Application of Transformers for Sleep Stage Classification

• David Vázquez-Lema ^[1] ; Diego Alvarez-Estevez ; Eduardo Mosqueira-Rey ^[1]
  1. [1] Universidade da Coruña

     Universidade da Coruña

     A Coruña, España

     
• Localización: Proceedings XoveTIC 2025: Impulsando el talento científico / coord. por Manuel Lagos Rodríguez, Hilda Romero-Velo, Álvaro Leitao , Javier Pereira-Loureiro , Manuel Penedo , 2026, ISBN 978-84-9749-925-5, págs. 99-106
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Transformer architectures have revolutionized the field of artificial intelligence. Thiswork investigates the application of Transformers to the sleep stage classification problem. De-spite promising results in recent studies, the clinical application of these methods remains lim-ited. In this paper, we propose a novel model that replaces the LSTM component of a state-of-the-art CNN+LSTM architecture with a Transformer encoder, comparing its performance againstour baseline and several state-of-the-art models. The results showed faster convergence, reducedcomplexity, and enhanced performance. Leveraging the Transformer architecture, we proposeand investigate an interpretability method based on attention mechanisms. Finally, we evaluatethe inter-database generalization performance of our model.