Abstract:PURPOSE The purpose of this study is to perform an extensive investigation of an approximately 2.5 mm long 90Sr/90Y source designed for treating wet age-related macular degeneration. METHODS As part of this investigation, a NIST-traceable absorbed dose to water calibration technique was established, and a source deployment verification test was developed. The influence of treatment cannula construction tolerance on the measurements as well as the dose delivered to the patient was investigated using the Monte Carlo code MCNP5. Variation between production cannulae was quantified experimentally using a well-type ionization chamber, and additional measurements along with Monte Carlo calculations of the collimating insert used for source deployment verification were performed to validate the model. RESULTS Maximum variation in the integrated target dose was seen when the source was shifted laterally within the treatment cannula. For the well chamber measurements, the observed standard deviation in ionization current for a single source placed in different reference cannulae was +/-0.3%, with a maximum observed range of less than +/-0.5%. Clinical cannulae in the collimating insert showed an average of 17.8% +/-0.4% of the reference signal when sources were fully deployed compared to 18.5% predicted by Monte Carlo calculations. This discrepancy has been attributed primarily to construction of the collimator since the collimation gap was observed to be approximately 0.025-0.075 mm smaller than specified. Construction tolerance of the well chamber insert as well as position tolerance of the cannula tip were both investigated, and their influence on the predicted signal was quantified. Additional measurements along with Monte Carlo based calculations of the collimating insert with polyethylene spacers added to the setup were performed to validate the Monte Carlo model. The shimmed Monte Carlo and measured data agree to within 1%, which is a magnitude difference of approximately 0.1% of the reference signal. CONCLUSIONS This investigation confirms that the signal for an acceptably deployed source in the collimating insert is between 17.5% and 21.5% of the reference signal, as calculated using Monte Carlo models. Clinical cannulae for which the source deployment verification measurement falls outside the acceptable range should not be used to treat patients.

TANet: Adversarial Network via Tokens Transformer for Universal Domain Adaptation.

Trust-aware Conditional Adversarial Domain Adaptation with Feature Norm Alignment.

Dual Contrastive Universal Adaptation Network.

Transferable adversarial masked self-distillation for unsupervised domain adaptation

Exploiting Both Domain-specific and Invariant Knowledge via a Win-win Transformer for Unsupervised Domain Adaptation

Towards Unsupervised Domain Adaptation via Domain-Transformer

Reusing the Task-specific Classifier as a Discriminator: Discriminator-free Adversarial Domain Adaptation

Making the Best of Both Worlds: A Domain-Oriented Transformer for Unsupervised Domain Adaptation

Transferable Adversarial Training: A General Approach to Adapting Deep Classifiers

Evidential Neighborhood Contrastive Learning for Universal Domain Adaptation

Transductive Adversarial Networks (TAN)

A Two-Stage Adaptation Network (TSAN) for Remote Sensing Scene Classification in Single-Source-Mixed-Multiple-Target Domain Adaptation ( S 2 M 2 T DA) Scenarios

CDTrans: Cross-domain Transformer for Unsupervised Domain Adaptation

Stochastic Adversarial Networks for Multi-Domain Text Classification

Adversarial open set domain adaptation via progressive selection of transferable target samples

Investigation of a 90Sr/90Y source for intra-ocular treatment of wet age-related macular degeneration.

A Two-Stage Adaptation Network (TSAN) for Remote Sensing Scene Classification in Single-Source-Mixed-Multiple-Target Domain Adaptation (S<SUP>2</SUP>M<SUP>2</SUP>T DA) Scenarios

Re-weighted Adversarial Adaptation Network for Unsupervised Domain Adaptation

Unveiling Class-Labeling Structure for Universal Domain Adaptation

Deep Asymmetric Transfer Network For Unbalanced Domain Adaptation

Universal multi-Source domain adaptation for image classification