TY - GEN
T1 - Fast reconstruction of quantitative tissue elasticity image based on modified 3-D finite-element model
AU - Yamakawa, Makoto
AU - Shiina, Tsuyoshi
AU - Matsumura, Takeshi
AU - Mitake, Tsuyoshi
PY - 2005
Y1 - 2005
N2 - For tissue elasticity imaging, strain imaging technique is being used in clinical tests [3], However, strain image corresponds to visualizing the qualitative information about tissue hardness. Therefore, in this study, we propose a method for reconstructing tissue hardness as quantitatively as possible from ultrasound data using 1-D array ultrasonic probe. In this method, we use a modified 3-D finite-element model. Using this model, it becomes possible to estimate Young's modulus only from the 2-D strain distribution. Moreover, since we can use the more realistic 3-D elastic equations in the modified 3-D model, it is possible to estimate Young's modulus more quantitatively. We compared the performance of the proposed method with other methods based on the 1-D model, 2-D model and 3-D model. Consequently, although the proposed method is comparable to the 3-D model method in accuracy, this method has 4.0 times precision better than the 1-D model method, 3.0 to 1.5 times precision better than the 2-D model method. Moreover, the proposed method can attain high-speed processing about 170 times faster than the method using 3-D model.
AB - For tissue elasticity imaging, strain imaging technique is being used in clinical tests [3], However, strain image corresponds to visualizing the qualitative information about tissue hardness. Therefore, in this study, we propose a method for reconstructing tissue hardness as quantitatively as possible from ultrasound data using 1-D array ultrasonic probe. In this method, we use a modified 3-D finite-element model. Using this model, it becomes possible to estimate Young's modulus only from the 2-D strain distribution. Moreover, since we can use the more realistic 3-D elastic equations in the modified 3-D model, it is possible to estimate Young's modulus more quantitatively. We compared the performance of the proposed method with other methods based on the 1-D model, 2-D model and 3-D model. Consequently, although the proposed method is comparable to the 3-D model method in accuracy, this method has 4.0 times precision better than the 1-D model method, 3.0 to 1.5 times precision better than the 2-D model method. Moreover, the proposed method can attain high-speed processing about 170 times faster than the method using 3-D model.
UR - http://www.scopus.com/inward/record.url?scp=33847124399&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33847124399&partnerID=8YFLogxK
U2 - 10.1109/ULTSYM.2005.1603208
DO - 10.1109/ULTSYM.2005.1603208
M3 - Conference contribution
AN - SCOPUS:33847124399
SN - 0780393821
SN - 9780780393820
T3 - Proceedings - IEEE Ultrasonics Symposium
SP - 1763
EP - 1766
BT - 2005 IEEE Ultrasonics Symposium
T2 - 2005 IEEE Ultrasonics Symposium
Y2 - 18 September 2005 through 21 September 2005
ER -