A JavaScript-based computational method for predicting decay activity of 131I: development, benchmark- and...

Area of Science:

• Nuclear physics and medical imaging.
• Computational methods in radiation science.

Background:

• Accurate prediction of iodine-131 (131I) decay and cumulative activity is crucial for dosimetry and research.
• Existing methods may require specialized software or extensive calibration.
• Development of accessible, open-access tools can enhance research reproducibility.

Purpose of the Study:

• To develop and validate an open-access JavaScript tool for predicting 131I decay and cumulative activity.
• To provide a reliable, client-side computational resource for researchers.
• To assess the accuracy and reliability of the developed prediction models.

Main Methods:

• Implementation of a client-side JavaScript engine simulating 131I decay using mono-exponential and piecewise multi-segmented models.
• Utilization of built-in JavaScript functions for logarithmic and exponential calculations.
• Validation of the tool against ion-chamber measurements using a thyroid phantom with a pre-calibrated 131I capsule.

Main Results:

• The JS tool demonstrated high accuracy, with mean absolute errors of 0.010 μSv/h (mono-exponential) and 0.008 μSv/h (multi-segmented).
• Excellent inter-observer agreement was confirmed via the intraclass correlation coefficient (ICC range: 0.998-0.999).
• Low intra-observer variability was observed, with a mean difference of 0.3% (SD = 0.2%).

Conclusions:

• The developed JavaScript tool offers a validated, open-access method for predicting 131I decay and activity.
• This tool can improve the confidence in isolating biological variability from instrument error in future studies.
• The findings support the robustness of patient-level interpretation and dose-prediction models in nuclear medicine and research.