In recent years, thanks to advances in sensor technology, imaging and data analysis, the use of AI-powered medical devices has seen an exponential increase, with mas-sive adoption by hospitals and specialized laboratories. These technologies also offer promising solutions for augmentative and alternative communication (AAC) devices needed for people with communication disorders, speech disabilities, and autism. This paper focuses on developing a virtual keyboard for patients with limited communication skills, specifically improving the AAC BCI (BrainControl Interface) device, which uses an EEG helmet to sense the users' brain activity to help them interact with the external world. By leveraging machine learning techniques, particularly using recurrent networks, we are able to accurately predict user intent and improve the typing experience. The experimental results show an average interaction time reduced by 45% compared to the previous scanning method, extremely significant for locked-in patients.
Prete, A.L., Landi, D., Andreini, P., Bianchini, M. (2023). Deep Learning Techniques for Text Generation to Support Augmentative and Alternative Communication. In 2023 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE) (pp.206-211). New York : IEEE [10.1109/MetroXRAINE58569.2023.10405778].
Deep Learning Techniques for Text Generation to Support Augmentative and Alternative Communication
Alessia Lucia Prete;Paolo Andreini;Monica Bianchini
2023-01-01
Abstract
In recent years, thanks to advances in sensor technology, imaging and data analysis, the use of AI-powered medical devices has seen an exponential increase, with mas-sive adoption by hospitals and specialized laboratories. These technologies also offer promising solutions for augmentative and alternative communication (AAC) devices needed for people with communication disorders, speech disabilities, and autism. This paper focuses on developing a virtual keyboard for patients with limited communication skills, specifically improving the AAC BCI (BrainControl Interface) device, which uses an EEG helmet to sense the users' brain activity to help them interact with the external world. By leveraging machine learning techniques, particularly using recurrent networks, we are able to accurately predict user intent and improve the typing experience. The experimental results show an average interaction time reduced by 45% compared to the previous scanning method, extremely significant for locked-in patients.File | Dimensione | Formato | |
---|---|---|---|
Deep_Learning_Techniques_for_Text_Generation_to_Support_Augmentative_and_Alternative_Communication.pdf
non disponibili
Tipologia:
PDF editoriale
Licenza:
NON PUBBLICO - Accesso privato/ristretto
Dimensione
778.06 kB
Formato
Adobe PDF
|
778.06 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11365/1254919