(L-475) Wearable optical peripheral vascular compensation sensor for early detection of postpartum hemorrhage

Introduction:: Postpartum hemorrhage (PPH) is both the leading and most preventable cause of maternal mortality, particularly in low resource settings¹. Alarmingly, within the United States PPH is responsible for 11% of maternal deaths and disproportionately impacts Black women². Prevention of dangerous PPH is largely dependent on early diagnosis and rapid treatment initiation³. However, current methods of PPH detection are insufficient in providing an early indicator in many cases. PPH is diagnosed through visual estimation of blood loss or monitoring of vital signs^3-7. Visual assessment routinely underestimates blood loss and extent of internal bleeding beyond the point of pharmaceutical intervention. Vital sign indicators are delayed due to compensatory changes that stabilize hemodynamics until hemorrhage is massive and high-risk invasive surgical interventions are required⁷. In low-income countries it is crucial to reduce the need for surgical intervention or blood transfusion as resources are limited. There is a clear and urgent need for a globally accessible technology that can aid in diagnosis and treatment monitoring of PPH. Quantitative monitoring of hemorrhage-induced compensatory processes, such as the constriction of peripheral vessels, may provide an early alert for PPH. To this end, we developed a low-cost, wearable optical device that continuously monitors peripheral perfusion via laser speckle flow index (LSFI) to detect hemorrhage-induced peripheral vasoconstriction. LSFI provides a measure of blood velocity, which is related to vasoconstriction and vasodilation, through laser speckle contrast imaging⁸.


Materials and Methods:: The laser speckle imaging system was designed using a Raspberry Pi Camera (V2), with a double lens system, and a 50 mW laser diode. The double lens system consists of a single fixed lens, a second identical inverted lens, and a 0.4 mm sapphire window to allow the device to be in focus when directly in contact with an object. The laser speckle system is controlled by a Raspberry Pi Zero computer and powered by a rechargeable Pi Sugar battery module. The battery is an uninterruptable power supply and can power the device for ~4 hours. The device is Bluetooth enabled, thus completely wireless, and can process and transmit LSFI signal to a receiving device (laptop/tablet) for near real-time data visualization. Speckle video acquisition and frame-by-frame processing can be performed at 80 Hz. The device was first validated by recording LSFI signal from an optical flow phantom with a 1.6 mm channel while pumping fluid through at physiologically relevant velocities 0-10 mm/s. Furthermore, a swine hemorrhage model was used to verify the ability of the device to wirelessly monitor blood loss and treatment in vivo. Swine (n=5) were anesthetized and wore the device on their left wrist (Figure 1). Blood was removed from the swine at a constant rate up to 40% estimated blood volume (EBV) and then crystalloids were reinfused into the swine up to 20% EBV. Vital signs were continuously monitored throughout the procedure, while the wearable device recorded and processed 10 seconds of speckle video every minute.


Results, Conclusions, and Discussions:: In phantom studies, the correlation coefficient between mean LSFI measured with the wearable device and flow velocities was 0.998, indicating sensitivity to changes in blood flow rate within the physiological range. During swine hemorrhage, the average correlation between LSFI and net fluids into the swine was -0.94, with correlation coefficients of -0.90, -0.95, -0.99, -0.90, and -0.98 for each individual case. The correlation between LSFI and %EBV during blood loss was higher and more consistent than correlations between vital signs and net fluids (Figure 2). Swine 1 and swine 4 experienced a vein collapse and blood clot, respectively, during hemorrhage, which were detected as artifacts in our LSFI signal, indicating our device may be able to identify certain vascular events in vivo.

During crystalloid infusion the average correlation between LSFI and net fluids was 0.79, with correlation coefficients of 0.99, 0.91, 0.23, 0.90, and 0.90 for each individual swine. Again, LSFI and net fluids correlated more strongly than LSFI and any of the vital sign measures (Figure 3). The LSFI signal measured from swine 3, unlike any other swine, approached zero towards the end of blood loss and did not increase again following crystalloid infusion. This indicates the swine may have exhausted its compensatory response and was then unresponsive to treatment. Our device has the potential to monitor treatment response and identify the turning point treatment may become ineffective.

The swine hemorrhage model is widely accepted as an effective model for human hemorrhage; however anesthesia likely reduces the compensatory response⁹. In future studies with human patients, we expect to see a greater delay in vital sign changes and a more dramatic vasoconstriction response¹⁰. In summary, we have demonstrated that a custom wearable laser speckle imaging device could rapidly detect hemorrhage-driven peripheral vasoconstriction and resuscitation via crystalloid infusion. This progress provides a framework for a novel low-cost and noninvasive technology that identifies ongoing blood loss when low-cost and accessible interventions are effective, with the hope of reducing the unacceptably high rates of maternal morbidity and mortality caused by hemorrhage in low- and high-resource settings alike.


Acknowledgements (Optional): : We thank Paul Chen, Dr. Joe Culver, Dr. Quing Zhu, Dr. Sam Achilefu, Dr. Sarah England, and Dr. Roxane Rampersad for helpful discussions. We are grateful for funding from Washington University Women’s Health Technologies Collaboration Initiation Grant, and NIH R00HD103954 & K99HD103954.


References (Optional): : 1. L. Say, D. Chou, A. Gemmill, Ö. Tunçalp, A.-B. Moller, J. Daniels, A. M. Gülmezoglu, M. Temmerman, L. Alkema. Global causes of maternal death: a WHO systematic analysis. The Lancet global health 2, e323-e333 (2014).

2. M. J. Tucker, C. J. Berg, W. M. Callaghan, J. Hsia. The Black-White disparity in pregnancy-related mortality from 5 conditions: differences in prevalence and case-fatality rates. Am J Public Health 97, 247-251 (2007).

3. B. Lawton, E. J. MacDonald, S. A. Brown, L. Wilson, J. Stanley, J. D. Tait, R. A. Dinsdale, C. L. Coles, S. E. Geller. Preventability of severe acute maternal morbidity. Am J Obstet Gynecol 210, 557.e551-556 (2014).

4. C. J. Berg, M. A. Harper, S. M. Atkinson, E. A. Bell, H. L. Brown, M. L. Hage, A. G. Mitra, K. J. Moise, Jr., W. M. Callaghan. Preventability of pregnancy-related deaths: results of a state-wide review. Obstet Gynecol 106, 1228-1234 (2005).

5. C. Deneux-Tharaux, M. Bonnet, J. Tort. Epidemiology of post-partum haemorrhage. J Gynecol Obstet Biol Reprod (Paris) 43, 936-950 (2014).

6. S. Miller, F. Lester, P. Hensleigh. Prevention and treatment of postpartum hemorrhage: new advances for low-resource settings. Journal of midwifery & women's health 49, 283-292 (2004).

7. H. L. Nathan, A. El Ayadi, N. L. Hezelgrave, P. Seed, E. Butrick, S. Miller, A. Briley, S. Bewley, A. H. Shennan. Shock index: an effective predictor of outcome in postpartum haemorrhage? BJOG : an international journal of obstetrics and gynaecology 122, 268-275 (2015).

8. M. Ghijsen, T. B. Rice, B. Yang, S. M. White, B. J. Tromberg. Wearable speckle plethysmography (SPG) for characterizing microvascular flow and resistance. Biomedical Optics Express 9, 3937-3952 (2018)

9. V. A. Convertino, N. J. Koons, M. R. Suresh. Physiology of Human Hemorrhage and Compensation. Comprehensive Physiology2021. p. 1531-1574.

10. J. Seagard, F. Hopp, Z. Bosnjak, J. Osborn, J. Kampine. Sympathetic efferent nerve activity in conscious and isoflurane-anesthetized dogs. Anesthesiology 61, 266-270 (1984).