Harmonic Imaging: tissues and microbubbles

Agent Detection Imaging (ADI) is a development of colour Doppler in which the directional and intensity information is sacrificed in favour of spatial resolution and depth penetration. It has been the subject of a multi-centre study in which the ability of the method to distinguish between focal benign and malignant liver lesions was assessed in a blinded review. The readers’ ability to diagnose lesions improved consistently from around 60 to almost 80% of cases.

A particular advantage of the SAE and ADI approaches is that the colour layer of the composite image contains only microbubble information while the grey scale layer contains only tissue structure information. The two sets of registered data can be switched on and off or viewed as an overlay according to the particular situation and preference.

The obvious way to separate the second (higher) harmonics from the fundamental frequencies is to apply a frequency filter to reject the fundamental and select the second harmonic for image formation. Filtered harmonics works quite well but suffers from the disadvantages of a narrower bandwidth (this reduces spatial resolution) and reduced sensitivity (because the transducer is operating away from its optimum sensitivity). Also, because of minor overlap between the skirts of the fundamental and second harmonic frequency responses, it is impossible to separate these two signals completely without such severe filtering that sensitivity is seriously compromised, so tissue and microbubble harmonics are inextricably mixed. This filtering approach is rarely used in current microbubble modes though some scanners use it for tissue harmonics.

An alternative way to extract the non-linear signals is to use another multi-pulse technique, the phase inversion mode. Here two pulses are sent down each line, the second being of opposite phase from the first. If there is a completely linear response (either from tissue or from microbubbles) the returning signals are mirror images of each other and cancel when they are added. Any non-linearity leaves a residual signal and this can be used to form images. This method gives high resolution images because the entire band width of the original pulses has been used, unlike the coarser images produced by the frequency filtering method referred to above. For this reason it’s sometimes called wide band harmonics. PIM is used for tissue harmonics. For microbubble studies both tissue conduction and microbubble reflection harmonics are registered and they are indistinguishable. Thus the final contrast image is a mixture of tissue and microbubble harmonics.