(D-145) Relationship between temperature and ultrasound image echogenicity in phantom meat models

 Duchenne muscular dystrophy (DMD) and spinal muscular atrophy (SMA) are muscle wasting diseases that cause fibrosis and intramuscular fat (IMF) accumulation within muscles.[1], [2] Ultrasound is an excellent non-invasive and accessible method to measure fat and fibrosis content through image echogenicity.[3] However, the direct relationship between echogenicity and IMF content is not well understood, therefore using meat phantoms to describe this relationship would aid in the measurement of disease progression. In order to characterize this relationship, we must first understand the impacts of temperature on echogenicity. It has been demonstrated that ultrasound can be used to measure temperature in tissue and blood via echo shifts due to changes in thermal expansion and speed of sound.[4], [5] Current work has shown that echogenicity increases with temperature. The goal of this study is to develop a relationship between temperature and echogenicity in meat phantoms, and to use this relationship to further investigate the relationship between fat fraction and echogenicity in meat phantoms.

Ground beef with fat contents of 7% and 23% were used as phantoms. To image the samples, ground meat was placed in a glass dish to hold its shape and a straw, filled with ultrasound gel, was inserted into a middle section of the sample as a depth marker for imaging. Samples were first imaged straight out of the refrigerator, at 6.7oC and then subsequently warmed in a ~40oC water bath to 6.7, 10.8, 14.7, 18.5, 23.5, 26, 28, 30, 32, 34, 36, 38, and 40oC, as measured with a meat thermometer. Two samples of each fat content were used and imaged at each temperature. Images of each specimen were obtained using a 40mm linear transducer (Telemed Medical Imaging, Milan, Italy). Echogenic properties were measured and analyzed using a custom Matlab code. Average echogenicity was determined by defining a region of interest and averaging the pixel intensities within the region. Average echogenicity at each corresponding temperature was plotted in Matlab and a linear fit was done in Matlab.

Results: The data showed a positive relationship between sample temperature and echogenicity, with echogenicity increasing with temperature. Echogenicities ranged from approximately 33 pixel intensity for cold temperatures, 7% fat meat to 78.46 pixel intensity for near body temperature 27% fat meat. All meat samples showed an increase in echogenicity of at least 24 points from 6.7oC to 40oC. When fit with a linear curve, the 7% fat samples had equations of echogenicity = 0.6919T+29.1536 with an R2 of 0.9107, and echogenicity = 0.8235T+24.4865 with an R2 of 0.9102 while the 27% fat samples had equations echogenicity = 0.9654T+32.6355 with an R2 of 0.8596 and echogenicity = 0.8422T+28.3985 and an R2 of 0.8349, as seen in Figure 1.

Discussion: The increase in echogenicity with both temperature and fat content are consistent with the literature.[3]–[5] Echogenicity values around body temperature (37oC) are similar to patient populations measured in our previous work and in the literature.[3] The findings of this study may also cause reconsideration of other techniques that use echogenicity to measure muscle parameters that may be sensitive to temperature such as glycogen stores where muscles may be imaged at different temperatures pre and post exercise.[6] By characterizing the strong linear relationship between echogenicity and temperature, we can define equations to correct the echogenicity measured in samples not at body temperature. This study was limited to only using ground meat, but now that we see that a clear relationship exists between temperature and echogenicity, we can extend the study to full cuts of meat to further validate the relationship. 

Conclusion: The knowledge that temperature affects tissue echogenicity will help us better design and use phantoms that can more accurately model human muscle, since current phantoms are artificial and commonly used at room temperature. We can apply these findings to more accurately develop the relationship between magnetic resonance image muscle fat fraction and ultrasound echogenicity, and thus better understand the effects of non-contractile material heterogeneity on muscle function using techniques such as medical image based finite element modeling.

[1] A. E. H. Emery, F. Muntoni, and R. C. M. Quinlivan, Duchenne Muscular Dystrophy. Oxford, UNITED KINGDOM: Oxford University Press, 2015. Accessed: Jan. 25, 2023. [Online]. Available: https://ebookcentral.proquest.com/lib/uva/detail.action?docID=1962957

[2] A. D’Amico, E. Mercuri, F. D. Tiziano, and E. Bertini, “Spinal muscular atrophy,” Orphanet J. Rare Dis., vol. 6, no. 1, p. 71, Nov. 2011, doi: 10.1186/1750-1172-6-71.

[3] S. Pillen and N. van Alfen, “Skeletal muscle ultrasound,” Neurol. Res., vol. 33, no. 10, pp. 1016–1024, Dec. 2011, doi: 10.1179/1743132811Y.0000000010.

[4] J. C. Beitler, B. Sigel, J. Machi, and J. R. Justin, “The effects of temperature on blood flow ultrasonic echogenicity in vitro,” J. Ultrasound Med. Off. J. Am. Inst. Ultrasound Med., vol. 2, no. 12, pp. 529–533, Dec. 1983, doi: 10.7863/jum.1983.2.12.529.

[5] D. Liu and E. S. Ebbini, “Real-Time 2-D Temperature Imaging Using Ultrasound,” IEEE Trans. Biomed. Eng., vol. 57, no. 1, pp. 12–16, Jan. 2010, doi: 10.1109/TBME.2009.2035103.

[6] D. C. Nieman, R. A. Shanely, K. A. Zwetsloot, M. P. Meaney, and G. E. Farris, “Ultrasonic assessment of exercise-induced change in skeletal muscle glycogen content,” BMC Sports Sci. Med. Rehabil., vol. 7, no. 1, p. 9, Dec. 2015, doi: 10.1186/s13102-015-0003-z.