braindecode.models.SPARCNet#

class braindecode.models.SPARCNet(n_chans=None, n_times=None, n_outputs=None, block_layers=4, growth_rate=16, bottleneck_size=16, drop_prob=0.5, conv_bias=True, batch_norm=True, activation=<class 'torch.nn.modules.activation.ELU'>, kernel_size_conv0=7, kernel_size_conv1=1, kernel_size_conv2=3, kernel_size_pool=3, stride_pool=2, stride_conv0=2, stride_conv1=1, stride_conv2=1, padding_pool=1, padding_conv0=3, padding_conv2=1, kernel_size_trans=2, stride_trans=2, chs_info=None, input_window_seconds=None, sfreq=None)[source]#

Seizures, Periodic and Rhythmic pattern Continuum Neural Network (SPaRCNet) from Jing et al (2023) [jing2023].

Convolution

This is a temporal CNN model for biosignal classification based on the DenseNet architecture.

The model is based on the unofficial implementation [Code2023].

Added in version 0.9.

Parameters:

  • n_chans (int) – Number of EEG channels.

  • n_times (int) – Number of time samples of the input window.

  • n_outputs (int) – Number of outputs of the model. This is the number of classes in the case of classification.

  • block_layers (int) – Number of layers per dense block. Default is 4.

  • growth_rate (int) – Growth rate of the DenseNet. Default is 16.

  • bottleneck_size (int) – The description is missing.

  • drop_prob (float) – Dropout rate. Default is 0.5.

  • conv_bias (bool) – Whether to use bias in convolutional layers. Default is True.

  • batch_norm (bool) – Whether to use batch normalization. Default is True.

  • activation (type[Module]) – Activation function class to apply. Should be a PyTorch activation module class like nn.ReLU or nn.ELU. Default is nn.ELU.

  • kernel_size_conv0 (int) – The description is missing.

  • kernel_size_conv1 (int) – The description is missing.

  • kernel_size_conv2 (int) – The description is missing.

  • kernel_size_pool (int) – The description is missing.

  • stride_pool (int) – The description is missing.

  • stride_conv0 (int) – The description is missing.

  • stride_conv1 (int) – The description is missing.

  • stride_conv2 (int) – The description is missing.

  • padding_pool (int) – The description is missing.

  • padding_conv0 (int) – The description is missing.

  • padding_conv2 (int) – The description is missing.

  • kernel_size_trans (int) – The description is missing.

  • stride_trans (int) – The description is missing.

  • chs_info (list of dict) – Information about each individual EEG channel. This should be filled with info["chs"]. Refer to mne.Info for more details.

  • input_window_seconds (float) – Length of the input window in seconds.

  • sfreq (float) – Sampling frequency of the EEG recordings.

Raises:

ValueError – If some input signal-related parameters are not specified: and can not be inferred.

Notes

This implementation is not guaranteed to be correct, has not been checked by original authors.

References

[jing2023]

Jing, J., Ge, W., Hong, S., Fernandes, M. B., Lin, Z., Yang, C., … & Westover, M. B. (2023). Development of expert-level classification of seizures and rhythmic and periodic patterns during eeg interpretation. Neurology, 100(17), e1750-e1762.

[Code2023]

Yang, C., Westover, M.B. and Sun, J., 2023. BIOT Biosignal Transformer for Cross-data Learning in the Wild. GitHub https://github.com/ycq091044/BIOT (accessed 2024-02-13)

Hugging Face Hub integration

When the optional huggingface_hub package is installed, all models automatically gain the ability to be pushed to and loaded from the Hugging Face Hub. Install with:

pip install braindecode[hub]

Pushing a model to the Hub:

from braindecode.models import SPARCNet

# Train your model
model = SPARCNet(n_chans=22, n_outputs=4, n_times=1000)
# ... training code ...

# Push to the Hub
model.push_to_hub(
    repo_id="username/my-sparcnet-model",
    commit_message="Initial model upload",
)

Loading a model from the Hub:

from braindecode.models import SPARCNet

# Load pretrained model
model = SPARCNet.from_pretrained("username/my-sparcnet-model")

# Load with a different number of outputs (head is rebuilt automatically)
model = SPARCNet.from_pretrained("username/my-sparcnet-model", n_outputs=4)

Extracting features and replacing the head:

import torch

x = torch.randn(1, model.n_chans, model.n_times)
# Extract encoder features (consistent dict across all models)
out = model(x, return_features=True)
features = out["features"]

# Replace the classification head
model.reset_head(n_outputs=10)

Saving and restoring full configuration:

import json

config = model.get_config()            # all __init__ params
with open("config.json", "w") as f:
    json.dump(config, f)

model2 = SPARCNet.from_config(config)    # reconstruct (no weights)

All model parameters (both EEG-specific and model-specific such as dropout rates, activation functions, number of filters) are automatically saved to the Hub and restored when loading.

See Loading and Adapting Pretrained Foundation Models for a complete tutorial.

Methods

forward(X)[source]#

Forward pass of the model.

Parameters:

X (Tensor) – The input tensor of the model with shape (batch_size, n_channels, n_times)

Returns:

The output tensor of the model with shape (batch_size, n_outputs)

Return type:

torch.Tensor