Dataset Information

Estimation of Patient Eye-Lens Dose During Neuro-Interventional Procedures using a Dense Neural Network (DNN).

ABSTRACT: The patient's eye-lens dose changes for each projection view during fluoroscopically-guided neuro-interventional procedures. Monte-Carlo (MC) simulation can be done to estimate lens dose but MC cannot be done in real-time to give feedback to the interventionalist. Deep learning (DL) models were investigated to estimate patient-lens dose for given exposure conditions to give real-time updates. MC simulations were done using a Zubal computational phantom to create a dataset of eye-lens dose values for training the DL models. Six geometric parameters (entrance-field size, LAO gantry angulation, patient x, y, z head position relative to the beam isocenter, and whether patient's right or left eye) were varied for the simulations. The dose for each combination of parameters was expressed as lens dose per entrance air kerma (mGy/Gy). Geometric parameter combinations associated with high-dose values were sampled more finely to generate more high-dose values for training purposes. Additionally, dose at intermediate parameter values was calculated by MC in order to validate the interpolation capabilities of DL. Data was split into training, validation and testing sets. Stacked models and median algorithms were implemented to create more robust models. Model performance was evaluated using mean absolute percentage error (MAPE). The goal for this DL model is that it be implemented into the Dose Tracking System (DTS) developed by our group. This would allow the DTS to infer the patient's eye-lens dose for real-time feedback and eliminate the need for a large database of pre-calculated values with interpolation capabilities.

SUBMITTER: Collins J

PROVIDER: S-EPMC8323862 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:PURPOSE:We retrospectively assessed the incidence of cataracts in patients with retinoblastoma (Rb) treated with either lens-sparing radiation therapy (LSRT) or whole-eye radiation therapy (WERT). A secondary aim of this study was to model the dose-response risk of cataract. METHODS AND MATERIALS:We reviewed 65 patients with Rb treated with radiation therapy (RT) at Children's Hospital, Los Angeles from 1997 to 2015. Eyes that were enucleated before RT or lacked follow-up eye examinations were excluded. All patients underwent computed tomography simulation, and mean lens dose data were collected. Follow-up ophthalmologic examinations and intraocular lens implant history were reviewed for cataracts. The primary event-free survival (EFS) outcome was cataract development. Eyes without cataracts were censored on the last date of eye examination or post-RT enucleation, if applicable. Kaplan-Meier estimates were used to compare EFS outcomes, and dose response was projected with Cox regression and logistic regression models. RESULTS:Sixty-one patients (94 eyes) were analyzed with a median follow-up of 51.8 months. For eyes treated with WERT, cataracts developed in 71.7% versus 35.3% for LSRT. Median EFS for WERT and LSRT were 20.8 and 67.9 months, respectively. Compared with WERT, a significant EFS benefit was demonstrated for LSRT (P < .001). Mean lens dose had a significant effect on cataracts in both Cox regression and logistic regression models (P < .01). The mean lens dose of 7 Gy was projected to have a 5-year cataract incidence of 20% and 25% with the logistic and Cox regression models, respectively. CONCLUSIONS:We report the first clinical data demonstrating significantly improved EFS in patients with Rb treated with LSRT. Through lens dose-response modeling, we validate a mean lens dose threshold of 7 Gy to keep cataract risk below 25%. Although RT is used less often for Rb owing to advances in chemotherapy delivery options, these findings are relevant for refining lens dose constraints, particularly in children who have received radiation dose near the orbit.

Dataset Information

Estimation of Patient Eye-Lens Dose During Neuro-Interventional Procedures using a Dense Neural Network (DNN).

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets