KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 5 of 5

Some of the metrics are blocked by your
consent settings
Evaluating Recurrent Neural Network Blind Source Separation of Event-Related Potentials Using Simulated Data
(2026-01-01)
O'Reilly, Jamie A.
Event-related potential (ERP) waveforms reflect spatiotemporal summation of potential differences resulting from the activity of current sources within the brain. Separating ERPs into underlying source waveforms and their scalp distributions is a potentially valuable for analyzing neurophysiology associated with psychophysiological events. However, ground-truth sources are unknown for real ERP data. This study aimed to evaluate recurrent neural network blind source separation (RNN-BSS) of ERP waveforms using simulated ERP data with known groundtruth source waveforms and scalp distributions. Simulated ERP waveforms were generated from seven simulated source waveforms and scalp distributions. Two simulations with positiveand negative-going source waveforms were evaluated to explore the effect of RNN source signal rectification on source representations. The source waveforms and scalp distributions extracted from applying RNN-BSS to simulated data were compared with ground-truth using Pearson's correlation coefficient. Source waveforms and scalp distributions extracted by RNN-BSS were highly correlated with their ground-truth counterparts. However, where scalp distributions of ground-truth sources were highly correlated, they were not separated perfectly by RNN-BSS. Negative-going simulations produced inverted scalp distributions. Overall, these results demonstrate efficacy of RNNBSS applied to simulated ERP waveforms with seven sources. Further evaluations are required to determine the limits of source separation when scalp distributions are highly correlated, and the influence of number of electrodes in RNN-BSS.
Some of the metrics are blocked by your
consent settings
Localized estimation of event-related neural source activity from simultaneous MEG-EEG with a recurrent neural network
(2024-12-01)
O'Reilly, Jamie A.
;
Zhu, Judy D.
;
Sowman, Paul F.
Estimating intracranial current sources underlying the electromagnetic signals observed from extracranial sensors is a perennial challenge in non-invasive neuroimaging. Established solutions to this inverse problem treat time samples independently without considering the temporal dynamics of event-related brain processes. This paper describes current source estimation from simultaneously recorded magneto- and electro-encephalography (MEEG) using a recurrent neural network (RNN) that learns sequential relationships from neural data. The RNN was trained in two phases: (1) pre-training and (2) transfer learning with L1 regularization applied to the source estimation layer. Performance of using scaled labels derived from MEEG, magnetoencephalography (MEG), or electroencephalography (EEG) were compared, as were results from volumetric source space with free dipole orientation and surface source space with fixed dipole orientation. Exact low-resolution electromagnetic tomography (eLORETA) and mixed-norm L1/L2 (MxNE) source estimation methods were also applied to these data for comparison with the RNN method. The RNN approach outperformed other methods in terms of output signal-to-noise ratio, correlation and mean-squared error metrics evaluated against reference event-related field (ERF) and event-related potential (ERP) waveforms. Using MEEG labels with fixed-orientation surface sources produced the most consistent estimates. To estimate sources of ERF and ERP waveforms, the RNN generates temporal dynamics within its internal computational units, driven by sequential structure in neural data used as training labels. It thus provides a data-driven model of computational transformations from psychophysiological events into corresponding event-related neural signals, which is unique among MEEG source reconstruction solutions.
Some of the metrics are blocked by your
consent settings
Evaluating synthetic neuroimaging data augmentation for automatic brain tumour segmentation with a deep fully-convolutional network
(2024-06-01)
Asadi, Fawad
;
Angsuwatanakul, Thanate
;
O'Reilly, Jamie A.
Gliomas observed in medical images require expert neuro-radiologist evaluation for treatment planning and monitoring, motivating development of intelligent systems capable of automating aspects of tumour evaluation. Deep learning models for automatic image segmentation rely on the amount and quality of training data. In this study we developed a neuroimaging synthesis technique to augment data for training fully-convolutional networks (U-nets) to perform automatic glioma segmentation. We used StyleGAN2-ada to simultaneously generate fluid-attenuated inversion recovery (FLAIR) magnetic resonance images and corresponding glioma segmentation masks. Synthetic data were successively added to real training data (n = 2751) in fourteen rounds of 1000 and used to train U-nets that were evaluated on held-out validation (n = 590) and test sets (n = 588). U-nets were trained with and without geometric augmentation (translation, zoom and shear), and Dice coefficients were computed to evaluate segmentation performance. We also monitored the number of training iterations before stopping, total training time, and time per iteration to evaluate computational costs associated with training each U-net. Synthetic data augmentation yielded marginal improvements in Dice coefficients (validation set +0.0409, test set +0.0355), whereas geometric augmentation improved generalization (standard deviation between training, validation and test set performances of 0.01 with, and 0.04 without geometric augmentation). Based on the modest performance gains for automatic glioma segmentation we find it hard to justify the computational expense of developing a synthetic image generation pipeline. Future work may seek to optimize the efficiency of synthetic data generation for augmentation of neuroimaging data.
Some of the metrics are blocked by your
consent settings
Localized estimation of electromagnetic sources underlying event-related fields using recurrent neural networks
(2023-08-01)
O'Reilly, Jamie A.
;
Zhu, Judy D.
;
Sowman, Paul F.
Objective. To use a recurrent neural network (RNN) to reconstruct neural activity responsible for generating noninvasively measured electromagnetic signals. Approach. Output weights of an RNN were fixed as the lead field matrix from volumetric source space computed using the boundary element method with co-registered structural magnetic resonance images and magnetoencephalography (MEG). Initially, the network was trained to minimise mean-squared-error loss between its outputs and MEG signals, causing activations in the penultimate layer to converge towards putative neural source activations. Subsequently, L1 regularisation was applied to the final hidden layer, and the model was fine-tuned, causing it to favour more focused activations. Estimated source signals were then obtained from the outputs of the last hidden layer. We developed and validated this approach with simulations before applying it to real MEG data, comparing performance with beamformers, minimum-norm estimate, and mixed-norm estimate source reconstruction methods. Main results. The proposed RNN method had higher output signal-to-noise ratios and comparable correlation and error between estimated and simulated sources. Reconstructed MEG signals were also equal or superior to the other methods regarding their similarity to ground-truth. When applied to MEG data recorded during an auditory roving oddball experiment, source signals estimated with the RNN were generally biophysically plausible and consistent with expectations from the literature. Significance. This work builds on recent developments of RNNs for modelling event-related neural responses by incorporating biophysical constraints from the forward model, thus taking a significant step towards greater biological realism and introducing the possibility of exploring how input manipulations may influence localised neural activity.
Some of the metrics are blocked by your
consent settings
Neural Generators of Intensity Mismatch Negativity Modelled with a Recurrent Neural Network : A Pilot Study on the Role of Sound Level Transitions
(2023-01-01)
Srivastava, Chandan K.
;
Gupta, Rashmi
;
O'Reilly, Jamie A.
Intensity mismatch negativity (MMN) is an electrophysiological response to auditory oddball stimulation with unexpected changes in sound pressure level. Studies using only quieter deviant stimuli leave uncertainty regarding the nature of resulting MMN, with louder deviant stimuli potentially evoking an opposite polarity amplitude shift due to intensity modulation of the auditory evoked response. To test this hypothesis, we conducted a pilot study with three subjects listening to an intensity oddball paradigm with 80 dB standards and 70 dB and 90 dB deviants. Event-related potentials and deviant-minus-standard difference waveforms were analyzed to determine whether MMN elicited by 70 dB and 90 dB reflects simple intensity modulation. We modelled the resulting neural activity with a recurrent neural network (RNN) to estimate generative signals from a distributed source space computed with a template head co-registered with standard 10-20 system electrode locations. Event-related potential waveforms suggest that intensity MMN elicited by quieter and louder auditory stimuli cannot be explained simply by intensity modulation of the auditory evoked response, given that both produced negative amplitudes from 0.1 to 0.25 s. Results from RNN-based source estimation are consistent with bilateral thalamic and right-prefrontal contributions to intensity MMN evoked by rising and falling sound level transitions.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results