Why the study?
This study aims to decode hemodynamic responses evoked by multiple sound-categories using functional near-infrared spectroscopy.
Can LSTM networks decode hemodynamic responses evoked by multiple sound categories using fNIRS in healthy subjects?
Population
18 healthy subjects
Comparison
10-s blocks of six sound-categories (English, non-English, annoying, nature, music, and gunshot)
Design
fNIRS decoding study using Long short-term memory networks
Authors
Loading...
fNIRS-LSTM decoding of sound-evoked HRs is feasible; leaves open validation for auditory neuroscience applications.
Can LSTM networks decode hemodynamic responses evoked by multiple sound categories using fNIRS in healthy subjects?
LSTM networks can decode fNIRS data of multiple sound categories subject-wise without feature selection, achieving slightly above chance level accuracy.
Yoo et al. (2021) studied this question.