Tomographic image reconstruction from dual modality ultrasound and electrical impedance data

Electrical resistance tomography (ERT) and ultrasound reflection tomography (URT) have shown some success in the detection of breast tumours. However, the physics behind these tomographic methods are classed as 'soft fields' and suffer from the fact that the interaction of the sensing field with the material distribution nonlinearly depends on the actual distribution. This leads to limited resolution and accuracy of the reconstructed images. Structures like tumours and cysts need to be reliably detected in biomedical imaging but often have dimensions in the order of the system resolution. In this paper the use of URT-EIT dual modality tomography is proposed to improve the detection efficiency of small tissue anomalies. ERT is mostly sensitive to the bulk properties of tissue, while URT responds to boundaries between different types of tissue. The complementarity of the data produces richer information about the imaging region than possible with a single sensing modality. Dual modality image reconstruction is achieved by minimising a cost functional comprising contributions from both data sets. Simulations suggest that the detection of small masses can be considerably improved.