Sensorless Reconstruction of Unconstrained Freehand 3D Ultrasound Data

Abstract

Freehand 3D ultrasound can be acquired without a position sensor by finding the separations of pairs of frames using information in the images themselves. Previous work has not considered how to reconstruct entirely freehand data, which can exhibit irregularly spaced frames, intersecting frames, non-monotonic out-of-plane probe motion and significant in-plane motion. This paper presents reconstruction methods that overcome these limitations and are able to robustly reconstruct unconstrained freehand data. The methods are assessed on freehand data sets and compared to reconstructions obtained with a position sensor.

Sensorless Reconstruction of Freehand 3D Ultrasound Data

“…Other scanning protocols are also possible, for example axial rotation, where there will certainly be intersections. Reconstructing these data sets is the subject of related work [10].…”

Probabilistic Speckle Decorrelation for 3D Ultrasound

“…Assuming that the relationship between elevational separation and the nominal correlation coefficient ρ 0 is one-to-one, as depicted by the decorrelation curves used in [4,13,6], p(ρ|δ, q) = p(ρ|ρ 0 , q). The statistical variability of the sample correlation coefficient depends on ρ 0 , and on the statistical distributions of X and Y .…”

“…In this paper, it is assumed that: (1) there is no in-plane probe motion or the images have been correctly aligned by some registration procedure; (2) the frames do not intersect; (3) the motion of the probe in the elevational direction is monotonic; (4) the images are made of fully developed speckle. These assumptions may be relaxed through the use of complementary algorithms [5,6] as an add-on to the proposed drift-reduction scheme. They are adopted here for simplicity, and in the experimental protocol to minimise the influence of sources of error outside the control of the proposed method in the analysis of results.…”

“…In-plane probe motion can be estimated through standard image registration techniques. This paper focuses on the estimation of out-of-plane motion from the elevational decorrelation of US speckle between image frames [4,13,10,3,6].…”

“…Recent work [6] has attempted to reduce the impact of measurement errors by averaging independent interleaved reconstructions using only correlation measurements lying on the "well-behaved" (i.e. linear) portion of the decorrelation curve.…”