Ole Bialas
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ole Bialas has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Edmund C. Lalor
Affiliations Department of Biomedical Engineering, University of Rochester, Rochester, New York, United States of America, Del Monte Institute for Neuroscience , University of Rochester, Rochester, New York, United States of America, Center for Visual Science, University of Rochester, Rochester, New York, United States of America
2 shared publications - Edmund C. Lalor
University of Rochester
1 shared publication - Jin Dou
University of Rochester
1 shared publication
Latest Publications
- Open Access
- 23 May 2025
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 20 (5) , e0323276
Appropriate data segmentation improves speech encoding models: Analysis and simulation of electrophysiological recordings
Background: In recent decades, studies modeling the neural processing of continuous, naturalistic, speech provided new insights into how speech and language are represented in the brain. However, the linear encoder models commonly used in such studies assume that the underlying data are stationary, varying to a fixed degree around a constant mean. Long, continuous, neural recordings may violate this assumption leading to impaired model performance. We aimed to examine the effect of non-stationary trends in continuous neural recordings on the performance of linear speech encoding models. Methods: We used temporal response functions (TRFs) to predict continuous neural responses to speech while splitting the data into segments of varying length, prior to model fitting. Our Hypothesis was that if the data were non-stationary, segmentation should improve model performance by making individual segments approximately stationary. We simulated and predicted stationary and non-stationary recordings to test our hypothesis under a known ground truth and predicted the brain activity of participants who listened to a narrated story, to test our hypothesis on actual neural recordings. Results: Simulations showed that, for stationary data, increasing segmentation steadily decreased model performance. For non-stationary data however, segmentation initially improved model performance. Modeling of neural recordings yielded similar results: segments of intermediate length (5–15 s) led to improved model performance compared to very short (1–2 s) and very long (30–120 s) segments. Conclusions: We showed that data segmentation improves the performance of encoding models for both simulated and real neural data and that this can be explained by the fact that shorter segments approximate stationarity more closely. Thus, the common practice of applying encoding models to long continuous segments of data is suboptimal and recordings should be segmented prior to modeling.
- 23 May 2025
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 20 (5) , e0323276
Open AccessAppropriate data segmentation improves speech encoding models: Analysis and simulation of electrophysiological recordings
Background: In recent decades, studies modeling the neural processing of continuous, naturalistic, speech provided new insights into how speech and language are represented in the brain. However, the linear encoder models commonly used in such studies assume that the underlying data are stationary, varying to a fixed degree around a constant mean. Long, continuous, neural recordings may violate this assumption leading to impaired model performance. We aimed to examine the effect of non-stationary trends in continuous neural recordings on the performance of linear speech encoding models. Methods: We used temporal response functions (TRFs) to predict continuous neural responses to speech while splitting the data into segments of varying length, prior to model fitting. Our Hypothesis was that if the data were non-stationary, segmentation should improve model performance by making individual segments approximately stationary. We simulated and predicted stationary and non-stationary recordings to test our hypothesis under a known ground truth and predicted the brain activity of participants who listened to a narrated story, to test our hypothesis on actual neural recordings. Results: Simulations showed that, for stationary data, increasing segmentation steadily decreased model performance. For non-stationary data however, segmentation initially improved model performance. Modeling of neural recordings yielded similar results: segments of intermediate length (5–15 s) led to improved model performance compared to very short (1–2 s) and very long (30–120 s) segments. Conclusions: We showed that data segmentation improves the performance of encoding models for both simulated and real neural data and that this can be explained by the fact that shorter segments approximate stationarity more closely. Thus, the common practice of applying encoding models to long continuous segments of data is suboptimal and recordings should be segmented prior to modeling.
- Open Access
Appropriate data segmentation improves speech encoding models
In recent decades, research on the neural processing of speech and language increasingly investigated ongoing responses to continuously presented naturalistic speech, allowing researchers to ask interesting questions about different representations of speech and their relationships. This requires statistical models that can dissect different sources of variance occurring in the processing of naturalistic speech. One commonly used family of models are temporal response functions (TRFs) which can predict neural responses to speech as a weighted combination of different features and points in time. TRFs model the brain as a linear time-invariant (LTI) system whose responses can be characterized by constant transfer functions. This implicitly assumes that the underlying signals are stationary, varying to a fixed degree around a constant mean. However, continuous neural recordings commonly violate this assumption. Here, we use simulations and EEG recordings to investigate how non-stationarities affect TRF models for continuous speech processing. Our results suggest that non-stationarities may impair the performance of TRF models, but that this can be partially remedied by dividing the data into shorter segments that approximate stationarity.
Open AccessAppropriate data segmentation improves speech encoding models
In recent decades, research on the neural processing of speech and language increasingly investigated ongoing responses to continuously presented naturalistic speech, allowing researchers to ask interesting questions about different representations of speech and their relationships. This requires statistical models that can dissect different sources of variance occurring in the processing of naturalistic speech. One commonly used family of models are temporal response functions (TRFs) which can predict neural responses to speech as a weighted combination of different features and points in time. TRFs model the brain as a linear time-invariant (LTI) system whose responses can be characterized by constant transfer functions. This implicitly assumes that the underlying signals are stationary, varying to a fixed degree around a constant mean. However, continuous neural recordings commonly violate this assumption. Here, we use simulations and EEG recordings to investigate how non-stationarities affect TRF models for continuous speech processing. Our results suggest that non-stationarities may impair the performance of TRF models, but that this can be partially remedied by dividing the data into shorter segments that approximate stationarity.
- Top Cited PapersOpen Access
- 12 September 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (89) , 5657
mTRFpy: A Python package for temporal response function analysis
Traditionally, studies on the neural processing of speech involved the repetitive display of isolated tokens (e.g., phonemes, words, sentences) where the properties of interest were carefully controlled. Recently, researchers have increasingly focused on investigating brain responses to more naturalistic speech like audiobooks (Hamilton & Huth, 2020). However, this approach demands statistical tools to account for the different sources of variance that naturally occur in speech. Among the most popular tools to model neural responses to naturalistic speech are multivariate temporal response functions (mTRFs). One of the most commonly used packages for computing mTRFs with regularized regression is the mTRF-toolbox (Crosse et al., 2016). However, this toolbox is implemented in the proprietary MATLAB language, restricting accessibility for parts of the scientific community. To overcome this constraint, we present mTRFpy, a Python package which replicates and advances the functionality of the original mTRF-toolbox.
- 12 September 2023
- Published by The Open Journal in Journal of Open Source Software
- Vol. 8 (89) , 5657
Top Cited PapersOpen AccessmTRFpy: A Python package for temporal response function analysis
Traditionally, studies on the neural processing of speech involved the repetitive display of isolated tokens (e.g., phonemes, words, sentences) where the properties of interest were carefully controlled. Recently, researchers have increasingly focused on investigating brain responses to more naturalistic speech like audiobooks (Hamilton & Huth, 2020). However, this approach demands statistical tools to account for the different sources of variance that naturally occur in speech. Among the most popular tools to model neural responses to naturalistic speech are multivariate temporal response functions (mTRFs). One of the most commonly used packages for computing mTRFs with regularized regression is the mTRF-toolbox (Crosse et al., 2016). However, this toolbox is implemented in the proprietary MATLAB language, restricting accessibility for parts of the scientific community. To overcome this constraint, we present mTRFpy, a Python package which replicates and advances the functionality of the original mTRF-toolbox.