In vivo ultrasonographic exposimetry: Human tissue–specific attenuation coefficients in the gynecologic examination

Objective: The purpose of the current study was to determine in vivo, tissue-specific ultrasonic attenuation coefficients for each of the tissue layers comprising the anterior abdominal wall, uterus, and vagina with use of a quantitative multilayer tissue model. We wanted to validate the “homogeneous” tissue model-based Food and Drug Administration derating factor of 0.3 dB/cm-MHz applied to obstetric-use ultrasonography systems. Study Design: With use of a 3.0-MHz mechanical sector scanner and our previously tested exposimetry equipment, we obtained a set of at least 5 separate acoustic pressure waveforms from each test subject by placing a calibrated 7-element linear-array hydrophone in the anterior vaginal fornix while she was undergoing transabdominal ultrasonography. Corresponding sets of reference in vitro acoustic pressure waveforms were also recorded for each test subject in a 37°C water bath. All linear measurements of individual layer thicknesses and total distances were made on-line with use of electronic calipers. A set of multiple and independent insertion loss values, denoted IL n , was calculated for path n between the abdominal surface and the hydrophone from n sonograms for each test subject. Each tissue layer type was identified and its thickness along each path n was measured. The thickness of tissue type m along path n was denoted by d nm . The only unknown quantities left were the attenuation coefficients A m of each of the m tissue layers for that test subject. The overestimated set of equations d nm A m = IL n was solved for A m with use of a nonnegative least-squares solution technique. Results: With use of data from 162 independent insertion loss estimate paths, the overall tissue-specific attenuation coefficients for each of the tissue layer types, expressed as mean value ± SD, were 2.3 ± 1.5 dB/cm-MHz for the skin and subcutaneous layer, 3.1 ± 2.5 dB/cm-MHz for skeletal muscle, 0.6 ± 0.5 dB/cm-MHz for myometrium, and 3.6 ± 2.7 dB/cm-MHz for the vaginal wall. The overall insertion loss assuming the “homogeneous” tissue model was 0.7 ± 0.3 dB/cm-MHz. Conclusions: We have determined the specific ultrasonic attenuation coefficients for each of the tissue layers comprising the anterior abdominal wall, uterus, and vagina and validated the Food and Drug Administration derating factor of 0.3 dB/cm-MHz applied to obstetric use ultrasonography systems. Of all the models proposed, the “homogeneous” tissue model appears to be the best model