Pitt Innovation Challenge 2020
PInCh 2020 Finalists<< Back to competition home
A novel EEG lead clip designed for people with coarse and curly hair, bringing innovative human-centered design to provide a 15x improvement in measurement accuracy for epilepsy, neurological disorder, stroke, and brain injuries in the Black population.
- Electroencephalography (EEG) is a neural sensing modality essential for diagnosis of epilepsy, stroke, brain injuries, and other neurological disorders.
- EEG is not reliable for individuals with coarse and curly hair common in Black people because of difficulty in applying electrodes on the scalp with this hair-type.
- Our novel Sevo clips obtain 15x reductions in electrode-skin impedance, improving data quality for individuals with coarse/curly hair (Sevo is “brain” in Haitian Creole).
- The proposed work will obtain rigorous clinical validation as well, thus obtaining necessary data to convince clinical centers across the world to adopt these techniques.
- Novel clips, that do not require braiding and work with all hair-types, will also be developed and tested.
More than 7 million Americans need EEG recordings each year due to epilepsy, brain injuries, and stroke. However, many Black patients are unable to receive reliable EEG recordings because state-of-the-art EEG systems are not compatible with coarse and curly hair (see here for some of many examples we are aware of). This can result in poor quality diagnoses and treatments for Black patients. Indeed, Black people are routinely denied participation in EEG experiments due to their hair-type (as noted in our website that collects the community response to Sevo).
In an attempt to make current systems work, Black patients are often advised to straighten their hair to improve electrode scalp contact. However, this only works for a short time (30-60 minutes – often not enough to travel to the clinic and get a test done). Sevo solves this problem using electrode clips that produce high quality EEG recordings. Sevo clips turn the problem upside down, viewing coarse and curly hair as a benefit instead of an impediment to reliable recordings, leveraging its strength to make improved contact with the scalp.
Sevo clips: the hair is braided in cornrowed fashion, exposing parts of the scalp needed for recording EEG in the clinical standard “10-20” arrangement of scalp locations. Wing-shaped Sevo clips are then placed on the scalp by leveraging the strength of braided hair. The braiding is consistent with the Black culture in many parts of the world. The overall solution leads to a striking improvement in performance: Sevo reduces electrode-skin impedance > 15x over state-of-the-art [Etienne et al., IEEE EMBC’20], and keeps impedance low all the time (compared to existing systems, which work 65% of the time, even after braiding). In doing so, Sevo provides a dignified and culturally competent solution compared to less effective current measures, e.g. straightening hair.
Advancing on Sevo, we will develop two additional designs, Sevo Firefly and Sevo Dragonfly, which will be the first solutions that do not require braiding. Our goal is to build on Sevo to release a line of self-contained entirely wearable, wireless Sevo systems that work with all hair-types, enabling sufferers of various disorders to wear electrodes without privacy concerns for multiple days (as needed, e.g., in planning epilepsy treatment), and enabling neuroenthusiasts to express themselves through their choice of electrodes.
The competitive landscape analysis below summarizes key features of this solution, and current competitors working to solve similar healthcare problems.
Christina Patterson, MD, PI, Director of Epilepsy Services and the Epilepsy Monitoring Unit, UPMC Children’s Hospital. Having first-hand experience with patients on whom existing electrode systems do not work, Dr. Patterson brings the clinical expertise and experience required for innovations that are aligned with the end goals.
Arnelle Etienne, BS, Co-PI, Director of Accessibility, Precision Neuro; Sevo electrodes lead inventor. An unstoppable innovator and problem solver, Ms. Etienne creatively utilizes her expertise with rigorous EEG testing and engineering, and her own experience as an African American using EEG, to identify and address novel issues in existing systems for the purpose of increasing accessibility.
Pulkit Grover, PhD, Co-PI, Associate Professor, Electrical & Comp Engineering, Biomedical Engineering, Neuroscience Institute, CMU; CTO, Precision Neuro. Dr. Grover uses his neuroengineering and information theory expertise to ensure designed EEG systems meet or exceed the signal-to-noise ratio of existing systems.
Shawn Kelly, PhD, CEO, Precision Neuro; Senior Systems Scientist, Engineering Research Accelerator, Electrical Engineering, Biomedical Engineering, CMU. Dr. Kelly brings decades of experience and expertise in biomedical instrumentation, including extremely complex retinal prosthetic systems, and combines it with his business expertise through involvement in several startups.
Ashwati Krishnan, PhD, Hardware Lead, Precision Neuro, CMU. Obsessed with low-cost and high-quality patient care, Dr. Krishnan is a hardware expert who helps designers in the team, such as Ms. Afelin and Ms. Etienne, bring their ideas to life.
Amber (Momi) Afelin, BA, Lead Electrode Designer, Precision Neuro, Neuroscience and Behavior. Synthesizing her background with neuroscience, entrepreneurship, and expertise in design, Ms. Afelin is designing novel EEG solutions that do not require braiding and work with all hair-types, and are fashionable, enabling individuals to wear EEG while protecting their privacy or expressing their individuality, as they desire.
Jasmine Kwasa, PhD student, Electrical & Computer Engineering. Ms. Kwasa, soon to receive her PhD at CMU in auditory neuroscience and attentional disorders, will use her neuroscience expertise to design the experimental protocols to rigorously characterize the gain in the signal to noise ratio.
Milestone 1.1: Adapt Sevo clips to currently used systems at UPMC Children’s Epilepsy Center, and begin clinical testing on developed prototypes. Pilot testing on five epilepsy patients to do iterations and finalize the design. Begin validation data collection on epilepsy patients. Milestone 1.2: Complete rigorous survey of potential customers to understand needs (e.g., setup time, number of electrodes).
Milestone 2.1: Begin design of prototypes for high-density, fast installation and bulk manufacturing, evaluating different materials for bulk processing. Identify vendors and services for molding/manufacturing. Milestone 2.2: Complete data collection using Sevo electrodes on 15 epilepsy patients, file for IP, and submit for publication.
Milestone 3.1: Complete design of Sevo Firefly and Sevo Dragonfly electrodes (that do not require braiding) with fast installation and test on healthy participants (stretch goal: test on epilepsy patients), with high electrode counts (64 or more) for all three designs. File for IP and submit for publication. Identify manufacturing process. Milestone 3.2: Submit FDA presubmission document.
Milestone 4.1 and Impact: Pilot use: Identified research labs who have expressed interest will start pilot use, enabling first direct impact (beyond the clinical studies ongoing throughout the year).
Path to Impact Plan
We have a five-pronged plan for impact:
(i) Obtain validation data in Pitt Epilepsy Center and present in clinical meetings and journals: Systematic and rigorous demonstration of how our system dramatically outperforms existing systems;
(ii) Provide Sevo solutions to existing research labs: to enable acquisition of more data and validation of the improvements in these labs. 10 such labs have already contacted us;
(iii) Continue IP development and publication in scientific and clinical conferences, networking with customers;
(iv) Reach out to EEG equipment companies to sell Sevo clips on their systems: We have been contacted by OpenBCI, who expressed enthusiasm for Sevo systems, and are planning to work with them as a start. We plan to reach out to other companies in this space as well;
(v) Continue publicizing our work to add to the first users’ list – in clinical and neuroscience space – who will act as brand ambassadors.
Frequently Asked Questions
Are these systems just “nice to have” or do we really need them?
Poor contact of just one electrode can lead to inaccurate/erroneous interpretation and misdiagnosis. Dr. Patterson has observed cases where EEG was uninstallable on patients’ thick hair, even by an experienced technologist. We are developing a catalog of clips for an individual’s preferences to obtain reliable recordings regardless of hair-type /braiding.
Why did it take almost 100 years after EEG was invented to identify this problem?
The most obvious culprit is lack of diversity in science and engineering, and consequently, the implicit assumptions researchers have about users of their products. However, that researchers work in silos also left this issue unaddressed. People making EEG devices are often not the people using them in the clinic.
Can we make existing solutions work?
There are cases where existing solutions completely fail. Ad-hoc solutions used by people facing the problem don’t work because loss of information from even one electrode can hurt diagnoses. This is not a cost we can afford as a society.
What is the Sevo team’s broad accessibility vision?
High-quality neural imaging is inaccessible for many reasons: a) equipment’s expense; b) need for trained technicians to install and maintain; c) provide limited information. We are inventing novel materials and wearable, wireless, high-density EEG, along with Artificial Intelligence algorithms that provide clinically relevant high-resolution information to the clinician in real-time.