Neonatal Intensive Care Unit Clinical Trial
Noninvasive Monitoring and Diagnosis of Physiologic Disturbances in Term and Preterm Newborns Using Video Analysis Techniques
Video recording for term and preterm newborns hosted at the AUBMC Neonate ICU will be collected. The videos will capture movements, skin color changes, positioning, and other features relevant to the diagnosis in question. Recorded video images will be analysed by a computer software using the Eulerian Video magnification technology; then images will be correlated with simultaneously recorded physiological parameters (heart rate, respiratory rate and saturation). Images will be annotated by the clinician. Correlation between the skin coloration differentials and annotated physiological parameters may establish physiological indices. These may be used to extrapolate the existence or absence of disease states.
Recent medical and technological advancements are progressively leading to increased survival rates of term and premature neonates. This project is concerned with detecting early signs of physiological disorders in term and preterm babies based on Eulerian Video Magnification, statistical analysis, and expert knowledge. Video recordings of skin coloration differentials for normal and sick neonates may be used to establish norms for each of the two major populations. Those and possibly other features will be used to develop an automatic contactless, noninvasive monitoring and diagnostic system that may detect early signs of disease. Early detection may lead to early diagnosis and preemptive treatment resulting in better outcome and lower costs of medical intervention. The objective of this project is to adapt the Eulerian Video Magnification for monitoring physiologic changes and then diagnosing potential disturbances in term and preterm neonates of different gestational ages. EVM uses video recordings of magnified skin color signals over time and thus permits analysis of physiological state changes such as heart rate and perfusion. Those changes would then be correlated with particular condition or disease states for the purpose of automatic early diagnosis and alarm issuance. Being contactless and non-invasive, the proposed system would be particularly helpful in vulnerable populations such as sick neonates. It may lead to earlier treatment, improved outcome, and possibly, decreased hospital stay. Parents of infants admitted to the neonatal intensive care unit (NICU) at AUBMC will be informed about the study by their Attending physician. The medical research assistant will approach only interested and potential participants. An informed consent will be obtained prior to enrollment. All infants admitted to the NICU are eligible. Those will be divided into four equal groups: 1) "normal" term, 2) "normal" preterm, 3) "sick" term, and 4) "sick" preterm babies, where a "normal" health state is defined by the medical team as "physiologically stable". The medical team considers "sick" state based on pathological symptoms such as bradycardia, apnea, and hypotension among others. We target collecting video recordings for a total of eighty babies as a pilot number over the period of eight months. Each baby will be recorded for a minimum of twenty four hours. A subset of the eighty babies will be considered as a control group for reference. The control group will be selected based on a normal physiological state excluding babies with sepsis, respiratory distress, and congenital anomalies. ;
|Source||American University of Beirut Medical Center|
|Start date||September 2015|
|Completion date||February 2022|
|Not yet recruiting||