We present a mathematical model to analyze transadmittance data for the detection of breast cancer using TransScan TS2000 commercial system. The model was constructed based on the assumption that a lesion exists near the surface of a breast region. The breast region that is considered as a background is assumed to be homogeneous at least near the surface where we attach a planar array of electrodes. Based on the model, we developed a lesion estimation algorithm utilizing single- or multifrequency transadmittance data. The approximate ratio of two conductivity values for the lesion and background needs to be known to estimate the size of the lesion even though the location estimate does not require this ratio. From the results of numerical simulations with added noise, we suggest better ways of interpreting TS2000 transadmittance images for the detection of breast cancer with improved accuracy. Since this study provides a rigorous mathematical modeling of TS2000 commercial system, it will be possible to apply the technique to lesion estimation problems based on more realistic models of breast regions in future studies.
Bibliographical noteFunding Information:
Manuscript received November 25, 2002; revised February 29, 2004. This work was supported by the Korea Science and Engineering Foundation under Grant R11-2002-103. The work of H. Ammari was supported by ACI Jeunes Chercheurs(0693) from the Ministry of Education and Scientific Research, France. Asterisk indicates corresponding author.
All Science Journal Classification (ASJC) codes
- Biomedical Engineering