Automatic Measurement of Human Subcutaneous Fat with Ultrasound

• Published in: IEEE transactions on ultrasonics, ferroelectrics, and frequency control Vol. 56; no. 8; pp. 1642 - 1653
• Main Authors: Ng, J., Rohling, R., Lawrence, P.D.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.08.2009; Institute of Electrical and Electronics Engineers; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Abdomen - anatomy & histology; Abdomen - diagnostic imaging; Acoustic signal processing; Acoustics; Adult; Algorithms; Anthropometry; Arm - anatomy & histology; Arm - diagnostic imaging; Biological and medical sciences; Biometry; Body fat; Data mining; Exact sciences and technology; Female; Fourier transforms; Frequency measurement; Fundamental areas of phenomenology (including applications); Humans; Image Processing, Computer-Assisted - methods; Investigative techniques, diagnostic techniques (general aspects); Male; Medical sciences; Miscellaneous. Technology; Physics; Radio frequency; RF signals; Skinfold Thickness; Subcutaneous Fat - anatomy & histology; Subcutaneous Fat - diagnostic imaging; Thickness measurement; Thigh - anatomy & histology; Thigh - diagnostic imaging; Ultrasonic imaging; Ultrasonic investigative techniques; Ultrasonic variables measurement; Ultrasonography - methods; Human; Averaging method; Visualization; Correlation; Fourier transformation; Acoustic measurement; Adipose tissue; Probabilistic approach; Grey level image; Threshold detection; Subcutaneous; Speckle; Fourier analysis; Radiofrequency; Modeling; Thickness measurement; Multiple view
• ISSN: 0885-3010; 1525-8955
• DOI: 10.1109/TUFFC.2009.1229

This paper presents an approach to measure human subcutaneous fat thickness automatically using ultrasound radio frequency (RF) signals. We propose using spatially compounded spectrum properties extracted from the RF signals of ultrasound for the purpose of fat boundary detection. Our fat detection framework consists of 4 main steps. The first step is to capture RF data from 11 ultrasound beam angles and at 4 different focal positions. Second, spectrum dispersion is calculated from the local spectrum of RF data using the shorttime Fourier transform and moment analysis. The values of the spectrum dispersion are encoded as gray-scale parametric images. Third, averaging is used to reduce speckle noise in the parametric image and improve the visualization of the subcutaneous fat layer. Finally, we apply Rosiniquests thresholding and random sample consensus boundary detection to extract the fat boundary. Our method was applied on 36 samples obtained in vivo at the suprailiac, thigh, and triceps of 9 human participants. In our study, high correlations between the manual and automatic ultrasound measurements (r > 0.7 at all body sites), and between the skinfold caliper and automatic ultrasound measurements (r > 0.7 at all body sites) were observed.