News

July 2015

Congratulations to Dr. Sarah Ostadabbas for accepting a faculty position at Northeastern University. We at the GT Bionics lab look forward to you making us all proud. We'll miss you!


March 2015

Dr. Ghovanloo was named an IEEE Circuits and Systems Society Distinguished Lecturer for 2015-2016 term. The topics on which he will lecture are "Implantable and Wearable Microelectronic Devices to Improve Quality of Life for People with Disabilities" and "Efficient Power and Wideband Data Transmission in Near Field."


November 2014

Tongue Drive featured in US NEWS and World Report: Wearable Tech for People With Disabilities.


August 2014

Congratulations to Temi for winning the best student paper competition award at the IEEE EMBC'14. All of us here at the GT Bionics lab are proud of you!


GT Bionics participated in the 36th Annual IEEE Engineering in Medicine and Biology Conference in Chicago with 6 papers.


July 2014

A new story featuring the Tongue Drive System was published in the New Scientist. Please click here for more details.


Congratulations to Dr. Mehdi Kiani for accepting a faculty position at Penn State. We at the GT Bionics lab look forward to you making us all proud. We'll miss you!


Congratulations to Temi for her EMBC'14 paper being selected among the Open Finalists for the student competition. Her paper was titled "Tracheal Activity Recognition Based on Acoustic Signals".


Low-Power Head-Mounted
Deep Brain Stimulator

This project seeks to develop wireless circuit interface and associated electronics for an implantable neural stimulating microsystem with a large number of stimulating sites for use in neural prostheses. The implantable microsystem should be inductively powered, button-sized, with 1024 sites, arranged in a 3-D configuration, with 128 simultaneous channels, each capable of sourcing ±100mA. The major challenges towards this goal are the implant size, microassembly method, large number of sites, effective and safe stimulation, low power consumption, and wideband wireless link between the implant and the external world.

The electrical connection to the neural tissue is formed through either a group of metal microwire electrodes or a micromachined silicon microelectrode array. For every recording channel, a low-noise low-power amplifier (LNA), which is capable of amplifying signals from mHz to kHz range, is used to amplify the acquired neural signals. A capacitive highpass filter at the input of every LNA rejects the large DC offset generated at the electrode-tissue interface but not the low-frequency evoked potentials that may contain significant neural information. 32 identical neural recording channels plus 4 monitoring channels that marks the beginning of each frame are multiplexed by a 36 to 1 multiplexer that is controlled by circular shift register (SHR). The SHR is run at 720 kHz by a triangular waveform generator, taking 20k samples/sec from every channel. This sampling rate should be enough for reconstruction of the neural signals which have a bandwidth of 8~10 kHz. A sample and hold (S&H) circuit follows the TDM to stabilize the acquired samples before pulse width modulation (PWM). The PWM is dedicated to convert the analog signal at the output of the S&H to a pseudo-digital signal that is more robust against noise. Using a pulse width modulator instead of an analog to digital converter (ADC) results in less power consumption, easier synchronization, and less complexity in the implantable device.

A voltage controlled oscillator (VCO) converts the PWM signal to a frequency shift keyed (FSK) carrier in the industrial, scientific, and medical (ISM) band. Due to the short range application of the system (within the animal cage), the VCO output can be directly applied to a miniature patch antenna with a proper off-chip matching circuit. A custom-designed ISM-band receiver is used as the external part of the system. The received PWM signal is directly converted to digitized samples using Time-to-Digital Conversion (TDC) technique on an FPGA, and transferred to a PC through USB. Finally by demultiplexing the TDM samples, the original neural signals are reconstructed. The wireless neural recording system also contains a receiver coil followed by an on-chip rectifier, filter, and regulator that provide the rest of the implant with a clean DC supply. The power carrier frequency is selected to have minimum interference with the neural signals and ISM carrier.

Related Publications