Centroiding algorithms and spatial resolution of photon counting detectors with cross-strip anodes

Abstract

Significant advances in readout elements of microchannel plate based sensors have led to the development of detectors with less than 10 µm spatial resolution. We have shown that cross strip (XS) anodes have spatial resolution as small as 5 µm FWHM when a simple and fast center of gravity centroiding technique is used. In this paper we investigate the variation of XS anode spatial resolution for several types of centroiding algorithms (N-finger center of gravity, Gaussian, Lorentzian, parabolic and hyperbolic cosine centroiding) and determine the optimum algorithm in terms of spatial resolution and image linearity. We found that for existing 32x32 mm 2 cross strip anodes and associated electronics, the best resolution and linearity is achieved with center of gravity centroiding with properly chosen thresholds on the data set. The images of USAF resolution test target obtained with MCP's with 6 µm pores on 7.5 µm centers resolve 71.8 line pairs per mm (Group 6 element 2, with line width of only ~7 µm). Thus the ~7 µm spatial resolution of the detector with cross strip anode is indeed limited now by the size of the MCP pores, while the resolution of XS readout is on the order of only few micrometers FWHM.

Sparse spectral techniques for emission imaging

“…This technique can achieve sub-detector element spatial resolution given the detected signal cluster spreads several detector elements. Therefore, the combination of a signal intensifier, which translates a single particle event into a many-particle shower, and a highly segmented detector provides a powerful and efficient tool for high spatial resolution imaging [10,11]. A centroiding algorithm computes the center of a signal distribution by estimating the shape of this (amplified) distribution.…”

“…Selected, often used centroiding algorithms for sparse emission data sets, are center-of-gravity (COG) centroiding, centroiding algorithms based on analytical functions (that describe the emission cloud) or "interpolated convolution". Center-of-gravity (COG) centroiding [9,10,13,[15][16][17] is a simple and popular centroiding method. The center coordinates of a signal cluster are determined by calculating a weighted average over the signal cluster.…”

Cross strip readouts for photon counting detectors with high spatial and temporal resolution

“…We also tried different centroiding algorithms (such as non-linear Gaussian fit, polynomial fit, etc.) in order to achieve the highest spatial resolution and found that the accuracy of simple centroiding is sufficient for current MCP pore dimensions (~6-10 µm) [10].…”

“…The cross strip anode [8]- [10] is a coarse period (~0.5 mm) multi-layer metal and ceramic cross strip pattern on a ceramic (alumina) substrate. On one surface of the substrate the conductors are fabricated as a set of fingers approximately 0.5mm wide.…”

A model of high resolution cross strip readout for photon and ion counting imaging detectors

“…The modeled images can be compared to experimentally obtained image of USAF resolution mask shown in Providing the accuracy of the anode geometry is adequate (see next section), the spatial resolution of the detector can be improved by application of more efficient centroiding algorithms, discussed in detail in references [6]- [8]. elements 1 and 2).…”