Effect of angle of incidence on backscatter methods of ultrasonic bone assessment

Ultrasonic backscatter techniques are being developed to detect changes in cancellous bone caused by osteoporosis. Clinical implementation of these techniques may use a hand-held transducer pressed against the body. Variations in transducer angle with respect to the bone surface may cause errors in the backscatter measurements. The goal of this study was to evaluate the sensitivity of backscatter parameters to these errors. Six parameters previously identified as potentially useful for ultrasonic bone assessment were investigated: apparent integrated backscatter (AIB), frequency slope of apparent backscatter (FSAB), frequency intercept of apparent backscatter, normalized mean of the backscatter difference, normalized backscatter amplitude ratio, and the backscatter amplitude decay constant. Measurements were performed on specimens prepared from a polymer open cell rigid foam coated with a thin layer of epoxy to simulate cancellous bone with an outer cortex. Data were collected using a 3.5 MHz transducer for angles of incidence ranging from 0° to 30° relative to the specimen surface perpendicular. AIB and FSAB demonstrated the greatest sensitivity to angle-dependent errors. The source of error was identified as reflection and attenuation losses caused by the cortex. A theoretical model was developed and experimentally validated to predict these losses.