PET/CT background noise and its effect on speech recognition

Iva Speck,Susan Arndt,Lars Frings,Valentin Rottmayer,Philipp T Meyer,Konstantin Wiebe,Thomas Wesarg,Antje Aschendorff,Johannes Thurow

doi:10.1038/s41598-021-01686-5

Iva Speck, Susan Arndt + Show 7 more

Open Access

https://doi.org/10.1038/s41598-021-01686-5

Copy DOI

Abstract

Positron emission tomography (PET) has been successfully used to investigate central nervous processes, including the central auditory pathway. Unlike early water-cooled PET-scanners, novel PET/CT scanners employ air cooling and include a CT system, both of which result in higher background noise levels. In the present study, we describe the background noise generated by two state-of-the-art air-cooled PET/CT scanners. We measured speech recognition in background noise: recorded PET noise and a speech-shaped noise applied in clinical routine to subjects with normal hearing. Background noise produced by air-cooled PET/CT is considerable: 75.1 dB SPL (64.5 dB(A)) for the Philips Gemini TF64 and 76.9 dB SPL (68.4 dB(A)) for the Philips Vereos PET/CT (Philips Healthcare, The Netherlands). Subjects with normal hearing exhibited better speech recognition in recorded PET background noise compared with clinically applied speech-shaped noise. Speech recognition in both background noises correlated significantly. Background noise generated by PET/CT scanners should be considered when PET is used for the investigation of the central auditory pathway. Speech in PET noise is better than in speech-shaped noise because of the minor masking effect of the background noise of the PET/CT.

Highlights

Positron emission tomography (PET) has been successfully used to investigate central nervous processes, including the central auditory pathway
We have investigated the influence of the background noise produced by air-cooled PET/CT scanners on speech recognition in young adults with normal hearing
Namely PET noise and OLnoise, temporal characteristics were assessed as the standard deviation of the root mean square (RMS) value and the mean and standard deviation of the CREST factor, the ratio between the absolute maximum amplitude and the RMS, obtained in short intervals lasting 5 ms and an overlap of 50% taken from a PET noise or OLnoise sample, each of 4 s duration, respectively

Summary

Introduction

Positron emission tomography (PET) has been successfully used to investigate central nervous processes, including the central auditory pathway. We measured speech recognition in background noise: recorded PET noise and a speech-shaped noise applied in clinical routine to subjects with normal hearing. Subjects with normal hearing exhibited better speech recognition in recorded PET background noise compared with clinically applied speech-shaped noise. Speech recognition in both background noises correlated significantly. Abbreviations CBF Cerebral blood flow CT Computed tomography FDG Fluorodeoxyglucose fMRI Functional magnetic resonance imaging PET Positron emission tomography. Since the 1980s, non-invasive functional imaging modality positron emission tomography (PET) has been used to investigate physiological (e.g., cerebral blood flow), biochemical (e.g., metabolism), and molecular (e.g., receptor binding) processes in living humans. Even with the new generation of MRI-compatible cochlear implant magnets, the magnet of an implant can cause an artifact on MRI scans[9,10,11]

Methods

Results

Discussion

Conclusion