HCI-BCI RESEARCH @ IIT KHARAGPUR

One such promising application area is the brain computing interface. With such technology in hand, it has become easy to monitor human behaviour before assigning them to critical tasks. Effective prevention of dangerous events is possible if a risk assessment process is considered in advance, particularly the role of man in accident dynamics. Out of several factors influencing human errors, two of them are of prime importance are: vigilance and mental workload. Vigilance or sustained attention is an act of careful observation of critical or rare events whose negligence may lead to some undesirable circumstances. In this regard, we propose to use Event Related Potentials (ERPs) of EEG signals to indicate level of vigilance. Mental Workload, on the other hand, is the demand placed on humans for mental resources while performing a task. One way to increase the performance of the operator is to constantly monitor their performance and then to take a strategy to avoid risk due to human resources. This influences the practical ramifications in many automated systems designed to forewarn monitors of impending, untoward events in the work world. Further, it would be more effective, if it can be accomplished in ground situation when operators are in operations as an instantaneous, online process without any uncomfortable setup that affects operators' normal operation.

Existing approaches assumes student’s cognitive state by analysing their response to tasks and activities during learning. If the e-learning system can comprehend student’s cognitive state precisely by considering their brain signals while learning, then it can make a significant improvement in building an adaptive e-learning system. In the direction of satisfying this goal, our research objective is to augment Brain Computing Interface (BCI) with an e-learning system to make it as an adaptive e-learning system. To achieve this task, the common BCI steps such as channel selection, feature extraction, feature selection and translation algorithm has to be processed to measure the student’s cognitive state. Finally, depending on the cognitive load level of the student, the e-course content has to be changed or revised, so that the user can enjoy his learning.

• The subjects were asked to perform imagery left hand, right hand and tongue according to a cue displayed on the screen.
• Each experiment includes several runs (at least 2) and trials (at least 20).
• Each trial includes the following order. The first 3s of the procedure will be quite and at t=3s an acoustic stimulus and a cross “+” will be displayed to indicate the beginning of trial.
• Then at t=4s an arrow to the left, right, or sphere in the centre will be displayed for 1s and at the same time the subject will be asked to imagine a left hand, right hand, and tongue movements respectively, till the cross appears on the screen.
• The cross "+" disappears at t=8s. Therefore, each trial takes 8s totally. Each of three cues will be displayed 20 times for each run in a randomised order.
• There was 5-10 minute gap between each run.

Subjects	Gender	Age	Session	Sleeping hours	Medicine Consumption
S1	M	28	FN	~8 hours	No
S2	F	29	FN	~6 hours	No
S3	F	31	AN	8 hours	No
S4	M	26	AN	6 hours	No
S5	M	32	FN	~7 hours	No

Vigilance Task: We used the popular Mackworth Clock Test for the vigilance detection experiment. In this task, the task of the participant is to monitor a red pointer moving in front of a black background in steps like the seconds’ hand of a clock and responds when the clock randomly skips a beat. In a session, whenever the clock skips a beat, the user has to promptly respond by pressing the "space bar" key.

Subject	Gender	Age	Sleeping hours	Medicine Intake	Tea/ Coffee Consumption
S1	M	28	~8 hours	No	No
S2	F	29	~6 hours	No	No
S3	F	31	8 hours	No	No
S4	M	26	6 hours	No	No
S5	M	32	~7 hours	No	No
S6	M	27	10 hours	No	No
S7	M	28	7 hours	No	No
S8	M	30	5 hours	No	No
S9	M	32	8 hours	No	No
S10	M	24	6 hours	No	No

Experimental Setup

The experiments were conducted in controlled environment in an electrically isolated BCI laboratory at IIT Kharagpur. We used R.M.S. EEG device and its 64 channel brain cap for this data collection. To minimize any of the artifacts generated, the subjects were asked to avoid any unnecessary physical movements and their hands were placed in a fixed position. Subjects were requested to refrain from excessive blinking of eye lid.

1. Each subject read a paragraph written in English language, appearing on the screen in front of him or her.
2. This experiment has included 6 trials for each subject.
3. Each trial included the following order. The first 6s of the procedure was quite, subject did nothing and at t=7s one paragraph has been displayed and subject started reading. This indicates the beginning of trial. The paragraph has been displayed for 1 minute.
4. After this, two more paragraphs of different level loads have been appeared in the screen for 1 minute each.
5. Totally, after 3minutes and 7s screen was black and this indicates the end of trial.
6. Same way one subject has gone through 6 trials.
7. There was a 10 minute gap between each trial. The total time taken for each trial is given in the table below.
8. Total time taken for each session that is for each subject is 1 hour 30 minutes.

Participants	Gender	Age	Medically fit?	Time of data collection
S1	M	26	Yes	FN
S2	M	27	Yes	AN
S3	M	29	Yes	FN
S4	M	24	Yes	FN
S5	M	26	Yes	AN
S7	M	29	Yes	FN
S8	M	31	Yes	FN
S9	F	27	Yes	FN
S10	F	29	Yes	FN
S11	F	28	Yes	AN