A deep learning‐based prediction model for gamma evaluation in patient‐specific quality assurance

Tomori, Seiji; Kadoya, Noriyuki; Takayama, Yoshiki; Kajikawa, Tomohiro; Shima, Katsumi; Narazaki, Kakutarou; Jingu, Keiichi

doi:10.1002/mp.13112

Cited by 108 publications

(150 citation statements)

References 27 publications

Supporting

Mentioning

145

Contrasting

Order By: Relevance

“…Another example is a comparison with the DL‐based method as an ultimate approach which automatically involves multiple characteristics. Tomori et al analyzed data of composite dose distributions measured using a gafchromic film and obtained a root mean square error of 1.11%, 1.50%, and 2.24% for the 3%/3, 3%/2, and 2%/2 mm tolerances, which were smaller than our results (2.3, 4.1, and 6.7%) by factor of 2, 3, and 3, respectively. Since our data were obtained from the Delta4 system with a 5‐mm pitch detector array, direct comparisons of these results do not provide an exact goal to achieve.…”

Section: Discussioncontrasting

confidence: 83%

“…The other approach is a direct estimation of the GPR. Our DUP‐based method, the DL‐based method, and the ML‐based method are classified into this direct estimation. This approach utilizes some parameters which have characteristics of the IM beam and good proportionality to the GPR.…”

Section: Discussionmentioning

confidence: 99%

“…A complexity metric (CM) that may be related to the GPR has been proposed by several authors, resulting in limited predictability of the GPR. Deep learning (DL)‐based prediction showed a good performance for the planar dose distribution measured using the film . Recently, machine learning (ML) technique was used to attempt the GPR prediction using the CMs as input data …”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Predictive gamma passing rate for three‐dimensional dose verification with finite detector elements via improved dose uncertainty potential accumulation model

Shiba¹,

Saito²,

Furumi³

et al. 2020

Medical Physics

View full text Add to dashboard Cite

Purpose We aim to develop a method to predict the gamma passing rate (GPR) of a three‐dimensional (3D) dose distribution measured by the Delta4 detector system using the dose uncertainty potential (DUP) accumulation model. Methods Sixty head‐and‐neck intensity‐modulated radiation therapy (IMRT) treatment plans were created in the XiO treatment planning system. All plans were created using nine step‐and‐shoot beams of the ONCOR linear accelerator. Verification plans were created and measured by the Delta4 system. The planar DUP (pDUP) manifesting on a field edge was generated from the segmental aperture shape with a Gaussian folding on the beam's‐eye view. The DUP at each voxel (u) was calculated by projecting the pDUP on the Delta4 phantom with its attenuation considered. The learning model (LM), an average GPR as a function of the DUP, was approximated by an exponential function aGPRu=e-qu to compensate for the low statistics of the learning data due to a finite number of the detectors. The coefficient q was optimized to ensure that the difference between the measured and predicted GPRs (dGPR) was minimized. The standard deviation (SD) of the dGPR was evaluated for the optimized LM. Results It was confirmed that the coefficient q was larger for tighter tolerance. This result corresponds to the expectation that the attenuation of the aGPRu will be large for tighter tolerance. The pGPR and mGPR were observed to be proportional for all tolerances investigated. The SD of dGPR was 2.3, 4.1, and 6.7% for tolerances of 3%/3 mm, 3%/2 mm, 2%/2 mm, respectively. Conclusion The DUP‐based predicting method of the GPR was extended to 3D by introducing DUP attenuation and an optimized analytical LM to compensate for the low statistics of the learning data due to a finite number of detector elements. The precision of the predicted GPR is expected to be improved by improving the LM and by involving other metrics.

show abstract

Section: Discussioncontrasting

confidence: 83%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Predictive gamma passing rate for three‐dimensional dose verification with finite detector elements via improved dose uncertainty potential accumulation model

Shiba¹,

Saito²,

Furumi³

et al. 2020

Medical Physics

View full text Add to dashboard Cite

show abstract

“…However, this measurement‐based QA still requires either setup or beam delivery times and cannot predict unacceptable‐quality plans . Recently, prediction of dosimetric accuracy has been developed as a more efficient patient‐specific QA method than measurement‐based QA . In fact, prediction does not require setup or beam delivery times, leading to increased adoption of IMRT and VMAT in clinical facilities.…”

Section: Introductionmentioning

confidence: 99%

“…Sum et al reported the prediction of beam output and derived MUs using machine learning . In recent years, there is a growing interest in applying machine learning for predicting the gamma passing rate for IMRT . For instance, Valdes et al predicted the gamma passing rate of 498 IMRT plans using 78 parameters, including MU, linac type, and position of multi‐leaf collimator .…”

Section: Introductionmentioning

confidence: 99%

Prediction of dosimetric accuracy for VMAT plans using plan complexity parameters via machine learning

et al. 2019

View full text Add to dashboard Cite

Purpose The dosimetric accuracies of volumetric modulated arc therapy (VMAT) plans were predicted using plan complexity parameters via machine learning. Methods The dataset consisted of 600 cases of clinical VMAT plans from a single institution. The predictor variables (n = 28) for each plan included complexity parameters, machine type, and photon beam energy. Dosimetric measurements were performed using a helical diode array (ArcCHECK), and the dosimetric accuracy of the passing rates for a 5% dose difference (DD5%) and gamma index of 3%/3 mm (γ3%/3 mm) were predicted using three machine learning models: regression tree analysis (RTA), multiple regression analysis (MRA), and neural networks (NNs). First, the prediction models were applied to 500 cases of the VMAT plans. Then, the dosimetric accuracy was predicted using each model for the remaining 100 cases (evaluation dataset). The error between the predicted and measured passing rates was evaluated. Results For the 600 cases, the mean ± standard deviation of the measured passing rates was 92.3% ± 9.1% and 96.8% ± 3.1% for DD5% and γ3%/3 mm, respectively. For the evaluation dataset, the mean ± standard deviation of the prediction errors for DD5% and γ3%/3 mm was 0.5% ± 3.0% and 0.6% ± 2.4% for RTA, 0.0% ± 2.9% and 0.5% ± 2.4% for MRA, and –0.2% ± 2.7% and –0.2% ± 2.1% for NN, respectively. Conclusions NNs performed slightly better than RTA and MRA in terms of prediction error. These findings may contribute to increasing the efficiency of patient‐specific quality‐assurance procedures.

show abstract

Auto‐Trending daily quality assurance program for a pencil beam scanning proton system aligned with TG 224

Shi

Chen

et al. 2020

J Applied Clin Med Phys

View full text Add to dashboard Cite

The Daily Quality Assurance (DQA) for a proton modality is not standardized. The modern pencil beam scanning proton system is becoming a trend and an increasing number of proton centers with PBS are either under construction or in planning. The American Association of Physicists in Medicine has a Task Group 224 report published in 2019 for proton modality routine QA. Therefore, there is a clinical need to explore a DQA procedure to meet the TG 224 guideline. The MatriXX PT and a customized phantom were used for the dosimetry constancy checking. An OBI box was used for imaging QA. The MyQA(TM) software was used for logging the dosimetry results. An in‐house developed application was applied to log and auto analyze the DQA results. Another in‐house developed program "DailyQATrend" was used to create DQA databases for further analysis. All the functional and easy determined tasks passed. For dosimetry constancy checking, the outputs for four gantry rooms were within ±3% with room to room baseline differences within ±1%. The energy checking was within ±1%. The spot location checking from the baseline was within 0.63 mm and the spot size checking from the baseline was within −1.41 ± 1.27 mm (left–right) and −0.24 ± 1.27 mm (in–out) by averaging all the energies. We have found that there was also a trend for the beam energies of two treatment rooms slowly going down (0.76% per month and 0.48 per month) after analyzing the whole data trend with linear regression. A DQA program for a PBS proton system has been developed and fully implemented into the clinic. The DQA program meets the TG 224 guideline and has web‐based logging and auto treading functions. The clinical data show the DQA program is efficient and has the potential to identify the PBS proton system potential issue.

show abstract

A deep learning‐based prediction model for gamma evaluation in patient‐specific quality assurance

Abstract: We developed a CNN-based prediction model for patient-specific QA of dose distribution in prostate treatment. Our results suggest that deep learning may provide a useful prediction model for gamma evaluation of patient-specific QA in prostate treatment planning.

Cited by 108 publications

References 27 publications

Predictive gamma passing rate for three‐dimensional dose verification with finite detector elements via improved dose uncertainty potential accumulation model

Predictive gamma passing rate for three‐dimensional dose verification with finite detector elements via improved dose uncertainty potential accumulation model

Prediction of dosimetric accuracy for VMAT plans using plan complexity parameters via machine learning

Auto‐Trending daily quality assurance program for a pencil beam scanning proton system aligned with TG 224

Contact Info

Product

Resources

About