Pileup mitigation in high rate spectroscopy using the deconvolution method

Abstract

Charge-sensitive readout chains are widely used to measure single-particle energy spectra in a variety of scientific fields. Such readout chains are used in microdosimetry, a discipline focused on measuring the energy deposition of ionizing radiation in microscopic sites. This information is particularly relevant in ion therapy, which utilizes ions to treat various types of cancer. The direct measurement of the radiation quality has been shown to exceed the capabilities of macroscopic dosimeters in clinical dosimetry and holds potential for enhancing treatment planning. Spectroscopic front-end electronics consist of a charge-sensitive preamplifier and a shaping network, which collectively generate well-defined pulses for digitization and pulse height measurement using a multichannel analyzer. Achieving a high amplitude resolution typically requires shaping times on the order of μ s. In microdosimetric measurements at clinical ion beams it is challenging to measure the amplitude with sufficient resolution and simultaneously be fast enough to manage pulse pileup at high dose rates. This work introduces an innovative pileup rejection algorithm based on the deconvolution method. By matching a custom filter to the analog front end, the timestamps of individual events can be reconstructed from severely piled-up signals with high accuracy. This approach allows the use of extended shaping times while maintaining pileup-rejection capabilities without requiring a secondary fast amplifier. Filters like this can be readily implemented in digital signal processing or potentially realized on chip. Beyond microdosimetric measurements, potential applications include high-throughput X-ray and gamma spectroscopy or scintillation counting. The method has been successfully demonstrated offline, using different signal sources.