Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Enformatik Enstitüsü, Sağlık Bilişimi Anabilim Dalı, Türkiye
Tezin Onay Tarihi: 2018
Öğrenci: UĞUR ACAR
Danışman: TOLGA ESAT ÖZKURT
Özet:The present thesis aims to contribute to the understanding of the effects of stimulus types and also cognitive aspects on Brain Computer Interfaces(BCIs) to promote the integration between BCIs and applications. In the model of the present thesis, three types of stimuli are presented. The first type are alphabet letters, the second type are symbols for basic needs, and the last type are words for creation of sentences. Technically the present model has two parts, a training phase and a testing (prediction) phase. In the training phase, the user concentrates on one of the stimuli among the entire set of alternative stimuli and the application detects a P300 ERP for that target stimulus which captures the selection of the subject. The training phase is used for understanding the neural patterns of each user and thus, it generates a user-specific training model. After the completion of the training phase, the subject is able to use the application for communication purposes. In the subsequent test phase, the subject carries out two sets of tasks. First, subjects use the three sets of items (letters, icons, and words), and create pre-specified target items, e.g., words (from letters), icon and word sequences. In case of incongruence of the intended and the produced item (by the application), a specific neural ERP signal, N400, is expected, which is statistically analyzed. Second, the cognitive aspects of subjects’ performance are investigated with respect to their motivation and mood. Thus, the effect of subjects’ current motivation and mood is measured before the experiment and after the trials of each modality and correlated with the success rate of the P300 speller performance. In sum, the proposed thesis aims at combining technical and cognitive aspects of the usage of BCI devices for communication. The analysis on predicting the success rate for the different stimuli types revealed that the P300 speller is performing better for characters and words than icons. Furthermore, The ERP data analysis within the time window of one second after onset of the predicted stimulus showed that there is no statistically significant N400 effect. Furthermore, the obtained results revealed that the current motivation of the participants might have a prominent role during the trials. The results are discussed with respect to the literature.