Universal automated classification of the acoustic startle reflex using machine learning

Timothy J. Fawcett; Ryan J. Longenecker; Dimitri L. Brunelle; Joel I. Berger; Mark N. Wallace; Alex V. Galazyuk; Merri J. Rosen; Richard J. Salvi; Joseph P. Walton

doi:10.1016/j.heares.2022.108667

Back

Universal automated classification of the acoustic startle reflex using machine learning

Journal article

Open access

Peer reviewed

Universal automated classification of the acoustic startle reflex using machine learning

Timothy J. Fawcett, Ryan J. Longenecker, Dimitri L. Brunelle, Joel I. Berger, Mark N. Wallace, Alex V. Galazyuk, Merri J. Rosen, Richard J. Salvi and Joseph P. Walton

Hearing research, Vol.428, pp.108667-108667

02/01/2023

DOI: 10.1016/j.heares.2022.108667

PMCID: PMC10734095

PMID: 36566642

Files and links (1)

url

https://doi.org/10.1016/j.heares.2022.108667View

Published (Version of record) Open Access

Abstract

•Acoustic startle reflex waveforms must be classified into true startle / non-startle to ensure high quality data.•Machine learning method generalized across multiple species and startle modalities.•Features derived from the time/period of continuous wavelet transform are key to success.•Classify startle waveforms for nearly any startle measurement on nearly any species. The startle reflex (SR), a robust, motor response elicited by an intense auditory, visual, or somatosensory stimulus has been widely used as a tool to assess psychophysiology in humans and animals for almost a century in diverse fields such as schizophrenia, bipolar disorder, hearing loss, and tinnitus. Previously, SR waveforms have been ignored, or assessed with basic statistical techniques and/or simple template matching paradigms. This has led to considerable variability in SR studies from different laboratories, and species. In an effort to standardize SR assessment methods, we developed a machine learning algorithm and workflow to automatically classify SR waveforms in virtually any animal model including mice, rats, guinea pigs, and gerbils obtained with various paradigms and modalities from several laboratories. The universal features common to SR waveforms of various species and paradigms are examined and discussed in the context of each animal model. The procedure describes common results using the SR across species and how to fully implement the open-source R implementation. Since SR is widely used to investigate toxicological or pharmaceutical efficacy, a detailed and universal SR waveform classification protocol should be developed to aid in standardizing SR assessment procedures across different laboratories and species. This machine learning-based method will improve data reliability and translatability between labs that use the startle reflex paradigm.

Acoustic startle response

Ensemble models

Machine learning

Pre-pulse inhibition

Waveform classification

Details

Title: Subtitle: Universal automated classification of the acoustic startle reflex using machine learning
Creators: Timothy J. Fawcett - University of South Florida
Ryan J. Longenecker - ITER
Dimitri L. Brunelle - University of South Florida
Joel I. Berger - University of Iowa
Mark N. Wallace - University of Nottingham
Alex V. Galazyuk - Northeast Ohio Medical University
Merri J. Rosen - Northeast Ohio Medical University
Richard J. Salvi - University at Buffalo, State University of New York
Joseph P. Walton - University of South Florida
Resource Type: Journal article
Publication Details: Hearing research, Vol.428, pp.108667-108667
DOI: 10.1016/j.heares.2022.108667
PMID: 36566642
PMCID: PMC10734095
NLM abbreviation: Hear Res
ISSN: 0378-5955
eISSN: 1878-5891
Publisher: Elsevier B.V
Grant note: DOI: 10.13039/100000002, name: National Institutes of Health, award: AG009524; DOI: 10.13039/100000055, name: National Institute on Deafness and Other Communication Disorders, award: R01 DC000937; DOI: 10.13039/100019324, name: Virginia Marine Resources Commission, award: R01 DC013314, TRIH 2018, U135097126
Language: English
Date published: 02/01/2023
Academic Unit: Neurosurgery
Record Identifier: 9984618519802771

Metrics

10 Record Views

6 Times Cited - Web of Science