Blaivas, Michael

ABSTRACT Traumatic brain injury (TBI), particularly among military personnel, poses significant challenges in acute management and long-term outcomes. Difficult-to-diagnose autoregulatory and inflammatory changes lead to neurocognitive impairment, anxiety, sleep disturbance, and chronic pain. To optimize readiness and test therapeutics for both diagnostic and future TBI treatments, this feasibility study introduces a novel wearable multimodal platform that integrates photoplethysmography and infra-red (IR) sensing from a camera and a wearable sensor to detect vital signs and then leverages machine learning (ML) to predict the need for acute treatment. This prospective observational study involving 28 healthy volunteers simulated a triage emergency room workflow to test sensor accuracy and mobile app functionality. The wearable device demonstrated high accuracy in pulse rate (mean error 7.3%) and oxygen saturation (mean error −1.2%) compared to standard-of-care devices. Camera-based pulse and respiratory rate, and wearable pulse, oxygen, and temperature sensing were tested in a mock triage setting to simulate activation of light-based treatment upon detection of physiological changes of concern after acute TBI. Preliminary ML models to predict hypotension and pain based on retrospective data achieved 98.4% and 93% accuracy but showed limitations in prospective real-time applications. This study provided insights to prepare for future iterations that will integrate near-IR spectroscopy to detect cerebral oximetry, and quantum dot light-emitting diodes to activate light-based treatment (photobiomodulation) from within a wearable head dressing. The findings underscore the potential of this platform to bridge critical gaps in acute TBI care through early detection and future therapeutic intervention. Further development is needed to address environmental challenges, improve accuracy across diverse populations, and refine the device for real-time clinical use. This innovative approach has the potential to revolutionize TBI care by bridging current gaps in diagnostic and therapeutic management.

Objective: To establish evidence-based guidelines for the use of bedside cardiac ultrasound, echocardiography, in the ICU and equivalent care sites. Methods: Grading of Recommendations, Assessment, Development and Evaluation system was used to rank the “levels” of quality of evidence into high (A), moderate (B), or low (C) and to determine the “strength” of recommendations as either strong (strength class 1) or conditional/weak (strength class 2), thus generating six “grades” of recommendations (1A–1B–1C–2A–2B–2C). Grading of Recommendations, Assessment, Development and Evaluation was used for all questions with clinically relevant outcomes. RAND Appropriateness Method, incorporating the modified Delphi technique, was used in formulating recommendations related to terminology or definitions or in those based purely on expert consensus. The process was conducted by teleconference and electronic-based discussion, following clear rules for establishing consensus and agreement/disagreement. Individual panel members provided full disclosure and were judged to be free of any commercial bias. Results: Forty-five statements were considered. Among these statements, six did not achieve agreement based on RAND appropriateness method rules (majority of at least 70%). Fifteen statements were approved as conditional recommendations (strength class 2). The rest (24 statements) were approved as strong recommendations (strength class 1). Each recommendation was also linked to its level of quality of evidence and the required level of echo expertise of the intensivist. Key recommendations, listed by category, included the use of cardiac ultrasonography to assess preload responsiveness in mechanically ventilated (1B) patients, left ventricular (LV) systolic (1C) and diastolic (2C) function, acute cor pulmonale (ACP) (1C), pulmonary hypertension (1B), symptomatic pulmonary embolism (PE) (1C), right ventricular (RV) infarct (1C), the efficacy of fluid resuscitation (1C) and inotropic therapy (2C), presence of RV dysfunction (2C) in septic shock, the reason for cardiac arrest to assist in cardiopulmonary resuscitation (1B–2C depending on rhythm), status in acute coronary syndromes (ACS) (1C), the presence of pericardial effusion (1C), cardiac tamponade (1B), valvular dysfunction (1C), endocarditis in native (2C) or mechanical valves (1B), great vessel disease and injury (2C), penetrating chest trauma (1C) and for use of contrast (1B–2C depending on indication). Finally, several recommendations were made regarding the use of bedside cardiac ultrasound in pediatric patients ranging from 1B for preload responsiveness to no recommendation for RV dysfunction. Conclusions: There was strong agreement among a large cohort of international experts regarding several class 1 recommendations for the use of bedside cardiac ultrasound, echocardiography, in the ICU. Evidence-based recommendations regarding the appropriate use of this technology are a step toward improving patient outcomes in relevant patients and guiding appropriate integration of ultrasound into critical care practice.