Affiliations: Medical Image Processing Group, Institute of
Automation Chinese Academy of Sciences, Beijing, China
Note: [] Corresponding author: Jie Tian, Medical Image Processing Group,
Institute of Automation Chinese Academy of Sciences, P.O. Box 2728, Beijing,
China. Tel.: +8610 82628760; Fax: +8610 62527995; E-mail:
tian@ieee.org,jie.tian@ia.ac.cn
Abstract: Various types of advanced imaging technologies have significantly
improved the quality of medical care available to patients. Corresponding
medical image reconstruction algorithms, especially 3D reconstruction, play an
important role in disease diagnosis and treatment assessment. However, these
increasing reconstruction methods are not implemented in a unified software
framework, which brings along lots of disadvantages such as breaking connection
of different modalities, lack of module reuse and inconvenience to method
comparison. This paper discusses reconstruction process from the viewpoint of
data flow and implements a free, accelerated, extensible Unified Reconstruction
Software Framework (URSF). The software framework is an abstract solution that
supports multi-modal image reconstruction. The goal of this framework is to
capture the common processing work flow for different modalities and different
methods, make the development of reconstruction for new devices much easier,
and implement a set of popular reconstruction algorithms, so that it is
convenient for researchers to compare against. The overall design and certain
key technologies are introduced in detail. Presented experiment examples and
practical applications commendably demonstrate the validity of this
framework.
Keywords: Unified Reconstruction Software Framework, work flow, CT, freehand 3D ultrasound imaging, acceleration schemes
