Optimal scintillation time estimation through analog filters

Optimal scintillation time estimation through analog filters

Scintillator

ToF-PET

CFD

Time measurement

Time estimation

Maximum likelihood

Bias correction

Adapted filter

Analog filter

620

Ingeniería

Tesis (Master of Science in Engineering)--Pontificia Universidad Católica de Chile, 2023

Scintillator crystals are a widespread solution for detecting high-energy radiation. These crystals spread the energy of the particle over time, which fundamentally limits the time resolution that can be obtained. This is relevant for applications requiring good time resolution, such as ToF-PET, which is currently aiming at resolutions in the order tens of picoseconds, and low energy measurements, where the limit is further degraded. The most common approach to measure scintillation time is leading-edge discrimination (LED), which performs better than other time measurement techniques because of the high variability of the measured waveforms, while more complex approaches using digital signal processing only present small improvements. In this work, an alternative detection technique is presented, by showing that a multi-variate MLE, which considers the shape and amplitude of the distribution, can be performed by feeding the output current of a SiPM to a filter, whose output is proportional to the log-probability of the scintillation time under a given distribution. It is shown that said filter can be implemented as an analog, passive pulse-shaper on a typical SiPM readout front-end, and the optimal estimation is made by evaluating a combined estimator of the time and amplitude of the maximum of the response. It is shown through a custom simulation that the proposed method gives an estimate independent of both energy and waveform variations, but has a slightly reduced resolution with respect to LED, so its convenience should be tested for each particular application.

Amoedo Nores, Gonzalo

2023-06-23T14:21:18Z

2023-06-23T14:21:18Z

2023

Guesalaga Meissner, Andrés

Abusleme Hoffman, Ángel Christian

Pontificia Universidad Católica de Chile. Escuela de Ingeniería

acceso abierto

xi, 75 páginas

application/pdf

en

tesis de maestría

https://repositorio.uc.cl/handle/11534/73523