Usability Evaluation of BCIs

摘要

Background & Purpose. Aim of the BCIs is to give paralyzed people a way to communicate. BCIs should be considered assistive technologies facilitating life activities. Therefore, an assessment process avoiding dissatisfaction and abandonment is required. In this study we aim to assess the usability of two BCIs measuring the predisposition to and the interaction with the systems. Experiment 1. Methodology. We used BCI simulations: Language Support Program (LSP) and P300 Speller (P3S). LSP allows users to choose between two alternatives (e.g. YES/NO). P3S allows users to use a flashing matrix of characters. We tested BCIs' learnability on 6 non-disabled users through the Thinking Aloud. All users used both BCIs and they were informed about their goal, although not about how they worked. Results. The Thinking Aloud task showed that, while all users on LSP easily learned how the system works, they failed on P3S. Experiment 2. Methodology. We tested the BCIs' efficiency on 30 participants (15 on LSP and 15 on P3S) through the Copy Spelling Task (CST), wherein users are asked to write on the screen through the BCI. We administered the SUS to measure usability and the SOTU scale of the MPT to measure predisposition to technology use. Results. On CST we found that P3S users are more accurate in selecting (F_((1,28))=5,18; p=.031) and recognising (F_((1,28))=7,67; p=.01) letters on the screen. Both SUS and SOTU did not show effects. Discussion. Although improvements are required, our methodology allowed us to assess BCIs. We found that LSP is more learnable than P3S. Nonetheless, the P3S allows users to be more accurate, although with less control.