DID YOU KNOW? DTIC has over 3.5 million final reports on DoD funded research, development, test, and evaluation activities available to our registered users. Click HERE
to register or log in.
Use of Advanced Machine-Learning Techniques for Non-Invasive Monitoring of Hemorrhage
ARMY INST OF SURGICAL RESEARCH SAN ANTONIO TX
Pagination or Media Count:
Hemorrhagic shock is a leading cause of death in both civilian and battlefield trauma. Currently available medical monitors provide the capabilities to measure standard vital signs that are often imprecise, subjective, and inconsistent. More important, the appearance of hypotension and other signs and symptoms of shock represent a point in time when it may be too late to apply effective life saving interventions. The resulting challenge is that early diagnosis is difficult because hemorrhagic shock is first recognized by late-responding vital signs and symptoms. The solution to this dilemma is to identify physiological signals that provides the best early indicators of blood volume loss and impending circulatory failure. We hypothesized that state-of-the-art machine learning techniques when integrated with novel non-invasive monitoring technologies could detect subtle, physiological changes in conscious, healthy humans who underwent progressive reduction in their central blood volume. Methods We exposed 28 healthy humans to progressive reductions in central blood volume using lower body negative pressure LBNP as a model of hemorrhage until the onset of hemodynamic decompensation. Continuous, non-invasively measured hemodynamic signals e.g., ECG, blood pressures, stroke volume were used for the development of machine-learning algorithms. Accuracy estimates were obtained by building models using 27 subjects and testing on the 28th. This process was repeated 28 times, each time using a different subject. Results Our method was 96.5 accurate in predicting the amount of central blood volume reduction i.e., level of LBNP. The correlation between predicted and actual LBNP level for hemodynamic decompensation was 0.89. Conclusion Machine modeling can accurately identify loss of central blood volume and predict the point at which an individual will experience hemodynamic decompensation onset of shock
APPROVED FOR PUBLIC RELEASE