B-spline parameterization of spatial response in a monolithic scintillation camera

TL;DR

This paper introduces a B-spline based method for efficiently parameterizing the light response functions in scintillation cameras, improving computational speed and reducing memory usage for detector response modeling.

Contribution

It presents a novel B-spline based framework for modeling light response functions, with techniques to minimize storage and computational requirements, integrated into an open source software package.

Findings

Smooth and accurate light response function approximation

Reduced computational and memory requirements

Effective integration into existing simulation software

Abstract

A framework for parameterization of the light response functions (LRFs) in a scintillation camera is presented. It is based on approximation of the measured or simulated photosensor response with weighted sums of uniform cubic B-splines or their tensor products. The LRFs represented in this way are smooth, computationally inexpensive to evaluate and require much less computer memory than non-parametric alternatives. The parameters are found in a straightforward way by the linear least squares method. Several techniques that allow to reduce the storage and processing power requirements were developed. A software library for fitting simulated and measured light response with spline functions was developed and integrated into an open source software package ANTS2 designed for simulation and data processing for Anger camera type detectors.