Vo, A, Nguyen, D, Pham, T, Ha, K & Dutkiewicz, E 2018, 'Subject-Independent ERP-based Brain-Computer Interfaces', IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 26, no. 4, pp. 719-728.
Brain–computer interfaces (BCIs) are desirable for people to express their thoughts, especially those with profound disabilities in communication.
The classification of brain patterns for each different subject requires an extensively time-consuming learning stage specific to that person, in order to reach satisfactory
accuracy performance. The training session could also be
infeasible for disabled patients as they may not fully understand the training instructions. In this paper, we propose a
unified classification scheme based on ensemble classifier,
dynamic stopping, and adaptive learning. We apply
this scheme on the P300-based BCI, with the subjectindependent manner, where no learning session is required for new experimental users. According to our theoretical analysis and empirical results, the harmonized integration of these three methods can significantly boost up the average accuracy from 75.00% to 91.26%, while at the same
time reduce the average spelling time from 12.62 to 6.78 iterations, approximately to two-fold faster. The experiments were conducted on a large public dataset which had been used in other related studies. Direct comparisons between our work with the others’ are also reported in details.
Vo, A & Eryk Dutkiewicz 2018, 'Optimal Length-Constrained Segmentation and Subject-Adaptive Learning for Real-time Arrhythmia Detection', IEEE/ACM International Conference on Connected Health: Applications, Systems and Engineering Technologies, Washington, DC, USA.
Vo, A, Diep Nguyen, Kha Ha & Eryk Dutkiewicz 2017, 'Real-Time Analysis on Ensemble SVM Scores to Reduce P300-Speller Intensification Time', IEEE Engineering in Medicine and Biology Society (EMBC’17), Korea, pp. 4383-4386.View/Download from: Publisher's site
Vo, A, Diep Nguyen, Kha Ha & Eryk Dutkiewicz 2017, 'Subject-Independent P300 BCI using Ensemble Classifier, Dynamic Stopping and Adaptive Learning', GLOBECOM 2017 - 2017 IEEE Global Communications Conference, IEEE Global Communications Conference, IEEE, Singapore.View/Download from: Publisher's site
Brain-computer interfaces (BCIs) are used to assist people, especially those with verbal or physical disabilities, communicate with the computer to indicate their selections, control a device or answer questions only by their mere thoughts. Due to the noisy nature of brain signals, the required time for each experimental session must be lengthened to reach satisfactory accuracy. This is the trade-off between the speed and the precision of a BCI system. In this paper, we propose a unified method which is the integration of ensemble classifier, dynamic stopping, and adaptive learning. We are able to both increase the accuracy, as well as to reduce the spelling time of the P300-Speller. Another merit of our study is that it does not require the training phase for any new subject, hence eliminates the extensively time-consuming process for learning purposes. Experimental results show that we achieve the averaged bit rate boost up of 182% on 15 subjects. Our best achieved accuracy is 95.95% by using 7.49 flashing iterations and our best achieved bit rate is 40.87 bits/min with 83.99% accuracy and 3.64 iterations. To the best of our knowledge, these results outperformed most of the related P300-based BCI studies.
Vo, A, Nguyen, DN, Kha, HH & Dutkiewicz, E 2017, 'Dynamic stopping using eSVM scores analysis for event-related potential brain-computer interfaces', 2017 11th International Symposium on Medical Information and Communication Technology (ISMICT), International Symposium on Medical Information and Communication Technology, IEEE, Lisbon, Portugal.View/Download from: Publisher's site
In brain-computer interface (BCI) research, there must be a trade-off between accuracy and speed of the BCI system, especially those based on event-related potentials (ERPs). This paper proposes a novel method which can significantly increase the spelling bit rate while also maintaining the desired accuracy. We provide an adaptive real-time stopping method based on the scores of ensemble support vector machine classifiers. We apply a criteria assessment process on the classifiers' scores to dynamically stop the ERP-evoked paradigms at any flashing sequence. Our experiments were conducted on three different P300-Speller data sets (BCI Competition II, BCI Competition III and Akimpech). Our proposed framework significantly outperformed the related state-of-the-art studies in terms of character output accuracy and elicitation bit rate rise between static and dynamic stopping schemes. We improve the average bit rate by over 80% while perfectly maintaining the best original static accuracy of over 96%.
Vo, K, Nguyen, D, Ha, K & Dutkiewicz 2017, 'Real-Time Analysis on Ensemble SVM Scores to Reduce P300-Speller Intensification Time', Proc. of the 39th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC'17), Annual International Conference of the IEEE Engineering in Medicine & Biology Society, IEEE, Korea, pp. 4383-4386.
In most Brain-Computer Interface systems, especially
the P300-Speller, there must be a harmonized balance
between the accuracy and the spelling time. One major drawback
of the classical 36-choice P300-Speller is the slow rate
of character elicitation. This paper aims to propose a realtime
signal processing method to decrease the spelling time by
exploiting the score margins of the ensemble Support Vector
Machine classifiers during real-time P300-Speller flashes, rather
than just getting the classifiers’ highest scores. Our experiments
were conducted on the dataset of the BCI Competition III and
resulted in a successful character rate of over 96% with just
approximately 15 to 20 seconds for each character spelling
session. As compared with the fixed 31.5 seconds of the best
original approach of the competition, our proposed method
significantly reduces the required spelling time by over 30%
while maintaining the desired classification accuracy