RT-induced myocardial damage is a significant predictor of cardiac events. LV V45 is associated with RT-induced myocardial damage and subsequent cardiac events.

Improvement of deep learning prediction model in patient-specific QA for VMAT with MLC leaf position map and patient's dose distribution.

PURPOSE: Deep learning-based virtual patient-specific quality assurance (QA) is a novel technique that enables patient QA without measurement. However, this method could be improved by further evaluating the optimal data to be used as input. Therefore, a deep learning-based model that uses multileaf collimator (MLC) information per control point and dose distribution in patient's CT as inputs was developed. METHODS: Overall, 96 volumetric-modulated arc therapy plans generated for prostate cancer treatment were used. We developed a model (Model 1) that can predict measurement-based gamma passing rate (GPR) for a treatment plan using data stored as a map reflecting the MLC leaf position at each control point (MLPM) and data of the dose distribution in patient's CT as inputs. The evaluation of the model was based on the mean absolute error (MAE) and Pearson's correlation coefficient (r) between the measured and predicted GPR. For comparison, we also analyzed models trained with the dose distribution in patient's CT alone (Model 2) and with dose distributions recalculated on a virtual phantom CT (Model 3). RESULTS: At the 2%/2 mm criterion, MAE[%] and r for Model 1, Model 2, and Model 3 were 2.32% ± 0.43% and 0.54 ± 0.03, 2.70% ± 0.26%, and 0.32 ± 0.08, and 2.96% ± 0.23% and 0.24 ± 0.22, respectively; at the 3%/3 mm criterion, these values were 1.25% ± 0.05% and 0.36 ± 0.18, 1.57% ± 0.35% and 0.19 ± 0.20, and 1.39% ± 0.32% and 0.17 ± 0.22, respectively. This result showed that Model 1 exhibited the lowest MAE and highest r at both criteria of 2%/2 mm and 3%3 mm. CONCLUSIONS: These findings showed that a model that combines the MLPM and dose distribution in patient's CT exhibited a better GPR prediction performance compared with the other two studied models.

Retrospective analysis of local recurrence pattern by computed tomography image-guided intracavitary and interstitial brachytherapy for locally advanced cervical cancer in a single Japanese institution.

PURPOSE: The purpose of this study was to investigate the treatment results with focus on local control (LC) by computed tomography (CT)-guided intracavity brachytherapy and interstitial brachytherapy (ICBT/ISBT) for locally advanced cervical cancer (LACC). METHODS AND MATERIALS: Patients with LACC undergoing ICBT/ISBT at least once in our institution between January 2017 and June 2019 were analyzed retrospectively. The primary endpoint was local control (LC), and the secondary endpoints were progression-free survival (PFS), overall survival (OS), and late toxicities. Differences between patient subgroups for prognostic factors in LC, PFS, and OS were analyzed using the log-rank test. The recurrence patterns of LC were also investigated. RESULTS: Forty-four patients were included in the present study. The median high-risk clinical target volume (HR-CTV) at the initial brachytherapy was 48.2 cc. The median total dose of HR-CTV D90 (EQD2) was 70.7 Gy. The median followup period was 39.4 months. The 3-year LC, PFS and OS rates in all patients were 88.2%, 56.6%, and 65.4% (95% CI 50.3-78.0%), respectively. Corpus invasion and large HR-CTV (70 cc or more) were significant prognostic factors in LC, PFS, and OS. Marginal recurrences at the fundus of the uterus were detected in 3 of 5 patients in whom local recurrence was observed. Late toxicities of Grade 3 or higher were detected in 3 patients (6.8%). CONCLUSIONS: Favorable LC was achieved by performing CT-guided ICBT/ISBT for LACC. The brachytherapy strategy for patients with corpus invasion or large HR-CTV may need to be reconsidered.

Development of a deep learning-based error detection system without error dose maps in the patient-specific quality assurance of volumetric modulated arc therapy.

To detect errors in patient-specific quality assurance (QA) for volumetric modulated arc therapy (VMAT), we proposed an error detection method based on dose distribution analysis using unsupervised deep learning approach and analyzed 161 prostate VMAT beams measured with a cylindrical detector. For performing error simulation, in addition to error-free dose distribution, dose distributions containing nine types of error, including multileaf collimator (MLC) positional errors, gantry rotation errors, radiation output errors and phantom setup errors, were generated. Only error-free data were employed for the model training, and error-free and error data were employed for the tests. As a deep learning model, the variational autoencoder (VAE) was adopted. The anomaly of test data was quantified by calculating Mahalanobis distance based on the feature vectors acquired from a trained encoder. Based on this anomaly, test data were classified as 'error-free' or 'any-error.' For comparison with conventional approaches, gamma (γ)-analysis was performed, and supervised learning convolutional neural network (S-CNN) was constructed. Receiver operating characteristic curves were obtained to evaluate their performance with the area under the curve (AUC). For all error types, except systematic MLC positional and radiation output errors, the performance of the methods was in the order of S-CNN ˃ VAE-based ˃ γ-analysis (only S-CNN required error data for model training). For example, in random MLC positional error simulation, the AUC of our method, S-CNN and γ-analysis were 0.699, 0.921 and 0.669, respectively. Our results showed that the VAE-based method has the potential to detect errors in patient-specific VMAT QA.

Development of a collapsed cone convolution/superposition dose calculation algorithm with a mass density-specific water kernel for magnetic resonance-guided radiotherapy.

This study aimed to develop and validate a collapsed cone convolution for magnetic resonance-guided radiotherapy (CCCMR). The 3D energy deposition kernels (EDKs) were generated in water in a 1.5-T transverse magnetic field. The CCCMR corrects the inhomogeneity in simulation geometry by referring to the EDKs according to the mass density between the interaction and energy deposition points in addition to density scaling. Dose distributions in a water phantom and in slab phantoms with inserted inhomogeneities were calculated using the Monte Carlo (MC) and CCCMR. The percentage depth dose (PDD) and off-axis ratio (OAR) were compared, and the gamma passing rate (3%/2 mm) was evaluated. The CCCMR simulated asymmetric dose distributions in the simulation phantoms, especially the water phantom, and all PDD and OAR profiles were in good agreement with the findings of the MC. The gamma passing rates were >99% for each field size and for the entire region. In the inhomogeneity phantoms, although the CCCMR underestimated dose in the low mass density regions, it could reconstruct dose changes at mass density boundaries. The gamma passing rate for the entire region was >95% for the field size of 2 × 2 cm2, but it was 68.9-86.7% for the field sizes of ≥5 × 5 cm2. Conclusively, in water, the CCCMR can obtain dose distributions comparable to those with the MC. Although the dose differences between them were mainly in inhomogeneity regions, the possibility of the effective use of the CCCMR in small field sizes was demonstrated.

Cardiac impacts of postoperative radiotherapy for breast cancer in Japanese patients.

Radiotherapy for breast cancer has attracted attention in Western countries because radiation to the heart can cause cardiac events. The purposes of this study were to evaluate the relationship between radiotherapy after breast-conserving surgery and the frequency of cardiac events in Japanese patients and to investigate the risk factors of cardiac events after postoperative radiotherapy in those patients. Female patients who received postoperative radiotherapy following breast-conserving surgery between 2007 and 2012 at our hospital were evaluated. In this study, we estimated the cumulative incidence of cardiac events including angina pectoris, myocardial infarction, ischemic heart disease, heart failure and cardiomyopathy after radiotherapy. Of 311 eligible patients, 7.1% of the patients had a smoking history, 20.3% of the patients were obese and 22.2% of the patients had hypertension. The median follow-up period was 118 months (interquartile range, 102-132 months). Twelve patients (3.9%) experienced cardiac events after treatment. The mean time to cardiac events was 126 months. The 10-year cumulative incidences of cardiac events after treatment were 4.2% and 4.3% for patients with left-sided and right-sided breast cancer, respectively, without a significant difference. Multivariate analysis showed that only hypertension was a risk factor for cardiac events (hazard ratio = 16.67, P = 0.0003). In conclusion, postoperative radiotherapy for breast cancer did not increase the incidence of cardiac events. Since at least 2007, postoperative radiotherapy for breast cancer has been safely performed without effects on the heart.

Impact of CT scan parameters on deformable image registration accuracy using deformable thorax phantom.

The purpose of this study was to evaluate the deformable image registration (DIR) accuracy using various CT scan parameters with deformable thorax phantom. Our developed deformable thorax phantom (Dephan, Chiyoda Technol Corp, Tokyo, Japan) was used. The phantom consists of a base phantom, an inner phantom, and a motor-derived piston. The base phantom is an acrylic cylinder phantom with a diameter of 180 mm, which simulates the chest wall. The inner phantom consists of deformable, 20 mm thick disk-shaped sponges with 48 Lucite beads and 48 nylon cross-wires which simulate the vascular and bronchial bifurcations of the lung. Peak-exhale and peak-inhale images of the deformable phantom were acquired using a CT scanner (Aquilion LB, TOSHIBA). To evaluate the impact of CT scan parameters on DIR accuracy, we used the four tube voltages (80, 100, 120, and 135 kV) and six reconstruction algorithms (FC11, FC13, FC15, FC41, FC44, and FC52). Intensity-based DIR was performed between the two images using MIM Maestro (MIM software, Cleveland, USA). Fiducial markers (beads and cross-wires) based target registration error (TRE) was used for quantitative evaluation of DIR. In case with different tube voltages, the range of average TRE were 4.44-5.69 mm (reconstruction algorithm: FC13). In case with different reconstruction algorithms, the range of average TRE were 4.26-4.59 mm (tube voltage: 120 kV). The TRE were differed by up to 3.0 mm (3.96-6.96 mm) depending on the combination of tube voltage and reconstruction algorithm. Our result indicated that CT scan parameters had moderate impact of TRE, especially for reconstruction algorithms for the deformable thorax phantom.

Development and validation of an [18F]FDG-PET/CT radiomic model for predicting progression-free survival for patients with stage II - III thoracic esophageal squamous cell carcinoma who are treated with definitive chemoradiotherapy.

BACKGROUND: Radiomics is a method for extracting a large amount of information from images and used to predict treatment outcomes, side effects and diagnosis. In this study, we developed and validated a radiomic model of [18F]FDG-PET/CT for predicting progression-free survival (PFS) of definitive chemoradiotherapy (dCRT) for patients with esophageal cancer. MATERIAL AND METHODS: Patients with stage II - III esophageal cancer who underwent [18F]FDG-PET/CT within 45 days before dCRT between 2005 and 2017 were included. Patients were randomly assigned to a training set (85 patients) and a validation set (45 patients). Radiomic parameters inside the area of standard uptake value ≥ 3 were calculated. The open-source software 3D slicer and Pyradiomics were used for segmentation and calculating radiomic parameters, respectively. Eight hundred sixty radiomic parameters and general information were investigated.In the training set, a radiomic model for PFS was made from the LASSO Cox regression model and Rad-score was calculated. In the validation set, the model was applied to Kaplan-Meier curves. The median value of Rad-score in the training set was used as a cutoff value in the validation set. JMP was used for statistical analysis. RStudio was used for the LASSO Cox regression model. p < 0.05 was defined as significant. RESULTS: The median follow-up periods were 21.9 months for all patients and 63.4 months for survivors. The 5-year PFS rate was 24.0%. In the training set, the LASSO Cox regression model selects 6 parameters and made a model. The low Rad-score group had significantly better PFS than that the high Rad-score group (p = 0.019). In the validation set, the low Rad-score group had significantly better PFS than that the high Rad-score group (p = 0.040). CONCLUSIONS: The [18F]FDG-PET/CT radiomic model could predict PFS for patients with esophageal cancer who received dCRT.

Dosimetric impact of deformable image registration using radiophotoluminescent glass dosimeters with a physical geometric phantom.

PURPOSE: To study the dosimetry impact of deformable image registration (DIR) using radiophotoluminescent glass dosimeter (RPLD) and custom developed phantom with various inserts. METHODS: The phantom was developed to facilitate simultaneous evaluation of geometric and dosimetric accuracy of DIR. Four computed tomography (CT) images of the phantom were acquired with four different configurations. Four volumetric modulated arc therapy (VMAT) plans were computed for different phantom. Two different patterns were applied to combination of four phantom configurations. RPLD dose measurement was combined between corresponding two phantom configurations. DIR-based dose accumulation was calculated between corresponding two CT images with two commercial DIR software and various DIR parameter settings, and an open source software. Accumulated dose calculated using DIR was then compared with measured dose using RPLD. RESULTS: The mean ± standard deviation (SD) of dose difference was 2.71 ± 0.23% (range, 2.22%-3.01%) for tumor-proxy and 3.74 ± 0.79% (range, 1.56%-4.83%) for rectum-proxy. The mean ± SD of target registration error (TRE) was 1.66 ± 1.36 mm (range, 0.03-4.43 mm) for tumor-proxy and 6.87 ± 5.49 mm (range, 0.54-17.47 mm) for rectum-proxy. These results suggested that DIR accuracy had wide range among DIR parameter setting. CONCLUSIONS: The dose difference observed in our study was 3% for tumor-proxy and within 5% for rectum-proxy. The custom developed physical phantom with inserts showed potential for accurate evaluation of DIR-based dose accumulation. The prospect of simultaneous evaluation of geometric and dosimetric DIR accuracy in a single phantom may be useful for validation of DIR for clinical use.

Radiation pneumonitis prediction model with integrating multiple dose-function features on 4DCT ventilation images.

PURPOSE: Radiation pneumonitis (RP) is dose-limiting toxicity for non-small-cell cancer (NSCLC). This study developed an RP prediction model by integrating dose-function features from computed four-dimensional computed tomography (4DCT) ventilation using the least absolute shrinkage and selection operator (LASSO). METHODS: Between 2013 and 2020, 126 NSCLC patients were included in this study who underwent a 4DCT scan to calculate ventilation images. We computed two sets of candidate dose-function features from (1) the percentage volume receiving > 20 Gy or the mean dose on the functioning zones determined with the lower cutoff percentile ventilation value, (2) the functioning zones determined with lower and upper cutoff percentile ventilation value using 4DCT ventilation images. An RP prediction model was developed by LASSO while simultaneously determining the regression coefficient and feature selection through fivefold cross-validation. RESULTS: We found 39.3 % of our patients had a ≥ grade 2 RP. The mean area under the curve (AUC) values for the developed models using clinical, dose-volume, and dose-function features with a lower cutoff were 0.791, and the mean AUC values with lower and upper cutoffs were 0.814. The relative regression coefficient (RRC) on dose-function features with upper and lower cutoffs revealed a relative impact of dose to each functioning zone to RP. RRCs were 0.52 for the mean dose on the functioning zone, with top 20 % of all functioning zone was two times greater than that of 0.19 for these with 60 %-80 % and 0.17 with 40 %-60 % (P < 0.01). CONCLUSIONS: The introduction of dose-function features computed from functioning zones with lower and upper cutoffs in a machine learning framework can improve RP prediction. The RRC given by LASSO using dose-function features allows for the quantification of the RP impact of dose on each functioning zones and having the potential to support treatment planning on functional image-guided radiotherapy.

Longitudinal analyses and predictive factors of radiation-induced lung toxicity-related parameters after stereotactic radiotherapy for lung cancer.

BACKGROUND AND PURPOSE: The purpose of this prospective study was to investigate changes in longitudinal parameters after stereotactic radiotherapy for lung cancer and to identify possible pretreatment factors related to radiation-induced lung toxicity and the decline in pulmonary function after radiotherapy. MATERIALS AND METHODS: Protocol-specified examinations, including 4-D CT, laboratory tests, pulmonary function tests (PFTs) and body composition measurements, were performed before SRT and at 1 month, 4 months and 12 months after stereotactic radiotherapy. Longitudinal differences were tested by using repeated-measures analysis of variance. Correlations were examined by using the Pearson product-moment correlation coefficient (r). RESULTS: Sixteen patients were analyzed in this study. During a median follow-up period of 26.6 months, grade 1 and 2 lung toxicity occurred in 11 patients and 1 patient, respectively. The mean Hounsfield units (HU) and standard deviation (SD) of the whole lung, as well as sialylated carbohydrate antigen KL-6 (KL-6) and surfactant protein-D (SP-D), peaked at 4 months after radiotherapy (p = 0.11, p<0.01, p = 0.04 and p<0.01, respectively). At 4 months, lung V20 Gy (%) and V40 Gy (%) were correlated with changes in SP-D, whereas changes in the mean HU of the lung were related to body mass index and lean body mass index (r = 0.54, p = 0.02; r = 0.57, p = 0.01; r = 0.69, p<0.01; and r = 0.69, p<0.01, respectively). The parameters of PFTs gradually declined over time. When regarding the change in PFTs from pretreatment to 12 months, lung V5 Gy (cc) showed significant correlations with diffusion capacity for carbon monoxide (DLCO), DLCO/alveolar volume and the relative change in DLCO (r = -0.72, p<0.01; r = -0.73, p<0.01; and r = -0.63, p = 0.01, respectively). CONCLUSIONS: The results indicated that some parameters peaked at 4 months, but PFTs were the lowest at 12 months. Significant correlations between lung V5 Gy (cc) and changes in DLCO and DLCO/alveolar volume were observed.

[Evaluation of Accuracy of Deformable Image Registration with Newly Developed Treatment Planning Support Software for Thoracic Images].

This study evaluated accuracy of deformable image registration (DIR) with twelve parameter settings for thoracic images. We used peak-inhale and peak-exhale images for ten patients provided by DIR-lab. We used a prototype version of iCView software (ITEM Corporation) with DIR to perform intensity, structure, and hybrid-based DIR with the twelve parameter settings. DIR accuracy was evaluated by a target registration error (TRE) using 300 bronchial bifurcations and the Dice similarity coefficient (DSC) of the lungs. For twelve parameter settings, TRE ranged from 2.83 mm to 5.27 mm, whereas DSC ranged from 0.96 to 0.98. These results demonstrated that DIR accuracy differed among parameter settings and show that appropriate parameter settings are required for clinical practice.