Modeling of Interpolation Methods for Robot Assisted and Freehand Ultrasound Imaging System

摘要

Abstract The 2D ultrasound is one of the popular imaging modality amongst all existing medical imaging methods such as X-Ray, MRI, CT, PET and SPECT. However, 3D ultrasound is also increasingly being used in clinical and image guidance applications such as echo cardiology and obstetrics by robot assisted probe and freehand probe, but 3D ultrasound imaging has some limitations such as lack of suitable interpolation methods that leads to wrong interpretation. To overcome this problem, this study presents a modelling approach for interpolation methods to get the better 3D results for enhanced diagnosis. For this modelling, this research mainly uses four (4) interpolation methods such as conventional Distance Weighted (DW), Improved Distance Weighted (IDW), Conventional Olympic (C-Oly) and Improved Olympic (I-Oly) interpolation methods. In order to evaluate the performance, this study used ten (10) test cases, which are testified by each interpolation method. Each method calculates a value, which would be used to fill-out the empty space during interpolation. Therefore, on the basis of obtained results, it is reported that the I-Oly method is producing more optimal results due to the lowest error rate (2.48%) and performance improvement of 26.23% as compared to other existing methods. The outcome of this research could be imparted to post-graduate level for students undertaking sensor and systems either in the robot assisted or freehand ultrasound imaging field.