This study investigated international standard systems for radiotherapy, focusing on the comparison of the radiation protection standards for medical linear accelerators adopted in China, the UK, and the USA. Despite some specific differences, the standards for radiotherapy rooms in the three countries generally adhere to the basic principles set by the International Commission on Radiological Protection (ICRP) and the International Atomic Energy Agency (IAEA). Regarding the zoning principle of radiotherapy rooms, the definitions of the controlled areas are similar in China, the UK, and the USA, while the classification of areas beyond the controlled areas differs across the three countries. In terms of measurement conditions, all the three countries require measurements under the maximum output dose of the radiotherapy equipment, with only minor differences in details. For dose limits and compliance criteria for radiation shielding of radiotherapy rooms, China adopts the highest instantaneous dose rate as the control threshold. In contrast, the UK and the USA base their standards on dose limits evaluated over certain time intervals (annual, weekly, and daily), assessing compliance through workload calculation. In terms of method for deducing and calculating effective dose limits, UK standards stipulate that annual personnel exposure should be calculated using instantaneous dose rates. In comparison, the USA provides specific method for calculating dose rates per week and any one hour from instantaneous dose rates. The comparative analysis indicates that China′s method, which is based on the maximum instantaneous dose rates, may lead to increased construction costs of radiotherapy rooms under the same conditions, hindering the application and development of novel radiotherapy technologies. To address these concerns while maintaining radiation safety, it is recommended that China consider introducing method based on average instantaneous dose rates or calculating the annual personnel exposure doses. This will help optimize protection standards and advance radiotherapy technology.

Objective:To construct a general deep learning dose prediction model applicable to radiotherapy for head and neck tumors, establish design methods for artificial intelligence (AI)-assisted radiotherapy plan and evaluate the accuracy of prediction.Methods:Radiotherapy plans of 818 patients who received radiotherapy for head and neck cancers from January 2018 to June 2021 in Cancer Hospital of Chinese Academy of Medical Sciences were enrolled. Patients involved 17 types of common head and neck cancers, and the prescribed dose covered 5 kinds of dose gradients ranging from 54 Gy to 73.92 Gy. And 1-2 cases per each cancer type (31 cases in total) were randomly selected as the validation set, and the remaining 787 cases were used as the training set to build a deep learning head and neck radiotherapy generalized dose prediction model. Then based on the dose prediction results of this model, a program was written to automatically generate inverse optimization condition scripts, which were sent back to the treatment planning system to achieve AI-assisted radiotherapy plan design. Among the patients who received radiotherapy in our hospital from June 2021 to January 2022, 1 patient for each disease type (17 cases in total) was selected to evaluate the AI-assisted plan design program and evaluate its clinical feasibility using paired t-test. Results:Dose prediction model accuracy evaluation revealed that in the 31-case validation set, there was no statistical difference in the evaluation metrics of clinical concern for organs at risks, except for the D 1 cm3 prediction for spinal cord planning risk volume, which was statistically different compared with the clinical reference plan. The AI-assisted plan design program had higher plan quality metric scores (37.88±6.42) than manual plans (35.00±7.63) in 17 test cases ( t=-1.00, P=0.166). The number of manual adjustments to the inverse optimization conditions was reduced from (5.47±2.97) times to (2.76±1.00) times for the AI-assisted plan compared to the manual-only plan ( t=4.12, P<0.001). And the number of outlined dose shaping structures was reduced from 7.35±3.98 to 3.12±1.18 ( t=5.61, P<0.001). Conclusions:The unified universal model of dose prediction established for different head and neck cancers has high accuracy in dose prediction for all types of head and neck tumor plans. The AI-assisted planning method established in this pattern can reduce the clinical workload of physicists and improve the efficiency of their work.

Artificial intelligence (AI) technologies are being widely applied in radiotherapy. However, the integration of AI into clinical workflows of radiotherapy faces a series of challenges, such as poor model interpretability, domain shifts between clinical application and training data, and the inherent model uncertainties. Therefore, case-specific quality assurance (QA) is essential before deploying AI models in clinical practice. This paper reviews and summarizes QA methodologies for the application of AI models in radiotherapy across four key areas: image registration, image generation, region of interest segmentation, and treatment planning.

Objective:To evaluate the feasibility of using the domestic ArcherQA system for fast and simplified independent verification of CyberKnife (CK) stereotactic radiotherapy (SRT) plans.Methods:SRT plans of 57 patients treated with CK at Tianjin Medical University Cancer Institute and Hospital from August 2021 to August 2022 were retrospectively analyzed, including 15 intracranial, 30 pulmonary, and 12 abdominal tumors cases. Point-dose and planar-dose verifications were performed using an ionization chamber and radiochromic films embedded in a homogeneous phantom, and the results were compared with those calculated by the treatment planning system (TPS). The localization CT images and corresponding SRT plans were imported into the ArcherQA system for independent dose verification and analysis. The correlation between ArcherQA results and phantom measurements was analyzed, with comparisons of target mean dose differences and γ pass rates.Results:Phantom measurement results showed, the measured point-dose differences for intracranial, lung, and abdominal plans were -0.94% ± 3.22%, 1.92% ± 2.05%, and 2.12% ± 0.77%, respectively. The mean dose differences in target dose calculation between ArcherQA and TPS: intracranial in the gross tumor volume (GTV) regions were 0.34% ± 2.21%, lung tumor GTV were -2.47% ± 2.46%, and abdominal tumor GTV were 0.80% ± 2.61%, respectively. Among them, the abdominal GTV region showed the highest correlation between ArcherQA and measured results ( r=0.78). The average two-dimensional γ pass rates (2 mm/2%, threshold=10%) measured using phantom films were 95.92% ± 2.35% for intracranial, 95.70% ± 2.74% for lung, and 96.74% ± 3.41% for abdominal tumors plans, respectively. The three-dimensional ArcherQA results showed comparable γ pass rates (1 mm/2%, threshold=10%) for lung and abdominal GTV and PTV regions, with similar medians and data dispersion to film measurements. Conclusions:The ArcherQA system enables rapid and efficient independent dose verification of CK SRT plans without the need for additional hardware. The verification results show good correlation with phantom measurements, supporting its potential as an auxiliary quality assurance tool in clinical CK SRT implementation.

This study investigated international standard systems for radiotherapy, focusing on the comparison of the radiation protection standards for medical linear accelerators adopted in China, the UK, and the USA. Despite some specific differences, the standards for radiotherapy rooms in the three countries generally adhere to the basic principles set by the International Commission on Radiological Protection (ICRP) and the International Atomic Energy Agency (IAEA). Regarding the zoning principle of radiotherapy rooms, the definitions of the controlled areas are similar in China, the UK, and the USA, while the classification of areas beyond the controlled areas differs across the three countries. In terms of measurement conditions, all the three countries require measurements under the maximum output dose of the radiotherapy equipment, with only minor differences in details. For dose limits and compliance criteria for radiation shielding of radiotherapy rooms, China adopts the highest instantaneous dose rate as the control threshold. In contrast, the UK and the USA base their standards on dose limits evaluated over certain time intervals (annual, weekly, and daily), assessing compliance through workload calculation. In terms of method for deducing and calculating effective dose limits, UK standards stipulate that annual personnel exposure should be calculated using instantaneous dose rates. In comparison, the USA provides specific method for calculating dose rates per week and any one hour from instantaneous dose rates. The comparative analysis indicates that China′s method, which is based on the maximum instantaneous dose rates, may lead to increased construction costs of radiotherapy rooms under the same conditions, hindering the application and development of novel radiotherapy technologies. To address these concerns while maintaining radiation safety, it is recommended that China consider introducing method based on average instantaneous dose rates or calculating the annual personnel exposure doses. This will help optimize protection standards and advance radiotherapy technology.

Objective:To construct a general deep learning dose prediction model applicable to radiotherapy for head and neck tumors, establish design methods for artificial intelligence (AI)-assisted radiotherapy plan and evaluate the accuracy of prediction.Methods:Radiotherapy plans of 818 patients who received radiotherapy for head and neck cancers from January 2018 to June 2021 in Cancer Hospital of Chinese Academy of Medical Sciences were enrolled. Patients involved 17 types of common head and neck cancers, and the prescribed dose covered 5 kinds of dose gradients ranging from 54 Gy to 73.92 Gy. And 1-2 cases per each cancer type (31 cases in total) were randomly selected as the validation set, and the remaining 787 cases were used as the training set to build a deep learning head and neck radiotherapy generalized dose prediction model. Then based on the dose prediction results of this model, a program was written to automatically generate inverse optimization condition scripts, which were sent back to the treatment planning system to achieve AI-assisted radiotherapy plan design. Among the patients who received radiotherapy in our hospital from June 2021 to January 2022, 1 patient for each disease type (17 cases in total) was selected to evaluate the AI-assisted plan design program and evaluate its clinical feasibility using paired t-test. Results:Dose prediction model accuracy evaluation revealed that in the 31-case validation set, there was no statistical difference in the evaluation metrics of clinical concern for organs at risks, except for the D 1 cm3 prediction for spinal cord planning risk volume, which was statistically different compared with the clinical reference plan. The AI-assisted plan design program had higher plan quality metric scores (37.88±6.42) than manual plans (35.00±7.63) in 17 test cases ( t=-1.00, P=0.166). The number of manual adjustments to the inverse optimization conditions was reduced from (5.47±2.97) times to (2.76±1.00) times for the AI-assisted plan compared to the manual-only plan ( t=4.12, P<0.001). And the number of outlined dose shaping structures was reduced from 7.35±3.98 to 3.12±1.18 ( t=5.61, P<0.001). Conclusions:The unified universal model of dose prediction established for different head and neck cancers has high accuracy in dose prediction for all types of head and neck tumor plans. The AI-assisted planning method established in this pattern can reduce the clinical workload of physicists and improve the efficiency of their work.

Artificial intelligence (AI) technologies are being widely applied in radiotherapy. However, the integration of AI into clinical workflows of radiotherapy faces a series of challenges, such as poor model interpretability, domain shifts between clinical application and training data, and the inherent model uncertainties. Therefore, case-specific quality assurance (QA) is essential before deploying AI models in clinical practice. This paper reviews and summarizes QA methodologies for the application of AI models in radiotherapy across four key areas: image registration, image generation, region of interest segmentation, and treatment planning.

Objective:To evaluate the feasibility of using the domestic ArcherQA system for fast and simplified independent verification of CyberKnife (CK) stereotactic radiotherapy (SRT) plans.Methods:SRT plans of 57 patients treated with CK at Tianjin Medical University Cancer Institute and Hospital from August 2021 to August 2022 were retrospectively analyzed, including 15 intracranial, 30 pulmonary, and 12 abdominal tumors cases. Point-dose and planar-dose verifications were performed using an ionization chamber and radiochromic films embedded in a homogeneous phantom, and the results were compared with those calculated by the treatment planning system (TPS). The localization CT images and corresponding SRT plans were imported into the ArcherQA system for independent dose verification and analysis. The correlation between ArcherQA results and phantom measurements was analyzed, with comparisons of target mean dose differences and γ pass rates.Results:Phantom measurement results showed, the measured point-dose differences for intracranial, lung, and abdominal plans were -0.94% ± 3.22%, 1.92% ± 2.05%, and 2.12% ± 0.77%, respectively. The mean dose differences in target dose calculation between ArcherQA and TPS: intracranial in the gross tumor volume (GTV) regions were 0.34% ± 2.21%, lung tumor GTV were -2.47% ± 2.46%, and abdominal tumor GTV were 0.80% ± 2.61%, respectively. Among them, the abdominal GTV region showed the highest correlation between ArcherQA and measured results ( r=0.78). The average two-dimensional γ pass rates (2 mm/2%, threshold=10%) measured using phantom films were 95.92% ± 2.35% for intracranial, 95.70% ± 2.74% for lung, and 96.74% ± 3.41% for abdominal tumors plans, respectively. The three-dimensional ArcherQA results showed comparable γ pass rates (1 mm/2%, threshold=10%) for lung and abdominal GTV and PTV regions, with similar medians and data dispersion to film measurements. Conclusions:The ArcherQA system enables rapid and efficient independent dose verification of CK SRT plans without the need for additional hardware. The verification results show good correlation with phantom measurements, supporting its potential as an auxiliary quality assurance tool in clinical CK SRT implementation.

Objective:To provide theoretical basis for strength and endurance training of waist and abdominal muscles and prevention of waist injuries in fighter pilots by exploring the muscle strength and work characteristics of the waist and abdominal flexor and extensor muscles in fighter pilots.Methods:Sixty male fighter pilots who were qualified for flight in aeromedical assessment, aged 24-46 years old, were randomly selected and divided into 24-30, 31-40, 41-46 years group. The Isomed2000 isokinetic dynamometer system was applied to measure the muscle strength and work performance of the abdominal and lumbar flexors and extensors of the fighter pilots with the angular velocities of 60°/s and 180°/s. The flexion and extension muscle peak torque, relative peak torque, flexion-extension ratio, and endurance ratio were compared among different age groups of pilots.Results:At the same angular velocity, the peak torque and relative peak torque of the lumbar and abdominal extensor muscles in fighter pilots were greater than those of the flexor muscles, with statistically significant differences ( t=7.01-9.13, all P<0.001). The peak torque and relative peak torque of the lumbar and abdominal flexor and extensor muscles significantly decreased with increasing test angular velocity ( t=13.63-17.25, all P<0.001). Under the angular velocity of 60°/s, there were no significant differences in the peak torque and relative peak torque of the flexor muscles among different age groups (both P>0.05); there were significant differences in the peak torque and relative peak torque of extensor muscles ( F=5.31, 6.61, P=0.008, 0.003) and 41-46 years groups were lower than the other 2 groups ( P=0.019, 0.003, 0.002, 0.004). Under an angular velocity of 180°/s, there were significant differences in the peak torque and relative peak torque of the waist and abdominal flexor and extensor muscles among different age groups ( F=3.82, 3.47, 3.83, 5.49, P=0.028, 0.043, 0.027, 0.008); the relative peak torque of the abdominal and lumbar flexor and extensor muscles in the 24-30 years group was higher than that in the 41-46 years group ( P=0.032, 0.006). The peak torque and the relative peak torque of the abdominal and lumbar flexor muscles in 31-40 years group were higher than those in 41-46 years group ( P=0.008, 0.013). The low qualification rate of peak torque ratio indicated the imbalance between flexor and extensor muscles and the poor endurance of abdominal and lumbar flexor and extensor muscles than that of flexor muscles. There was no statistically significant difference in the endurance ratio of the abdominal and lumbar flexor and extensor muscles among different age groups ( P>0.05). Conclusions:The balance between flexor and extensor muscles of waist and abdominal muscles should be paid more in fighter pilot′s fitness training. For the pilots older than 40 the training targeted to slow the muscles decline is necessary.