About

Ralph Etienne-Cummings is the Julian S. Smith Professor of electrical and computer engineering at Johns Hopkins University and serves as the vice provost for faculty affairs. He is a pioneer in mobile robotics and legged locomotion, with innovations that aim to enable computers to perform recognition tasks as effortlessly and efficiently as humans. His research includes developing systems and algorithms for biologically inspired and low-power processing, biomorphic robots, closed-loop neural prosthetics, and computer integrated surgical systems and technologies. Etienne-Cummings has made numerous contributions to the field, including helping to develop the first large-scale neural computer using very-large-scale integration (VLSI) chips and publishing the first paper on pulse-based, inter-pixel time of travel motion chips based on fly motion detection. His work has extended into silicon retinas, guiding micro unmanned aerial vehicles, and interfacing electronics with the nervous system. His recent projects involve implantable devices for intra-spinal micro-stimulation to mitigate spinal cord injuries, wireless physiological sensors for cardiac health monitoring, and ultrasonic imaging systems for infertility treatment. He has published over 230 technical articles, holds 16 patents, and authored a book and multiple book chapters. Throughout his career, Etienne-Cummings has contributed to the development of silicon Central Pattern Generators for biped locomotion, which enabled real-time adaptation and smooth movement in legged robotics. His work has evolved into brain-machine interfaces and neural prostheses aimed at restoring function after injury and human augmentation. He has held leadership roles, including founding director of the JHU Institute of Neuromorphic Engineering, and has consulted for numerous technology firms. Recognized for his impact, he has received multiple awards, including being named a Fellow of AIMBE and IEEE, and has participated in DARPA projects. His educational background includes a BSc in Physics from Lincoln University and MSEE and PhD degrees in electrical engineering from the University of Pennsylvania.

Research topics

• Computer Science
• Artificial Intelligence
• Telecommunications
• Physics
• Computer vision
• Optoelectronics
• Optics

Selected publications

• BLEscope: A Bluetooth Low Energy (BLE) Microscope for Wireless Multicontrast Functional Imaging

  IEEE Transactions on Biomedical Engineering · 2024 · 1 citations

  • Computer Science
  • Artificial Intelligence
  • Computer Science

  Recent advances in low-power wireless-capable system-on-chips (SoCs) have accelerated diverse Internet of Things (IoT) applications, encompassing wearables, asset monitoring, and more. Concurrently, the field of neuroimaging has experienced escalating demand for lightweight, untethered, low-power systems capable of imaging in small animals. This article explores the feasibility of using a low-power asset monitoring system as the basis of a new architecture for fluorescence and hemodynamic contrast-based wireless functional imaging. The core system architecture hinges on the fusion of a Bluetooth Low Energy (BLE) 5.2 SoC and a low-power 560 × 560, 8-bit monochrome CMOS image sensor module. Successful integration of a multicontrast optical front-end consisting of a fluorescence channel (FL) and an intrinsic optical signal (IOS) channel resulted in the creation of a wireless microscope called 'BLEscope'. Next, we developed a wireless (i.e., BLE) protocol to remotely operate the BLEscope via a laptop and acquire in vivo images at 1 frame per second (fps). We then conducted a comprehensive characterization of the BLEscope to assess its optical capabilities and power consumption. We report a new benchmark for continuous wireless imaging of ∼1.5 hours with a 100 mAh battery. Via the FL channel of the BLEscope, we successfully tracked the kinetics of an intravenously injected fluorescent tracer and acquired images of fluorescent brain tumor cells in vivo. Via the IOS channel, we characterized the differential response of normal and tumor-associated blood vessels to a carbogen gas inhalation challenge. When miniaturized, the BLEscope will result in a new class of low-power, implantable or wireless microscopes that could transform preclinical and clinical neuroimaging applications.

  PublisherOA PDFDOI

• Pixel-wise programmability enables dynamic high-SNR cameras for high-speed microscopy

  Nature Communications · 2024 · 14 citations

  • Computer Science
  • Artificial Intelligence
  • Computer Science

  High-speed wide-field fluorescence microscopy has the potential to capture biological processes with exceptional spatiotemporal resolution. However, conventional cameras suffer from low signal-to-noise ratio at high frame rates, limiting their ability to detect faint fluorescent events. Here, we introduce an image sensor where each pixel has individually programmable sampling speed and phase, so that pixels can be arranged to simultaneously sample at high speed with a high signal-to-noise ratio. In high-speed voltage imaging experiments, our image sensor significantly increases the output signal-to-noise ratio compared to a low-noise scientific CMOS camera (~2-3 folds). This signal-to-noise ratio gain enables the detection of weak neuronal action potentials and subthreshold activities missed by the standard scientific CMOS cameras. Our camera with flexible pixel exposure configurations offers versatile sampling strategies to improve signal quality in various experimental conditions.

  PublisherOA PDFDOI

• A closed-loop all-electronic pixel-wise adaptive imaging system for high dynamic range video

  arXiv (Cornell University) · 2019-06-24 · 1 citations

  preprintOpen access

  We demonstrated a CMOS imaging system that adapts each pixel's exposure and sampling rate to capture high dynamic range (HDR) videos. The system consist of a custom designed image sensor with pixel-wise exposure configurability and a real-time pixel exposure controller. These parts operate in a closed-loop to sample, detect and optimize each pixel's exposure and sampling rate to minimize local region's underexposure, overexposure and motion blurring. Exposure control is implemented using all-integrated electronics without external optical modulation. This reduces overall system size and power consumption. The image sensor is implemented using a standard 130nm CMOS process while the exposure controller is implemented on a computer. We performed experiments under complex lighting and motion condition to test performance of the system, and demonstrate the benefit of pixel-wise adaptive imaging on the performance of computer vision tasks such as segmentation, motion estimation and object recognition.

  PublisherOA PDFDOI

• Real Time Compressive Sensing Video Reconstruction in Hardware

  IEEE Journal on Emerging and Selected Topics in Circuits and Systems · 2012-09-01 · 13 citations

  articleSenior author

  Compressive sensing has allowed for reconstruction of missing pixels in incomplete images with higher accuracy than was previously possible. Moreover, video data or sequences of images contain even more correlation, leading to a much sparser representation as demonstrated repeatedly in numerous digital video formats and international standards. Compressive sensing has inspired the design of a number of imagers which take advantage of the need to only subsample a scene, which reduces power consumption by requiring acquisition and transmission of fewer samples. In this paper, we show how missing pixels in a video sequence can be estimated using compressive sensing techniques. We present a real time implementation of our algorithm and show its application to an asynchronous time-based image sensor (ATIS) from the Austrian Institute of Technology. The ATIS only provides pixel intensity data when and where a change in pixel intensity is detected, however, noise randomly causes intensity changes to be falsely detected, thereby providing random samples of static regions of the scene. Unlike other compressive sensing imagers, which typically have pseudo-random sampling designed in at extra effort, the ATIS used here provides random samples as a side effect of circuit noise. Here, we describe and analyze a field-programmable gate array implementation of a matching pursuit (MP) algorithm for compressive sensing reconstruction capable of reconstructing over 1.9 million 8 × 8 pixel regions per second with a sparsity of 11 using a basis dictionary containing 64 elements. In our application to ATIS we achieve throughput of 28 frames per second at a resolution of 304 × 240 pixels with reconstruction accuracy comparable to that of state of the art algorithms evaluated offline.

  PublisherDOI

• Optimization Methods for Spiking Neurons and Networks

  IEEE Transactions on Neural Networks · 2010-10-20 · 52 citations

  articleOpen accessSenior author

  Spiking neurons and spiking neural circuits are finding uses in a multitude of tasks such as robotic locomotion control, neuroprosthetics, visual sensory processing, and audition. The desired neural output is achieved through the use of complex neuron models, or by combining multiple simple neurons into a network. In either case, a means for configuring the neuron or neural circuit is required. Manual manipulation of parameters is both time consuming and non-intuitive due to the nonlinear relationship between parameters and the neuron's output. The complexity rises even further as the neurons are networked and the systems often become mathematically intractable. In large circuits, the desired behavior and timing of action potential trains may be known but the timing of the individual action potentials is unknown and unimportant, whereas in single neuron systems the timing of individual action potentials is critical. In this paper, we automate the process of finding parameters. To configure a single neuron we derive a maximum likelihood method for configuring a neuron model, specifically the Mihalas-Niebur Neuron. Similarly, to configure neural circuits, we show how we use genetic algorithms (GAs) to configure parameters for a network of simple integrate and fire with adaptation neurons. The GA approach is demonstrated both in software simulation and hardware implementation on a reconfigurable custom very large scale integration chip.

  PublisherOA PDFDOI

• Discriminating Multiple Nearby Targets Using Single-Ping Ultrasonic Scene Mapping

  IEEE Transactions on Circuits and Systems I Regular Papers · 2010-06-25 · 3 citations

  articleSenior author

  We present a software simulation and a hardware proof of concept for a compact low-power lightweight ultrasonic echolocation design that is capable of imaging a 120 field of view with a single ping. The sensor uses a single transmitter and a linear array of ten microphones, followed by a bank of eight spatiotemporal filters to determine the bearing angle of returned echoes. The sensor is capable of detecting multiple objects with a single omnidirectional ping, even if their echoes interfere with each other at the microphone array. The hardware implementation detects the bearing of nearby objects with an rms accuracy of 1.6° and can reliably detect a 70-cm-long 5-cm-diameter metal table leg at a range of 3 m. Stronger reflectors, such as building corners, can be reliably detected at a range of 9 m.

  PublisherDOI

• Decoding Individuated Finger Movements Using Volume-Constrained Neuronal Ensembles in the M1 Hand Area

  IEEE Transactions on Neural Systems and Rehabilitation Engineering · 2008-02-01 · 76 citations

  article

  Individuated finger and wrist movements can be decoded using random subpopulations of neurons that are widely distributed in the primary motor (M1) hand area. This work investigates 1) whether it is possible to decode dexterous finger movements using spatially-constrained volumes of neurons as typically recorded from a microelectrode array; and 2) whether decoding accuracy differs due to the configuration or location of the array within the M1 hand area. Single-unit activities were sequentially recorded from task-related neurons in two rhesus monkeys as they performed individuated movements of the fingers and the wrist. Simultaneous neuronal ensembles were simulated by constraining these activities to the recording field dimensions of conventional microelectrode array architectures. Artificial neural network (ANN) based filters were able to decode individuated finger movements with greater than 90% accuracy for the majority of movement types, using as few as 20 neurons from these ensemble activities. Furthermore, for the large majority of cases there were no significant differences (p < 0.01) in decoding accuracy as a function of the location of the recording volume. The results suggest that a brain-machine interface (BMI) for dexterous control of individuated fingers and the wrist can be implemented using microelectrode arrays placed broadly in the M1 hand area.

  PublisherDOI

• The feeling of color: A haptic feedback device for the visually disabled

  2008-01-01 · 15 citations

  articleOpen accessSenior author

  We describe a sensory augmentation system designed to provide the visually disabled with a sense of color. Our system consists of a glove with short-range optical color sensors mounted on its fingertips, and a torso-worn belt on which tactors (haptic feedback actuators) are mounted. Each fingertip sensor detects the observed object’s color. This information is encoded to the tactor through vibrations in respective locations and varying modulations. Early results suggest that detection of primary colors is possible with near 100% accuracy and moderate latency, with a minimum amount of training.

  PublisherOA PDFDOI

• Asynchronous Decoding of Dexterous Finger Movements Using M1 Neurons

  IEEE Transactions on Neural Systems and Rehabilitation Engineering · 2008-02-01 · 100 citations

  articleOpen access

  Previous efforts in brain-machine interfaces (BMI) have looked at decoding movement intent or hand and arm trajectory, but current cortical control strategies have not focused on the decoding of dexterous [corrected] actions such as finger movements. The present work demonstrates the asynchronous decoding (i.e., where cues indicating the onset of movement are not known) of individual and combined finger movements. Single-unit activities were recorded sequentially from a population of neurons in the M1 hand area of trained rhesus monkeys during flexion and extension movements of each finger and the wrist. Nonlinear filters were designed to detect the onset of movement and decode the movement type from randomly selected neuronal ensembles (assembled from individually recorded single-unit activities). Average asynchronous decoding accuracies as high as 99.8%, 96.2%, and 90.5%, were achieved for individuated finger and wrist movements with three monkeys. Average decoding accuracy was still 92.5% when combined movements of two fingers were included. These results demonstrate that it is possible to asynchronously decode dexterous finger movements from a neuronal ensemble with high accuracy. This work takes an important step towards the development of a BMI for direct neural control of a state-of-the-art, multifingered hand prosthesis.

  PublisherOA PDFDOI

• Finite element modeling of tissue for optimal ultrasonic transducer array design

  2008-05-01 · 1 citations

  articleSenior author

  Tissue characterization using ultrasound scattering has been routinely used to extract the cellular properties of tissue. Ultrasonic backscattered radio frequency (RF) data is analyzed to provide estimates of the size, shape and concentration of a wide range of tissues. These tissue parameters form a feature space that can be used to discriminate between tissue types as well as indicate the presence of disease. In this work we develop numerical finite element models of tissue by using spectrum analysis to extract the acoustic properties of three distinct tissue-mimicking phantoms. The tissue-mimicking phantoms are constructed using glass microspheres of diameters 15-45 mum, 40-70 mum and 90-150 mum, embedded in a uniform concentration of gelatin derived from porcine skin. Spectrum analysis of the backscattered data from the tissue phantoms yields acoustic properties (of scatterer size, shape and distribution) which are used to create finite element models (FEM) of the tissue phantom. Simulations of acoustic scattering from an FEM phantom consisting 40 mum scatterers located in a transducer resolution cell is performed. Spectral analysis of the backscattered from the FEM phantom yields an estimate of the scatterer size of 36.7 mum giving an accuracy of 91%.

  PublisherDOI

Recent grants

• SST: Minimally-Attended Integrated Visual Surveillance Network

  NSF · $412k · 2004–2008

• Annual Telluride Workshop on Neuromorphic Engineering: Telluride, CO 6/27/0407/17/04; 2004-2009

  NSF · $400k · 2004–2009

• Learning Shape Representation in Somatosensory Cortex and Their Applications to Upper Limb Prosthetics

  NSF · $196k · 2011–2013

Frequent coauthors

• Francesco V. Tenore

  Johns Hopkins University Applied Physics Laboratory

  7 shared

• Nitish V. Thakor

  Johns Hopkins University

  5 shared

• Soumyadipta Acharya

  4 shared

• Garrick Orchard

  4 shared

• C. Posch

  3 shared

• Clyde Clarke

  Johns Hopkins University

  3 shared

• Jonathan Tapson

  University of Technology Sydney

  3 shared

• Amir Fahmy

  Johns Hopkins University

  3 shared

Awards & honors

• Fellow of AIMBE (2021)
• Fellow of IEEE (2012)
• Johns Hopkins Discovery Awards (2019, 2018)
• Research competition first place at the 5th annual Johns Hop…
• Indispensable Role of African Americans at JHU honor