Post-Crash Noninvasive Occupant Monitoring for Emergency Response Kevin Joseph and Kristofer D. Kusano, Ph.D. Center of Injury Biomechanics, Virginia Tech Wake Forest University, Blacksburg VA Background CDC studies on traumatic brain injuries(TBI): Risk of death decreases by 25% when a seriously injured patient is treated at a trauma center vs a local hospital Only 27% of 1.5 million TBI go to a trauma center Car crashes are the second leading cause of TBI and leader of TBI deaths stimuli. •Accurately detects unconscious and stable occupants only because of the lack of many GCS motor tests •Has to estimate GCS motor based off other vitals and sensor Weight has to be calibrated based off seat and sitting position for exact occupant weight since voltage conversion yields weight being pushed onto the sensor Heart Rate frequency/signal weak and varied but still valid Non-invasive occupant monitoring shows promise for post- crash injury assessment. Listening Move in seat back and forward Unbuckle and buckle seatbelt Press button on steering wheel 1 2 3 4 5 6 Moto r Makes no movement s Extension to painful stimuli Abnormal flexion to painful Flexion to painful Localizes painful stimuli Obeys commands 1 2 3 4 5 Eye Does not open eyes Opens eyes in response to painful stimuli Opens eyes in response to voice Opens eyes spontaneou sly Verbal Makes no sounds Incomprehensible sounds Utters inappropriate words Confused, disoriente d Oriented , converse normally GCS program is a GUI that will be displayed in a car. It will display different pages that will poise different challenges to test responsiveness and assign a score via different car mechanics Filter pressure signal with a 4 th order Butterworth low pass filter at three Hz: Weight: 0.1Hz; Respiratory: 3Hz; Heart: 10Hz Aim: test for consciousness • Weight • Respiratory Rate • Heart Rate • Glasgow Coma Scale(GCS) i. GCS is used to test the level of TBI and consciousness with a score from 3(Unconscious) to 15(Stable) ii.Composed of Motor, Verbal, and Eye Sections Use these vital signs to help the car diagnose the occupant’s condition in order to speed emergency medical response from the medical responders and/or car. Method: All Vital Signs: MATLAB program that features data acquisition, filtering, and different tests to extrapolate proper value for each vital sign GCS Testing: Self-Coded Java Graphical User Interface(GUI) Biomechanics REU • Virginia Tech School of Biomedical Engineering • National Science Foundation IMARI (Increasing Minority Admissions into Research Institutions) • Oakwood Biology Department • NIH Grant: 1R25GM106994-01 Listening Move in seat back and forward Unbuckle and buckle seatbelt Press button on steering wheel ---> Chronological order of GCS testing ---> Tests eye and verbal responsiveness and control Uses a visual prompt and a microphone • Visual display first • Verbal Prompt if there is no response Tests motor responsiveness and physical control Uses various car mechanics to sense physical response • OCS Seat Sensor • Steering Wheel • Seatbelt, etc. System Overview Abstract/Aim Introduction Respiratory Rate Heart Rate Glasgow Coma Scale Acknowledgements Conclusion Diagnoses Diagnosing runs every time new vitals or GCS score is acquired assessment •Starts with Unconscious Test, then Severe Injury Test, and ending with Minor Injury Test •Has the ability to recognize if vitals have changed or there was a miscalculation and then restart the algorithm via Misdiagnosed diagnoses Uses the previously acquired vitals as metrics to base parameters for each diagnoses • Unstable: pulse not between 50 and 100, weak breaths, light headed and pain or headaches • Undiagnosable: no weight in seat, volatile readings, or misdiagnosed 3 times • Extreme Blood Loss: major weight drop over time, and increased average heart rate variability and pulse from starting average Uses a self adjusted IQR test to remove outliers • Copies the lowest quartile to the opposite side to lower the IQR value • Done because of small dataset and one number can largely extend the outlier range 1 2 3 1 2 3b 3a