In the general U.S. population, traumatic brain injury (TBI) is the leading cause of mortality in people under age 45 and accounts for over $75 billion in healthcare costs each year. Unfortunately, up to 60% of severely injured TBI patients are under-triaged and admitted to non-trauma hospitals - a critical flaw which in the US leads to an excess mortality rate of 25%. This problem is replicated in many other forms of brain injury, including acute stroke and acute substance intoxication.
In many cases the pupillary light reflex (PLR) is the only objective information a first responder or clinician may have in determining the severity of a patient’s neurological status. A normal PLR is defined as symmetric constriction or dilation of both pupils. The technique of measuring the PLR is known as pupillometry. The PLR can be represented as a curve with multiple characteristics, each of which has a range of normal values, and each of which may be altered to varying degrees. Because the neural pathways underlying the PLR include multiple brain regions, it is sensitive to a wide variety of injuries including stroke, TBI, Alzheimer’s, diabetes, substance use, and sleep deprivation.
There are two methods currently used by clinicians to measure the PLR – digital and manual pupillometry. Digital infrared pupillometry is a current quantitative option. While these devices are accurate and reliable, they suffer from several disadvantages that severely limit their adoption. The upfront cost is about $9,000 per unit and disposable parts are required for each new patient which adds significantly to the ongoing operating cost of the device. Their expense and fragility also make deployment in the field and low resource/underserved settings impractical. Extensive training in both the acquisition and interpretation of results is required and thus their use outside a neurological intensive care unit setting is limited.
The current worldwide standard of practice is manual penlight pupillometry. This low-cost alternative consists of an unaided clinician watching for pupillary constriction while directing a penlight toward and away from the patient’s eyes. While available to any trained clinician with a light source (and part of every neurological assessment), this exam is inherently subjective, and many meaningful changes in PLR are too small to be appreciated by the human eye no matter how expert the examiner is. The highly subjective nature of this manual penlight assessment is well-demonstrated in the medical literature to be inaccurate and imprecise.
Product. A technology that integrates the accessibility of the pen-light exam and the accuracy of a digital pupillometer while being more affordable would be useful. We have developed PupilScreen, a smartphone-based pupillometer, which is optimally positioned to bridge this gap. Like a digital infrared pupillometer it is non-invasive and objectively measures each component of the PLR. However, since the technology is standalone app-based without the need for extra hardware or disposables it can be deployed in a wide range of environments at an affordable cost. In the future, PupilScreen also can extend beyond TBI and stroke to previously untapped markets (the military, collegiate and professional athletics, industrial sites, home, law enforcement) to help determine need for evaluation by healthcare professionals in a timely manner.
Current work. For the current project, we have developed a relationship with collaborators at the University of Missouri-Kansas City and Mosaic Health system in rural Missouri (IRB approval in process as of 12/27/2023). We propose to fund both FDA quality manual technical writing with quality system development (we are Class I, 510k exempt) and concurrent research of our smartphone pupillometry app neuro-biomarker in patients with the acute neurological conditions of traumatic brain injury, acute ischemic stroke, and acute substance intoxication. These conditions are important because they can all confound one another and are not easily distinguished despite requiring very different emergency treatments with mobilization of very different medical resources. We hypothesize that recording of the pupillary light reflex within the rural Missouri healthcare system will enable use of the PLR to differentiate these conditions during triage (e.g., when ambulances need to know whether to send patients to a comprehensive stroke center versus a high-level trauma center versus a local hospital without advanced capabilities). We believe that funding patient-centered research in an ideal use environment (rural emergency care setting) and funding FDA application components will accelerate the development of this practice-changing technology with the potential to impact morbidity and mortality in a variety of patients with acute neurological disease. Please see linked references below for examples of our work on this app so far.
Lynn B. McGrath: I have a unique combination of experience and expertise that will allow me to contribute meaningfully to the proposed work. As an Assistant Professor in neurological surgery at one of the premier neurosurgical institutions in the world (Weill Cornell Medicine), I am responsible for the triage and management of all neurosurgical emergencies, especially traumatic brain injury. I am often the first neurosurgeon to directly assess a new patient and rapidly determine their need for emergency surgery. In making these decisions I am forced to rely on an archaic set of tools to obtain certain physiological data points, an experience which has made strikingly clear both the insufficiency of the tools and their potential for meaningful improvement. My confrontation with this technological void when a resident neurosurgeon at the University of Washington drove me to start Apertur with the goal of developing machine learning based tools on smartphones as the platform for next generation clinical tools. Our research has focused primarily on utilizing the sensors native to smartphones to capture novel audiovisual physiological data, optimize the data for training machine learning algorithms, and to deploy these novel diagnostic capabilities back to the smartphone platform. Our research in pupillometry resulted in the device PupilScreen which is the focus of the current funding proposal. An initial publication demonstrating the potential of PupilScreen generated significant interest from the international press, featured by Newseek, the Washington Post, GeekWire, and The Times and was named one of GE’s “5 Coolest Things On Earth This Week”. I have received funding from Amazon, UW CoMotion and the Washington Research Foundation to support my work in the development of next generation medical technologies through Apertur and was named the Seattle “Health Innovator of the Year” for 2018 and the 2019 Congress of Neurological Surgery “Innovator of the Year”. My position on the frontlines of both emergency neurosurgical care and medical device innovation makes me uniquely qualified to contribute meaningfully to the execution of both the underlying research and the commercialization phases of the proposed work.
Anthony J. Maxin: As a medical student research scientist, I have the research experience, vision, and motivation that will ensure the successful completion of the proposed work. My early clinical research experience in the field of otolaryngology piqued my interest in the development and use of accessible and affordable technological innovations to improve patient access to care and patient outcomes. As an undergraduate at the University of Washington, I helped in the development of a smartphone application for the detection of laryngeal disease by collecting data from and interviewing patients in a busy academic clinical setting. I was also involved in the management of a survey-based assessment of clinician diagnostic accuracy of voice disorders. I subsequently focused and combined my long-standing interest in neuroscience and neurological surgery with my interest in technological innovation by working initially as the leader of a team of medical students for video annotation, database management, and data processing related to the development of a smartphone application for detection of pupillary light reflex abnormalities in patients with traumatic brain injury. I took on extensive responsibility, including IRB and grant writing, developing software including multiple machine learning and data analysis pipeline programs to process our raw data into pupillary light reflex parameters. I also conducted statistical analysis for several abstracts and manuscripts. As a result of my work in both otolaryngology and neurological surgery, I have presented several abstracts and have written several papers related to these projects. I am confident that given my prior experience managing a clinical data collection team in the neurological ICU, emergency department, and a neurology clinic along with my knowledge of data processing and analysis as well as experience with development of machine learning smartphone applications, I have the motivation, dedication, and scientific knowledge to coordinate the success of the proposed work.
that's fine.
https://www.linkedin.com/in/anthony-maxin/
https://www.linkedin.com/in/lynnmcgrath/
https://www.liebertpub.com/doi/abs/10.1089/neu.2022.0516
https://www.sciencedirect.com/science/article/abs/pii/S1052305723004536
https://dl.acm.org/doi/abs/10.1145/3584371.3613049
https://www.frontiersin.org/articles/10.3389/fnins.2022.893711/full
https://dl.acm.org/doi/abs/10.1145/3131896
https://www.liebertpub.com/doi/abs/10.1089/neu.2023.0468
https://scholar.google.com/citations?user=zNyS9ywAAAAJ&hl=en
https://scholar.google.com/citations?user=WwfKIlEAAAAJ
$60,000 ($30,000 to support FDA 510(k) quality manual writing along with FDA-focused final software development steps and FDA submission and $30,000 to support salary and employee benefits for a 0.5 FTE research assistant for 12 months or a 1.0 FTE research assistant for 6 months to do data collection in acute difficult-to-enroll brain injured patient populations in rural Missouri).
No response.
Our chances of success within one year of receiving funding are as follows. We feel that we have a 95% probability of success in writing a quality manual with assistance from a professional technical writer for FDA 510(k) application and having a research team collecting data on brain injured patients in rural St. Joseph, Missouri and surrounding small towns through collaboration with an established team at University of Missouri-Kansas City medical school and Mosaic Health hospital system. The 5% probability estimate of failing is related only to the risk associated with the Institutional Review Board (IRB) approval process at UMKC and Mosaic Health, which may be prolonged once our application is fully reviewed. We have already begun conversations with the IRBs and our collaborators in Missouri to attempt to ensure that we do not have to wait through a prolonged review process before beginning to collect data on brain injured patients.