An Electroencephalogram Signal based Triggering Circuit for controlling Hand Grasp in Neuroprosthetics G. Karthikeyan1, Debdoot Sheet2 and M. Manjunatha2 1

2

Department of Biomedical Engineering, SSN College of Engineering, Chennai, INDIA. School of Medical Science and Technology, Indian Institute of Technology, Kharagpur, INDIA.

Abstract— Quadriplegia is a serious problem in the case of patients with neurological disorders. Functional Electrical Stimulation (FES) has been a very good rehabilitation technique to treat this condition and help the patient to lead a near-normal life by aiding him/her to move the limbs of the upper and lower extremities with less difficulty. Various techniques have been proposed to trigger the FES system. In this paper, we describe the design of a novel circuit used to trigger a FES device by a person’s EEG signals i.e., by his thoughts. The total project was divided into three modules. The first module was to design a proper interface between the electrodes placed on the scalp and the electronic system which was to be used as a trigger. The second module was to amplify the signal to a sufficient level such that the strength of the signal is high enough to drive the third module which served as the classifier part. The classifier part of the circuit was built out of commercially available IC’s and external discrete components. Though there was some tolerance errors induced due to the external components, the error was at a minimal rate when compared with the actual signal considered. The circuit was powered by a 9V battery and the only input to it was the thought waves through EEG signals from the subject/patient considered. The circuit is low power efficient with a wide operational range of ±3V to ±18V. Keywords— Circuit synthesis, Electroencephalography, Electronic equipment, Filters, Instrument amplifiers, Integrated circuits, Logic design.

I. INTRODUCTION Quadriplegia is a medical condition in which one half of the body is completely paralyzed, greatly reducing mobility of the affected person. In the case of upper extremity paralysis (which is the condition dealt with in this study), the affected person suffers from an inability to move his/her hands, resulting in severe inability to grasp objects and perform normal tasks. The two reasons for occurrence of Quadriplegia are Congenital and Result of a stroke or illness. Though there is little development as of now for treatment of the congenital type of Quadriplegia, recent developments in the field of Functional Electrical Stimulation (FES) have lead to ease in the treatment of the second kind of patients. The treatment of stroke induced Quadriplegia is made possible by a FES system due to the fact that

though the Upper Motor Neurons (Connecting the brain with the spinal cord) are damaged, the Lower Motor Neurons (Connecting the spinal cord with the limbs) are healthy. As a result of this, any stimulation given to the nerves connecting the affected limb with the spinal cord stimulates the muscles of the limb and facilitates motion. Recent developments in the acquisition and analysis of Electroencephalogram (EEG) signals have shown that EEG signals are unique for each person and for each thought activity. This property of the EEG has been found to be suitable for usage in Brain- Computer Interface (BCI) applications [1]. Most of the BCI applications use variations in EEG patterns to classify a person’s thoughts and make an external device act according to the patterns recognized by the machine [2]. The EEG based trigger circuit developed in this project acts using the same principle as that of a BrainComputer Interface, detecting patterns in EEG signals using a circuit tailor- made for stimulation of the FES device. II. FUNCTIONAL ELECTRIC STIMULATION:

AN OVERVIEW

A Functional Electrical Stimulator is a neuroprosthetic device, the main function of which is to activate the inactive and weak nerves of the upper/lower extremities of the body. Usage of a FES is fruitful in a way that the affected nerves are rejuvenated by continual usage of the device and this helps in restoring movement to the patient. The affected portion of the patient is stimulated whenever movement is required by the patient. Most of the early FES devices were based upon pure analog designs which rendered the usage and accuracy of usage of the FES device an unpredictable one. But, recent developments in digital technologies and usage of microcontrollers have rendered the usage of FES systems a reliable method of rehabilitation. The schematic of the device used here is shown in Fig. 1 and is controlled using a microcontroller and the output waveform of the circuit is shaped in such a manner that the patient feels negligible fatigue of muscles. Bio-potentials as a command for FES. Numerous works have been done to study the feasibility of using EEG signals for control of FES. A significant result with regards to this

was given by Juul et al (2000) [3] stating that there is a remarkable change in the EEG patterns during preparation of hand or leg movements. This experiment used different kinds of movement trainings to record the Movement Related Potentials (MRPs) for different kinds of movement attempts. Seven recording sites were used according to the 10–20 electrode placement system. The promising results shown by this study prompted us to use C3 and C4 recording sites for tapping the Beta waves of the EEG which are the control signals to be used in triggering the FES.

Fig. 1. Block setup of FES.

III. IDEA OF THE

CIRCUIT

The basic idea of the circuit is divided into two partsThe first part would amplify the circuit to the required level and the second part would classify the amplified signals for triggering of the circuit. The block set up of the total circuit is given is Fig.2 and Fig.3.

The second portion of the circuit is involved in choosing and classifying the frequencies of interest which would facilitate triggering of the FES device. The output/ Gain calculation of the amplification portion of the circuit is shown in Table 1. Table 1 Gain of the various stages of the amplification portion of the circuit Stage

Gain

Preamplifier - I Preamplifier - II Non-Inverting Amplifier - I Non-Inverting Amplifier – II Net Gain of circuit in Fig. 2

2.08 74.46 24.33 7.76 29358

Voltage dectection and FES triggering using comparators and AND gates:A very critical component of the circuit is the Frequency to Voltage (F–V) converter which is used to convert the various frequencies present in the EEG signal into proportional voltages. The output of an F–V converter is a voltage signal which would make this portion of the circuit act as a classification precursor stage. The output of the F–V converter is given into two comparators with one comparator being sensitive to voltages corresponding to frequencies greater than F1 and another comparator being sensitive to frequencies lesser than F2. The outputs of these comparators are treated as logic levels of 1’s and 0’s. These logic levels are given as inputs to the AND gate and this AND gate responds to the logic level inputs and gives out an output which is given as a trigger signal to the FES device and the algorithm for the output and input of the comparators and AND gate is as seen in Table 2. The output of the AND gate, is a logic level of 1 or 0 and would be suitable for use as an input to the trigger switch of the FES device. Table 2: Logic levels of the comparators and AND gate for triggering the FES

Fig.2 Amplifier and Preprocessing set up of the circuit.

Input to AND gate from Comparator 1

Input to AND gate from Comparator 2

Output to the AND gate

0 (F
0 (F≥F2) 1 (F
0 0 0 1

IV. CONCLUSIONS Fig.3 Classification portion of the circuit

The concept and design of the circuit is in such a manner that it would help quadriplegic and with some fine-tuning, paraplegic patients with foot drop to lead a near- normal life. Though the circuit would help in bringing a near-

normal lifestyle back into the life of the patients, it is extremely difficult for any machine to bring a normal life back to the patient at least with present day technology. The best thing which can be done to improve the lifestyle of disabled people is using rehabilitation techniques such as FES and neural engineering methodologies. The major drawback with the circuit described above or for that matter any kind of Brain-Computer Interface technologies is the fact that they require cooperation from the patient’s side and ultimately however strong the technology developed is, for the technology to be successful, cooperation from the patient side is very much essential.

ACKNOWLEDGMENT The authors would like to acknowledge the efforts of Mr. Sourajit Das and Mr. Sukanta Kr. Sabut at the BioInstrumentation Laboratory, School of Medical Science & Technology, IIT, Kharagpur, for their whole hearted cooperation and assistance.

4.

5. 6. 7. 8. 9. 10.

11.

12.

13.

14.

REFERENCES 15. 1.

2.

3.

Thomas Sinkjaer et al.(2002), “Biopotentials as command and feedback signals in functional electrical stimulation systems,” Medical Engineering and Physics. Reinhold Scherer et al.(2007) “The Self-Paced Graz BrainComputerInterface: Methods and Applications,” Computational Intelligence and Neuroscience Juul PR, Ladouceur M, Nielsen K. D. (2000), “Coding of lower limb muscle force generation in associated EEG movement related potentials: preliminary studies toward a feed-forward control of FESassisted walking,” Proc. 5th Ann. Conf. Int. Funct. Stim. Society, pp. 335–337.

16.

Donchin, E., Spencer, K.M., and Wijesinghe, R.(2000), “The mental prosthesis: Assessing the speed of a P300-based brain-computer interface,” IEEE Tran. Rehabilitation Engg., vol. 8 no. 2, pp. 174-179, John G. Webster, Medical Instrumentation- Application and Design, Wiley Publications, Analog Devices 620 AN Datasheet at http://www.datasheetcatalog.com Texas Instruments OP07 CP Datasheet at http://www.datasheetcatalog.com Intersil LM339 Quad Voltage Comparators Datasheet at http://www.datasheetcatalog.com Telecom Semiconductors, TC9400 F–V Converter Datasheet at http://www.datasheetcatalog.com Phil Tresadern et al. (2006), “Artificial Neural Network Prediction Using Accelerometers to Control Upper Limb FES During Reaching and Grasping Following Stroke,” Proc. 28th IEEE EMBS Ann. Int. Conf., NY US Nai-Jen-Huan et al. (2004), “Classification of mental tasks using fixed and adaptive autoregressive models of EEG signals,” Proc. 26th Ann. Int. Conf. IEEE EMBS, San Francisco, CA, USA. Smith J, Jones M Jr, Houghton L et al. (1999), “Future of Health Insurance,” N. Engl. J. Med 965:325–329 DOI 10.10007/s002149800025 Lock I, Jerov M and Scovith S (2003), “Future of modeling and simulation,” IFMBE Proc. vol. 4, World Congress on Med. Phys. & Biomed. Eng., Sydney, Australia, 2003, pp 789–792 Keirn Z. A. and Aunon J. I. (1990), “A new mode of communication between man and his surroundings,” IEEE Trans. Biomed. Eng., 37 1209–14 J.Wolpaw (2006), “Brain-computer interfaces as new brain output pathways,” J Physiol, 579:613–619. IFMBE at http://www.ifmbe.org Corresponding author: Author: G. Karthikeyan Institute: SSN College of Engineering Street: Old Mahabalipuram Road, Kalavakkam, City: Chennai Country: INDIA Email: [email protected]