More Examples and Open Questions

Goswami, Debashish; Bhowmick, Jyotishman

doi:10.1007/978-81-322-3667-2_10

Cited by 1 publication

(1 citation statement)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Advances in time series machine learning have focused recently on self-supervised, representation learning using the transformer architecture [35][36][37][38] to develop foundation models in various time series classification and forecasting tasks. Generally, these models use low frequency data, primarily < 1 Hz, limiting their fine-tuning ability to high frequency sleep datasets.…”

Section: Introductionmentioning

confidence: 99%

A foundational transformer leveraging full night, multichannel sleep study data accurately classifies sleep stages

Fox,

Jiang,

Wickramaratne

et al. 2024

Preprint

View full text Add to dashboard Cite

Study ObjectivesTo investigate whether a foundational transformer model using 8-hour, multi-channel data from polysomnograms can outperform existing artificial intelligence (AI) methods for sleep stage classification.MethodsWe utilized the Sleep Heart Health Study (SHHS) visits 1 and 2 for training and validation and the Multi-Ethnic Study of Atherosclerosis (MESA) for testing of our model. We trained a self-supervised foundational transformer (called PFTSleep) that encodes 8-hour long sleep studies at 125 Hz with 7 signals including brain, movement, cardiac, oxygen, and respiratory channels. These encodings are used as input for training of an additional model to classify sleep stages, without adjusting the weights of the foundational transformer. We compared our results to existing AI methods that did not utilize 8-hour data or the full set of signals but did report evaluation metrics for the SHHS dataset.ResultsWe trained and validated a model with 8,444 sleep studies with 7 signals including brain, movement, cardiac, oxygen, and respiratory channels and tested on an additional 2,055 studies. In total, we trained and tested 587,944 hours of sleep study signal data. Area under the precision recall curve (AUPRC) scores were 0.82, 0.40, 0.53, 0.75, and 0.82 and area under the receiving operating characteristics curve (AUROC) scores were 0.99, 0.95, 0.96, 0.98, and 0.99 for wake, N1, N2, N3, and REM, respectively, on the SHHS validation set. For MESA, the AUPRC scores were 0.56, 0.16, 0.40, 0.45, and 0.65 and AUROC scores were 0.94, 0.77, 0.87, 0.91, and 0.96, respectively. Our model was compared to the longest context window state-of-the-art model and showed increases in macro evaluation scores, notably sensitivity (3.7% increase) and multi-class REM (3.39% increase) and wake (0.97% increase) F1 scores.ConclusionsUtilizing full night, multi-channel PSG data encodings derived from a foundational transformer improve sleep stage classification over existing methods.

show abstract

Section: Introductionmentioning

confidence: 99%