Organ dose prediction for patients undergoing radiotherapy CBCT chest examinations using artificial intelligence.

PURPOSE: To propose an artificial intelligence (AI)-based method for personalized and real-time dosimetry for chest CBCT acquisitions. METHODS: CT images from 113 patients who underwent radiotherapy treatment were collected for simulating thorax examinations using cone-beam computed tomography (CBCT) with the Monte Carlo technique. These simulations yielded organ dose data, used to train and validate specific AI algorithms. The efficacy of these AI algorithms was evaluated by comparing dose predictions with the actual doses derived from Monte Carlo simulations, which are the ground truth, utilizing Bland-Altman plots for this comparative analysis. RESULTS: The absolute mean discrepancies between the predicted doses and the ground truth are (0.9 ± 1.3)% for bones, (1.2 ± 1.2)% for the esophagus, (0.5 ± 1.3)% for the breast, (2.5 ± 1.4)% for the heart, (2.4 ± 2.1)% for lungs, (0.8 ± 0.6)% for the skin, and (1.7 ± 0.7)% for integral. Meanwhile, the maximum discrepancies between the predicted doses and the ground truth are (14.4 ± 1.3)% for bones, (12.9 ± 1.2)% for the esophagus, (9.4 ± 1.3)% for the breast, (14.6 ± 1.4)% for the heart, (21.2 ± 2.1)% for lungs, (10.0 ± 0.6)% for the skin, and (10.5 ± 0.7)% for integral. CONCLUSIONS: AI models that can make real-time predictions of the organ doses for patients undergoing CBCT thorax examinations as part of radiotherapy pre-treatment positioning were developed. The results of this study clearly show that the doses predicted by analyzed AI models are in close agreement with those calculated using Monte Carlo simulations.

A fully automated machine learning-based methodology for personalized radiation dose assessment in thoracic and abdomen CT.

PURPOSE: To develop a machine learning-based methodology for patient-specific radiation dosimetry in thoracic and abdomen CT. METHODS: Three hundred and thirty-one thoracoabdominal radiotherapy-planning CT examinations with the respective organ/patient contours were collected retrospectively for the development and validation of segmentation 3D-UNets. Moreover, 97 diagnostic thoracic and 89 diagnostic abdomen CT examinations were collected retrospectively. For each of the diagnostic CT examinations, personalized MC dosimetry was performed. The data derived from MC simulations along with the respective CT data were used for the training and validation of a dose prediction deep neural network (DNN). An algorithm was developed to utilize the trained models and perform patient-specific organ dose estimates for thoracic and abdomen CT examinations. The doses estimated with the DNN were compared with the respective doses derived from MC simulations. A paired t-test was conducted between the DNN and MC results. Furthermore, the time efficiency of the proposed methodology was assessed. RESULTS: The mean percentage differences (range) between DNN and MC dose estimates for the lungs, liver, spleen, stomach, and kidneys were 7.2 % (0.2-24.1 %), 5.5 % (0.4-23.0 %), 7.9 % (0.6-22.3 %), 6.9 % (0.0-23.0 %) and 6.7 % (0.3-22.6 %) respectively. The differences between DNN and MC dose estimates were not significant (p-value = 0.12). Moreover, the mean processing time of the proposed workflow was 99 % lower than the respective time needed for MC-based dosimetry. CONCLUSIONS: The proposed methodology can be used for rapid and accurate patient-specific dosimetry in chest and abdomen CT.

Rapid estimation of patient-specific organ doses using a deep learning network.

BACKGROUND: Patient-specific organ-dose estimation in diagnostic CT examinations can provide useful insights on individualized secondary cancer risks, protocol optimization, and patient management. Current dose estimation techniques mainly rely on time-consuming Monte Carlo methods or/and generalized anthropomorphic phantoms. PURPOSE: We proposed a proof-of-concept rapid workflow based on deep learning networks to estimate organ doses for individuals following thorax Computed Tomography (CT) examinations. METHODS: CT scan data from 95 individuals undergoing thorax CT examinations were used. Monte Carlo simulations were performed and three-dimensional (3D) dose distributions for each patient were obtained. A fully connected sequential deep learning network model was constructed and trained for each organ considered in this study. Water-equivalent diameter (WED), scan length, and tube current were the independent variables. Organ doses for heart, lungs, esophagus, and bones were calculated from the Monte Carlo 3D distribution and used to train the deep learning networks. Organ dose predictions from each network were evaluated using an independent data set of 19 patients. RESULTS: The trained networks provided organ dose predictions within a second. There was very good agreement between the deep learning network predictions and reference organ dose values calculated from Monte Carlo simulations. The average difference was -1.5% for heart, -1.6% for esophagus, -1.0% for lungs, and -0.4% for bones in the 95 patients dataset, and -5.1%, 4.3%, 0.9%, and 1.4% respectively in the 19 patients test dataset. CONCLUSIONS: The proposed workflow demonstrated that patient-specific organ-doses can be estimated in nearly real-time using deep learning networks. The workflow can be readily implemented and requires a small set of representative data for training.

Calculation of organ doses from breast cancer radiotherapy: a Monte Carlo study.

The current study aimed to: a) utilize Monte Carlo simulation methods for the assessment of radiation doses imparted to all organs at risk to develop secondary radiation induced cancer, for patients undergoing radiotherapy for breast cancer; and b) evaluate the effect of breast size on dose to organs outside the irradiation field. A simulated linear accelerator model was generated. The in-field accuracy of the simulated photon beam properties was verified against percentage depth dose (PDD) and dose profile measurements on an actual water phantom. Off-axis dose calculations were verified with thermoluminescent dosimetry (TLD) measurements on a humanoid physical phantom. An anthropomorphic mathematical phantom was used to simulate breast cancer radiotherapy with medial and lateral fields. The effect of breast size on the calculated organ dose was investigated. Local differences between measured and calculated PDDs and dose profiles did not exceed 2% for the points at depths beyond the depth of maximum dose and the plateau region of the profile, respectively. For the penumbral regions of the dose profiles, the distance to agreement (DTA) did not exceed 2 mm. The mean difference between calculated out-of-field doses and TLD measurements was 11.4% ± 5.9%. The calculated doses to peripheral organs ranged from 2.32 cGy up to 161.41 cGy depending on breast size and thus the field dimensions applied, as well as the proximity of the organs to the primary beam. An increase to the therapeutic field area by 50% to account for the large breast led to a mean organ dose elevation by up to 85.2% for lateral exposure. The contralateral breast dose ranged between 1.4% and 1.6% of the prescribed dose to the tumor. Breast size affects dose deposition substantially.

Radiation dose and risk from fluoroscopically guided percutaneous transluminal angioplasty and stenting in the abdominal region.

The objective of this study was to estimate the radiation dose and associated risks resulting from fluoroscopically guided percutaneous transluminal angioplasty with or without stent placement in the abdominal region. Average examination parameters for renal and aortoiliac procedures were derived using data from 80 consecutive procedures performed in our institute. Organ and effective doses were estimated for endovascular procedures with the use of a Monte Carlo (MC) transport code and an adult mathematical phantom. Thermoluminescent dosimeters were used in an anthropomorphic phantom to verify MC calculations. Radiation-induced risks were estimated. Results are presented as doses normalized to dose area product, so that the patient dose from any technique and X-ray unit can be easily calculated for iliac and renal PTA/stenting sessions. The average effective dose varied from 75 to 371 microSv per Gycm(2) depending on the beam quality, procedure scheme and sex of the patient. Differences up to 17% were observed between MC-calculated data and data derived from thermoluminescent dosimetry. The radiation-induced cancer risk may be considerable for younger individuals undergoing transluminal angioplasty with stent placement.

Radiation dose and risk from fluoroscopically guided percutaneous transhepatic biliary procedures.

PURPOSE: To estimate radiation dose and associated risks after fluoroscopically guided percutaneous transhepatic biliary (PTB) drainage and stent implantation procedures. MATERIALS AND METHODS: Organ and effective doses, normalized to dose-area product (DAP), were estimated for PTB procedures with use of a Monte Carlo transport code and an adult mathematical phantom. Exposure parameters from 51 consecutive patients were used to determine average examination parameters for biliary drainage and stent implantation procedures. Thermoluminescent dosimeters were used in an anthropomorphic phantom to verify Monte Carlo calculations. Radiation-induced cancer and genetic risks were estimated. RESULTS: The results consist of doses normalized to DAP so patient dose from any technique and x-ray unit can be easily calculated for left and right biliary access and for separate or combined biliary and metallic stent implantation sessions. A good agreement was found between Monte Carlo-calculated data and data derived from thermoluminescent dosimetry. The average effective dose varied from 1.8 to 5.4 mSv depending on procedure approach (left vs right access) and procedure scheme. A maximum effective dose of 13 mSv was estimated for 30 minutes of fluoroscopy. CONCLUSIONS: Doses delivered to patients undergoing PTB procedures are comparable to those that arise from computed tomography protocols. Radiation-induced cancer risk may be considerable for young patients undergoing PTB drainage and stent implantation procedures.

Influence of initial electron beam parameters on Monte Carlo calculated absorbed dose distributions for radiotherapy photon beams.

Our aim in the present study was to investigate the effects of initial electron beam characteristics on Monte Carlo calculated absorbed dose distribution for a linac 6 MV photon beam. Moreover, the range of values of these parameters was derived, so that the resulted differences between measured and calculated doses were less than 1%. Mean energy, radial intensity distribution and energy spread of the initial electron beam, were studied. The method is based on absorbed dose comparisons of measured and calculated depth-dose and dose-profile curves. All comparisons were performed at 10.0 cm depth, in the umbral region for dose-profile and for depths past maximum for depth-dose curves. Depth-dose and dose-profile curves were considerably affected by the mean energy of electron beam, with dose profiles to be more sensitive on that parameter. The depth-dose curves were unaffected by the radial intensity of electron beam. In contrast, dose-profile curves were affected by the radial intensity of initial electron beam for a large field size. No influence was observed in dose-profile or depth-dose curves with respect to energy spread variations of electron beam. Conclusively, simulating the radiation source of a photon beam, two of the examined parameters (mean energy and radial intensity) of the electron beam should be tuned accurately, so that the resulting absorbed doses are within acceptable precision. The suggested method of evaluating these crucial but often poorly specified parameters may be of value in the Monte Carlo simulation of linear accelerator photon beams.

Broadband ultrasound attenuation imaging: algorithm development and clinical assessment of a region growing technique.

This paper presents a computerized method for the selection of an irregular region of interest (ROI) in broadband ultrasound attenuation (BUA) images. A region growing algorithm searches an initial region in the posterior part of the calcaneus until the pixel with the lowest attenuation value is found; this is the starting seed. Then, the algorithm evaluates the values of the eight pixels neighbouring the starting seed. Pixels that have the closest value to the starting seed are accepted. This procedure is the first processing level. The procedure is repeated for the group of pixels neighbouring those accepted from the previous processing level. The algorithm ceases when the number of accepted pixels reaches a user-specified number. The clinical part of this study compares measurements of BUA at an automatic ROI implemented on a quantitative ultrasound imaging device, defined as the circular region of lowest attenuation in the posterior part of the calcaneus, and at irregular ROIs of various sizes generated by the algorithm developed in this study. The algorithm was applied to BUA images obtained from 24 post-menopausal women with hip fractures and 26 age-matched healthy female subjects. The use of the irregular ROI with a size of 2400 pixels is proposed because that region yielded better clinical results compared to irregular ROIs with different size and the circular automatic ROI.