Music genre neuroimaging dataset

Abstract

This dataset includes functional magnetic resonance imaging (fMRI) data collected while five subjects extensively listened to 540 music pieces from 10 music genres over the course of 3 days. Behavioral data are also available. Data are separated into training and test samples to facilitate the application of machine learning algorithms. Test stimuli were repeated four times and can be used to evaluate the signal to noise ratio of brain activity. Using this dataset, both neuroimaging and machine learning researchers can test multiple algorithms regarding the prediction performance of brain activity induced by various music stimuli. Although two previous studies have used this dataset, there remains much room to apply different acoustic models. This dataset can contribute to integration of the fields of auditory neuroscience and machine learning.

Exploring an EM-algorithm for banded regression in computational neuroscience

“…In this section, we investigate how the inclusion of stimulus-irrelevant predictors - here defined as simulated noise predictors that are unrelated to the task or stimuli - can impact results in an fMRI encoding analysis and how the EM-banded estimator may help suppress their contribution. The example uses a publicly available BOLD fMRI dataset [50] described in [37, 51]. The dataset contains BOLD fMRI data acquired from five participants listening to excerpts from music in 12 training runs and 6 test runs.…”

“…The Python code relies on numpy [78], scipy [79] and numba [80]. EEG data and BOLD fMRI data used in this study have been presented previously [5, 37, 50] and is available on https://doi.org/10.5061/dryad.070jc and https://openneuro.org/datasets/ds003720, respectively.…”

“…The example uses a publicly available BOLD fMRI dataset [50] described in [37,51]. The dataset contains BOLD fMRI data acquired from five participants listening to excerpts from music in 12 training runs and 6 test runs.…”

“…In this paper, we describe the modeling framework and provide simulations to illustrate the properties, limitations and potential use-cases of the described algorithm. Additionally, we illustrate properties of the model by using it to fit stimulus-response models on two publicly available datasets (Broderick et al, 2018; Nakai et al, 2022) and compare these model fits with the widely used Ridge estimator (Hoerl and Kennard, 1970). In the first dataset, we consider encoding models that use a set of audio features to predict BOLD fMRI responses to music (Nakai et al, 2022).…”

“…Additionally, we illustrate properties of the model by using it to fit stimulus-response models on two publicly available datasets (Broderick et al, 2018; Nakai et al, 2022) and compare these model fits with the widely used Ridge estimator (Hoerl and Kennard, 1970). In the first dataset, we consider encoding models that use a set of audio features to predict BOLD fMRI responses to music (Nakai et al, 2022). In the second, we fit decoding models that use features of multi-channel time-lagged EEG data to predict an audio stimulus feature during a speech listening task (Broderick et al, 2018).…”