Conjugate gradient and regularized Inverse Problem-Based solutions applied to ultrasound image reconstruction

Image reconstruction based on inverse problems has the potential for providing good image quality in ultrasound imaging. In this paper, two Inverse Problems-Based (IPB) methods, i.e., Regularized Least Squares (RLS) and the iterative Conjugated Gradient (CG), are evaluated in comparison to the traditional Delay-And-Sum (DAS) Beamrfoming method. It is shown an improvement from 0:36mm to 0:2mm for the axial resolution when comparing the traditional DAS method with the IPB methods, showing the high potential of the method for improving image quality. Similarly, an improvement from 0:5mm to 0:2mm in lateral resolution was found. Further evaluation need to be carried out using data from actual ultrasound systems when noise and speckle is expected.