|
|
|||||
|
ClearView
|
|||||
|
Advantages of ultrasound as diagnostic tool are well known: it is a safe, low cost, real-time and portable medical imaging technique, but some drawbacks of ultrasound imaging limit its use. A major and well-known problem is the so called speckle and noise, which significantly affects human interpretation of the images. As a result, ultrasound diagnosis demands a high level of operator experience. Speckle in ultrasound imaging (and all coherent imaging systems) is caused by the interference of energy from randomly distributed scatters, too small to be resolved by the imaging system. Speckle degrades both the spatial and contrast resolution in ultrasound images and thereby reduces the diagnostic value of the images. The intent of speckle reduction is to remove the distracting speckle pattern without reducing the detail in the ultrasound image, in other words, to make ultrasound images less granular and easier to "read".
Designers over the world have worked on methods to reduce or eliminate speckle. Some of these well-known techniques, like temporal averaging (frame averaging) and post-processing filtering, have been used in previous versions of Echo Blaster scanners and UltraView software, designed by Telemed. Each of these techniques has drawbacks: • Temporal averaging reduces real frame rate, because output frame is a result of superposition of the up to eight acquired frames. Moving tissues looks blurry if a high number of frames are used. • Result of median filtering ( “smooth” or “smooth more” mode in Echo Blaster Pro and Echo Wave software) is lost of detail resolution, despite that readability of processed image is better then original image. Method Leveraging the computational power modern personal computers in combination with the open architecture on the Echo Blaster family, Telemed recently introduced software plug-in for real-time speckle reduction called ClearView. This plug-in works together with Echo Wave software. High-speed algorithm (see image below) analyses ultrasound image features low level features like edges and lines and higher level image representations such as texture, regions, object boundaries, objects themselves, relations between objects and then averages or emphasizes the image based on the results of this comparison.
Results Initial clinical trials demonstrate improvement the contrast resolution as well as overall clarity with no perceptible loss of frame rate or detail resolution. |
|||||
![]() |
![]() |
||||
| Left image: not processed image | Right image: processed with ClearView technology | ||||