Data acquisition system for a 146-channel counter of protons in particle therapy

TL;DR

This paper presents a novel 146-channel proton counter prototype using segmented Low Gain Avalanche Diodes and custom ASICs for high sensitivity and speed, suitable for real-time beam monitoring in particle therapy.

Contribution

Development of a high-speed, high-sensitivity proton counter with 146 channels using innovative ASICs and sensors for improved clinical beam monitoring.

Findings

Achieved less than 1% pileup counting inefficiency at clinical fluence rates.

Designed a compact frontend system with custom ASICs and FPGA-based digital processing.

Demonstrated real-time online counting and data storage capabilities.

Abstract

A prototype of proton counter was developed by the University and the National Institute for Nuclear Physics of Torino to be used as online fluence beam monitor in particle therapy. The single particle identification approach aims at increasing the sensitivity and readout speed with respect to the state-of-the-art gas ionization chambers. The sensitive area is 2,7 x 2,7 cm^2 to cover the clinical beam cross section characterized by a full width at half maximum of about 1 cm at the isocenter. The sensor is a thin Low Gain Avalanche Diode segmented in 146 strips with 180 micrometer pitch and with 50 micrometer active thickness, designed and produced by the Fondazione Bruno Kessler (Trento, Italy). The frontend readout to identify the single proton signal provided by each strip is based on a 24channel custom ASICs, named ABACUS, optimized to discriminate the signal pulses in a wide charge range (3-150 fC) with a maximum dead-time of 10 ns. With these specifications, at the maximum fluence rate of 10^8 p/(cm^2s) in the clinical energy range (60-230 MeV) and considering the silicon strips described above, a maximum pileup counting inefficiency less than 1 percent is achieved. A frontend board housing 6 ABACUS chips to readout the 146 strips was developed, the digital outputs being sent to 3 FPGAs (Kintex7) for the counting. A LabVIEW program implements the interface with the FPGAs, displays online the counting rate from each strip and stores the data for offline analysis.