Principal Investigator Richard Cohen
Tissue perfusion, or blood flow in capillaries, is a very important clinical parameter. Fall in tissue perfusion below a critical level results in tissue injury, cell death, and impaired organ function. Even if flow of blood to the tissue is reestablished, there is a risk of ischemia-reperfusion injury, which can lead to sepsis, hepatic cirrhosis, heart failure, and many other critical illnesses.
Powerful drugs are currently available to modify tissue perfusion. These drugs would allow physicians to prevent and limit the extent of disease on different organs (e.g. limit the damage to the heart after myocardial infarct, decrease the effects of a stroke, or decrease morbidity associated with multi-organ failure in the intensive care setting). However, the use of these drugs is hampered by the lack of reliable non-invasive techniques for the measurement of perfusion.
Although diagnostic ultrasound is commonly used for blood flow measurement, it is mainly restricted to the fast-moving blood in large vessels. The signal from blood is often masked by much stronger signal from the movement of tissue (vibrations, movement of arterial walls, etc.). Conventional signal processing methods employed in modern ultrasound scanners simply filter out all slow-motion signals assuming that slow motion corresponds to tissue and fast motion to blood. Obviously, this approach will not work for perfusion measurement as blood flow in capillaries is very slow, comparable to the velocity of tissue movement
We use another approach based on the difference between spatially correlated tissue motion and uncorrelated blood flow. We apply our processing algorithm offline to the raw ultrasound data acquired by a commercial ultrasound scanner with a research interface. We have qualitatively demonstrated that our method is capable of non-invasively measuring perfusion. Our data show a marked decrease in muscle perfusion following the application of a tourniquet and an increase in perfusion following exercise.