MULTIDIMENSIONAL HEART IMAGING WITH ULTRASOUND

我们最近成功地制作了三维阴影

We have recently successfully produced three-dimensional shaded surface ultrasound images of the left ventricular chamber of a human using a new method of transducer staqblization and rotation. Daa for producing multidimensional volumetric images were acquired by incrementally rotating a sector scanner transducer through 180% about its central beam centered at an appropriate acousic window. Fifty angles of view were recorded for subsequent selection by gating with the ECG and respiration. Sctor scan images were then selected from each view at appropriate points in the heart and respiratory cycles and inserted into a three-dimensional volume. The resulting imates promise new methods of analyzing complex dynamic geometries of the heart and great vessels using advanced workstation software. This proposal is a collaborative effort between one of the largest, most experienced echo cardiography laboratories and an active, widely knwon ultrasonic research laboratory. The Biodynamics Research Unit, of which the Ultrasound Laboratory is a part, has been developing multidimensional imaging devices (such as the Dynamic Spatial Reconstructor) and software for more than 20 years. One major result from this effort has been ANALYZE a widely used Mayo-developed multidimensional data nalaysis package. It is this vast experience in multidimensional software and pioneering efforts in clinical echocardiography that provide the expertise for this proposed reseach. The proposal consists of four aims. First, we will refine our methods of multidimensional scanning by developing methods of transducer stabilizaation, image synchronization, and image digitization and storage (on digital cassette tape). Second, we will evaluate the fidelity of the multidimensional images and calculated parameters using data obtained from fresh fixed hearts in a water tank, anesthetized animals (also scanned on the DSR), and selected volunteers and patients. Third, we will develop advanced software for cliniclly specific analysis and measurement of myocardial structure and function from the ultrasound vol.umetric data. Fourth, using ROC analysis, we will test the accuracy and clinical usefulness of the data acquisition and analysis workstation concept using specifically selected cardiac lesions proven clinically or surgically in a large number of patients.