- Neuroscience and Neural Engineering
- Energy Harvesting in Wireless Networks
- Wireless Power Transfer Systems
- Advanced Sensor and Energy Harvesting Materials
- Modular Robots and Swarm Intelligence
- Wireless Body Area Networks
- Analog and Mixed-Signal Circuit Design
- Robot Manipulation and Learning
- Indoor and Outdoor Localization Technologies
- Underwater Vehicles and Communication Systems
- Photoacoustic and Ultrasonic Imaging
- Advanced Memory and Neural Computing
- RFID technology advancements
- Ultrasound Imaging and Elastography
- Ultrasound and Hyperthermia Applications
- EEG and Brain-Computer Interfaces
Stanford University
2018-2024
Johns Hopkins University
2015-2016
Johns Hopkins Medicine
2016
University of Alberta
2003
We present a bidirectional neural interface with 4-channel biopotential analog-to-digital converter (bioADC) and current-mode stimulator in 180 nm CMOS. The bioADC directly transduces microvolt biopotentials into digital representation without voltage-amplification stage. Each channel comprises continuous-time first-order <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\Delta...
The paper describes a novel design of robot tactile sensor, called hybrid based on analog resistive technology. employs set parallel sensing strips, rather than two-dimensional (2D) grid elements, to sample the sensor surface. It is therefore in one dimension and discrete other. As result, number samples that need be obtained significantly reduced, as connectors circuitry. In addition, scales easily large surface area, offers flexibility geometry. However, only contact location shape can...
We introduce a touch sensor design, which we refer to as hybrid tactile sensor. It combines the simplicity of analog sensing, is widely used in electronic screens and touchpads, with ability shape discrimination array found sensors designed for robotics applications. describe construction techniques sensor, its modelling, interpretation algorithms that determine an object contact Future work outlined using inexpensive materials conductive silicone rubber fabric mesh are described.
Single modality wireless power transfer has limited depth for mm-sized implants across air / tissue or skull interfaces because they either suffer from high loss in (RF, Optical) reflection at the medium interface (Ultrasound (US)). This paper proposes an RF-US relay chip media avoiding boundary, and enabling efficient powering to deep multiple media. The rectifies incoming RF through 85.5% inductive link (across air) using a multi-output regulating rectifier (MORR) with 81% conversion...
We present the first fully wireless implantable imager for distributed long-term high-resolution imaging in body. The consists of two mm-sized piezoelectric transducers (piezos) ultrasound (US) power and data communication; a capacitive micromachined transducer (CMUT) array; 65 nm CMOS IC, which includes harvesting, high voltage charge pump, low-power 16-channel front-end with up to $125\ \text{dB}\Omega$ gain 13 MHz bandwidth, memory system store digitized at 8.6 Gbps, 120 kbps US...
Neural interfaces, including retinal and brain implants, require increasing post-compression uplink data rates $\gt10$ Mb/s to send from larger denser recording arrays. In electrical large arrays can be leveraged identify cell types, stimulation thresholds, locations construct better patterns. Among current solutions, RF has the highest rate but limited depth inside body for small transmitter (Tx) sizes [1, 2]. Galvanic solutions achieve $\sim10$ [3] with low power form factor have scaling...
We present a 4-channel integrated biopotential analog-to-digital converter (bioADC) for electrocortical recording in 180nm CMOS. The bioADC directly transduces microvolt biopotentials into digital representation without voltage-amplification stage. Each channel comprises continuous-time first-order ΔΣ modulator with OTA input and current feedback, followed by second-order comb-filter decimator. Powered from 1.5V supply, the analog circuits each draw 2.1uA 1.4uA of supply current,...
We introduce CRADLE, or combined RF/acoustic detection and localization, for the remote parametric ranging of passive tags. In this approach, a tag consists an RF antenna connected to ultrasonic transducer via electrical network. A reader looks tags by transmitting both continuous-wave pulsed ultrasound into region interest; when is within that region, incident on varies capacitance, creating time-varying load thus parametrically modulating waves reradiated from tag. The localizes...
Single modality wireless power has limited depth for mm-sized implants across air-tissue (e.g. retina implants) or skull-tissue interfaces because of high loss in tissue (RF, optical) reflection at the medium interface (ultrasound). The proposed RF-ultrasound (US) relay uses an RF link to a chip which rectifies and transmits focused ultrasound implant using 6-channel phased array with 251% efficiency improvement 2.5 cm over fixed focusing.