Kieran Allan
Research Mentor(s): Mark Burns
Mentor Department: Chemical Engineering
Authors: Sanaz Habibi, Kieran Allan
Session: Session 4 (1:00pm – 1:50pm)
Presentation Type: Poster 96
Abstract
Traumatic brain injury is one of the leading causes of death and disability across the world, being induced in warzones, in car crashes, and even in sports. Due to the current devices like the i-STAT (Abbott Laboratories, IL, USA) needing centrifuged blood and being large in size, the development of a smaller, commercially available credit-card-sized chip that can use whole blood to quickly detect the presence of biomarkers for TBI was created. The device measures glial fibrillary acidic protein (GFAP) biomarkers ranging from 0 to 10000 pg/mL, which is released into the bloodstream when TBI occurs, using beads that have been covered in antibodies to detect the protein. This allows for the detection and intensity of the TBI to be measured. The results from this stage of development were favorable, and new mechanisms are hoping to be able to improve its performance even further. As artificial intelligence and machine learning have begun to grow in popularity and efficiency, their role in the chip is currently being analyzed. Using machine learning, each bead in the chip will be analyzed, instead of a “summary†of all the beads being used to make a prediction. This will result in more accurate forecasts than the current optical signaling method. Using Python, R, and machine learning, many different algorithms are being tested to develop the most efficient and accurate way to detect the TBI biomarkers in the blood. Current and further work on this project will concern improving the algorithms to make the device more optimized, strengthening its validity in the medical world.
