LQL-Equiv-web: a Python library and browser application for linear-quadratic-linear dose equivalence in radiotherapy
Abstract
Dose equivalence calculations decide how radiotherapy schedules are compared, compensated for after an interruption, and summed in reirradiation. They are also applied inconsistently: across identical clinical cases, inter-practitioner agreement reaches only 25.8 % and nearly half of surveyed centres report no usable tool. LQL-Equiv computes biologically effective dose, equivalent dose in a reference fractionation, normal-tissue complication probability, tumour control probability and radiation-induced cancer risk under the linear-quadratic-linear model, with corrections for incomplete inter-fraction repair, accelerated proliferation and treatment protraction, over any number of successive courses separated by gaps. Version 3.0 is a dependency-free Python library and a browser application that runs entirely on the user’s machine through WebAssembly : no installation, no licence fee, and no computation on a server. It was regression-tested case by case against the 2014 MATLAB reference implementation over 4438 treatment schedules, agreeing on 69,125 of 69,131 compared values, its solver was checked against the original exhaustive search across 10,554 searches, and its equations are pinned independently of both by closed-form tests. The complication and control probabilities are point estimates at a single equivalent dose and have not been validated against clinical outcomes. It is intended for research and education and is not a medical device.
Related
- MilAMin in the browser: a million unknowns a minute in TypeScript and WebAssembly
- A library for differentiable signal processing and machine learning on the sphere
- Screening and Activity Evaluation of Peptides With Anti-Ice-Nucleation Activity Using the Phage Display Library Method.
- Multi-sensor IoT data and AI-driven adaptive control model for library indoor environments
- antasena: A Python Library for Nitrox Diving Calculations
BibTeX
@article{voyant2026lqlequivweb,
title = {{LQL-Equiv-web: a Python library and browser application for linear-quadratic-linear dose equivalence in radiotherapy}},
author = {Cyril Voyant and Daniel Julian},
journal = {SoftwareX},
year = {2026},
month = sep,
doi = {10.1016/j.softx.2026.103085},
url = {https://hal.science/hal-05768419v1},
note = {CC BY},
}
RIS
TY - JOUR
TI - LQL-Equiv-web: a Python library and browser application for linear-quadratic-linear dose equivalence in radiotherapy
AU - Cyril Voyant
AU - Daniel Julian
PY - 2026
DA - 2026/09/29
JO - SoftwareX
DO - 10.1016/j.softx.2026.103085
UR - https://hal.science/hal-05768419v1
AB - Dose equivalence calculations decide how radiotherapy schedules are compared, compensated for after an interruption, and summed in reirradiation. They are also applied inconsistently: across identical clinical cases, inter-practitioner agreement reaches only 25.8 % and nearly half of surveyed centres report no usable tool. LQL-Equiv computes biologically effective dose, equivalent dose in a reference fractionation, normal-tissue complication probability, tumour control probability and radiation-induced cancer risk under the linear-quadratic-linear model, with corrections for incomplete inter-fraction repair, accelerated proliferation and treatment protraction, over any number of successive courses separated by gaps. Version 3.0 is a dependency-free Python library and a browser application that runs entirely on the user’s machine through WebAssembly : no installation, no licence fee, and no computation on a server. It was regression-tested case by case against the 2014 MATLAB reference implementation over 4438 treatment schedules, agreeing on 69,125 of 69,131 compared values, its solver was checked against the original exhaustive search across 10,554 searches, and its equations are pinned independently of both by closed-form tests. The complication and control probabilities are point estimates at a single equivalent dose and have not been validated against clinical outcomes. It is intended for research and education and is not a medical device.
SN - 2352-7110
ER -