Signal Processing Techniques for Improving B-Mode Echoencephalography

Abstract

In spite of many previous research efforts, B-mode echoencephalography is not used clinically for adults because of the poor quality of the images obtained with current, techniques. The intervening presence of the skull results in low sensitivity, reverberations, distortions in the range and lateral dimension, and, most importantly, poor lateral resolution (White, 1970). Lateral resolution, which can be described by the system point spread function (PSF), is a function of the relative phase variation introduced across the transducer aperture by the presence of the skull. A technique has been described (Phillips, et al., this volume) whereby this phase aberration can be measured by a secondary transducer and compensated for in a phased array imaging system, thus restoring the PSF to its unaberrated condition. An alternate technique will be described here which requires no such a priori knowledge of the skull phase variation. This is accompushed by optimizing the transducer frequency and size, and by signal processing the returning echoes. It has been shown (Phillips, et al., this volume) that if one accurately measures the relative phase variation across the transducer due to the skull, the point spread function can be calculated to a close approximation. A simple but accurate model of the skull predicts that the phase variation, $\Delta \phi$, is a linear function of skull thickness, y, and ultrasound frequency, f.