On the use of data fusion in ultrasonic imaging

This paper presents a new approach to fuse the fundamental ultrasonic image and second harmonic ultrasonic image using the wavelet transform. When images are merged in a wavelet space, we can process frequency ranges differently. For example, high frequency information from one image can be combine with lower frequency information from another to perform edge enhancement. Wavelet efficiently separates an image into different subspaces that preserve the features of the original images after fusion. Feature classification information can be directly applied to the coarse approximation of the wavelet decomposed image without segmenting the original image. The proposed method reduces artefacts and enables physicians to make better diagnoses, which is robust in processing noisy images.