I am currently a researcher at the U.S. FDA in the Office of Science and Engieering Labs (OSEL). My research interests are in machine learning for signal and image processing with biomedical applications, particularly for medical device algorithms and computational neuroscience. My current projects involve database testing of intelligent medical device alarm algorithms and big data analytics of electroencephalography (EEG).
I obtained my Ph.D. in Electrical and Computer Engineering from the University of Maryland while conducting research in the Human-Device Interaction group at the FDA. I was co-advised by Dr. Kimberly Kontson (FDA) and Prof. Jonathan Simon (UMD).
When I am not working, I am:
Staff researcher and medical device review consultant in the Advanced Patient Monitoring and Controls (AMPC) group in the Division of Biomedical Physics (DBP) under the Office of Science and Engineering Laboratories (OSEL) within the Center for Devices and Radiological Health (CDRH) at the Food and Drug Administration (FDA).
My research aims to advance regulatory science through better understanding and analysis of intelligent medical device alarm algorithms for patient monitoring applications through quantitative metrics and database testing.
Data science, machine learning, statistical and information technology consulting firm focusing on local technology and education companies. Client domains have included: biotechnology, finance, and education.
Aid clients with large amounts of data, generally proprietary, to improved product performance, new product features, and direct future product development.
ORISE Fellow in the Human-Device Interaction (HDI) group in the Division of Biomedical Physics (DBP) under the Office of Science and Engineering Laboratories (OSEL) within the Center for Devices and Radiological Health (CDRH) at the Food and Drug Administration (FDA).
My research aims to advance regulatory science through better understanding and analysis of electrophysiology and other biomarkers as well as through the study of human interaction with existing and emerging medical device technologies.
Manage and support the School of Public Health's technology infrastructure.
Support and troubleshoot computer hardware, software, servers, networking, virtualization, mobile/tablet computing, cloud resources, database, and enterprise systems.
Manage Epic’s database system, via InterSystem’s Caché database, both through development and implementation of database server systems for new Epic customers and support of existing implementations of Epic software at various health organizations.
Unix database administrator and Caché system manager.
Done in the Intelligent Systems Division of the Engineering Lab (E.L.) at NIST. Managed a full independent project on AGV (Autonomously Guided Vehicle) obstacle detection and avoidance.
Designed and constructed computer controlled mechanical system to move obstacles, such as test pieces or mannequins, with controllable displacement and velocity. Further developed AGV’s sensing ability and testing procedure for industry standards.
Details regarding the project cannot be posted outside of NIST as per their regulations.
Software engineering position that demanded learning new languages and software packages as each project was developed in a collaborative environment.
Explored and documented WeBWorK’s capabilities as an online tool for math courses. Created assignments and solutions that are now the sole WeBWorK assignments for introductory math courses at Swarthmore.
Lead problem sessions for students to assist with weekly homeworks, tutored students one-on-one as needed, guide review sessions for exams and grade homework assignments.
The aim of this conference poster and extended abstract is to present preliminary results of research done at the FDA on qEEG. Presentented at the 2017 International Symposium on Wearable Robotics and Rehabilitation (WeRob). Extensions of this work appear in future journal publications.
This paper explores the spectral graph theory and Fourier Transforms on graphs along with their potential applications to electroencephalography.
The aim of this paper is to construct a neural-network based video tracker that learns to associate the motion of tracked objects with a sequence of video frames.
This paper explores the Group Invariant Scattering Transform proposed by Stéphane Mallat in 2012.
For my senior thesis in engineering I designed and constructed a system to acquire and analyze electromyography (EMG) data and then using artificial neural networks recognize various hand gestures to control a mouse and keyboard in real-time. The system had over 99% accuracy for four hand gestures. [Swarthmore Publication]
Presented at Swarthmore's Agora Talks May 2014 as one of seven senior projects across all disciplines.
My mathematics paper for my senior conference delved into the derivation and applications of the Haar and Daubechies-4 wavelets specifically in signal processing.
This research, which spawned from an experiment I designed for a class, has shown that bilinguals represent syntactic structures over multiple languages in their mind as one intertwined syntactic structure rather than separate ones.
This project implemented a version of panorama image stitching where the program searches through a folder of images (for example, a folder of summer vacation photos), and determines if there are appropriate matches for some of the images. If there exist panoramas in the folder, the program outputs all possible panoramas that can be validly stitched together.
Projects delved into path planning, kinematics, stability and implementation of various algorithms for humanoid robots. In particular the Darwin-OP and the HUBO were modeled and used. Projects built up from determinig simple statically stable poses to implementing Zero Moment Point (ZMP) walking while avoiding obsticle. These projects were documented via a blog, with videos, at DJrobotic [so named for the obvious pun and first two letters of the names of myself and Jackie Kay.]
This project given a net-list uses a Modified Nodal Analysis (MNA) algorithm to find the solution to circuits that incorporate dependent and independent, both AC and DC, sources; resistors; inductors; capacitors; transformers; and idealized versions of the operational amplifier (op amp). Additionally, we implanted solution techniques for the simple circuits and single sinusoidal sources and find symbolic solutions for all other cases.
Continuing with tradition at Swarthmore Engineering, I led the prank to 'extend' the Crum Creak meander, a then new art peice on campus, through the dinning hall.
I am also interested in applying these technical skills to other interesting domains with challenging problems.
Recipient of an ORISE Fellowship for research at FDA fall 2016 - fall 2019.
Recipient of a Joshua Lippincott Fellowship in 2015.
Coursework: Statistical Pattern Recognition; Neural Modeling; Advanced Numerical Optimization; Image Understanding; Harmonic Analysis; Selected Topics in Analysis: Wavelets, Time-Frequency Analysis, and Frames; Advanced Digital Signal Processing; Random Processes in Communication and Control; Information Theory; Optimal Control; Embedded Systems; Intellectual Property Seminar; Microelectronics Seminar.
Audited: Neural Basis of Human Movement; Estimation and Detection Theory; Unsupervised Learning; Mathematical Methods in Machine Learning; Computational Linguistics I.
Engineering Design Thesis: Electromyography Analysis and Recognition for Human Device Interface.
Mathematics Thesis: Development and Applications of Wavelets in Signal Processing.
Recipient of The Albert Vollmecke Engineering Award for 2014.
Nominated and elected as an associate member of Sigma Xi, Scientific Research Honors Society.
Certificate of Merit from State of Maryland
College Board AP Scholar
© Copyright 2020 | Designed&Developed by David Nahmias