Mouse livers machine learning identification based on hyperspectral x-ray computed tomography reconstructed x-ray absorption spectra

Zheng Fang,Siyuan Cheng,Shuo Zhong,Weifeng Hu

doi:10.1063/5.0010463

Abstract

X-ray computed tomography (X-CT) is often used to examine organs, but the reconstructed images can only be used for structural identification. Whether the organs are healthy or not requires a professional doctor to examine the reconstructed image and judge from his or her own experience. The purpose of this paper is to identify the cirrhotic mouse liver and normal mouse liver with hyperspectral x-ray CT (HXCT) and machine learning. HXCT is proposed to reconstruct the x-ray absorption spectrum (XAS) characteristics of a single pixel in the reconstructed mouse liver images. HXCT uses a cadmium telluride photon counter as the x-ray detector, which can improve the spectral resolution and separate spectral lines. Filtered back-projection and algebra reconstruction technique reconstruction algorithms are used for image and XAS reconstruction. In the machine learning model, principal component analysis is utilized to reduce the dimensionality of XAS. Besides, the neural network algorithm Artificial Neural Network (ANN) is used to train and identify the reconstructed XAS of two different kinds of livers. These two different mouse livers can be well recognized since the accuracy goes to almost 100% based on ANN. It is feasible to employ the machine learning algorithm to identify the XAS of different mouse livers.

Highlights

In order to explore the microworld, many methods, such as x-ray fluorescence (XRF), x-ray diffraction (XRD), and x-ray electron spectroscopy techniques, have been invented.1 all these methods are not applicable for biomedical testing
The average reconstructed x-ray absorption spectrum (XAS) from Filtered Back-Projection (FBP) and Algebra Reconstruction Technique (ART) are shown in Figs. 6 and 7
With the data collected by the hyperspectral x-ray CT (HXCT) system, we are able to reconstruct the image and the XAS of the interested region

Summary

Introduction

In order to explore the microworld, many methods, such as x-ray fluorescence (XRF), x-ray diffraction (XRD), and x-ray electron spectroscopy techniques, have been invented. all these methods are not applicable for biomedical testing. In order to explore the microworld, many methods, such as x-ray fluorescence (XRF), x-ray diffraction (XRD), and x-ray electron spectroscopy techniques, have been invented.. DECT becomes more and more mature, and many companies have made superior products, such as Discovery CT750 HD10 launched by General Electric in 2009 It can switch from 80 kVp to 140 kVp in 50 ms by two x-ray tubes so that 101 singleenergy images could be reconstructed in different tube voltage settings. Standard DECT may introduce interference during the material discrimination process because broadband radiation includes overlapping spectral components for using broadband radiation sources. This shortcoming leads to the low material scitation.org/journal/adv identification ability of DECT. In order to overcome the deficiency of DECT, spectral CT was proposed to maximize the utilization of spectral information

Objectives

Results

Conclusion