ViPS: A novel visual processing system architecture for medical imaging

https://doi.org/10.1016/j.bspc.2017.06.003Get rights and content

Highlights

  • Developed a low-power and high-performance medical visual processing system (ViPS).

  • ViPS takes images with complex data structures from medical imaging interfaces.

  • It manages complex images by using on-chip Specialized Medical Memory.

  • To process the application ViPS uses heterogeneous multi-core processors.

  • ViPS provides easy to use programming model for medical applications.

Abstract

Imaging has become an essential tool in modern medicine science. Numerous powerful platforms to register, store, analyze and process medical imaging applications appear in recent years. In this article, we have designed an advanced visual processing system (ViPS) that stores and processes complex and multi-dimensional medical imaging application. The ViPS provides a user-friendly programming environment and high-performance architecture for data acquisition, registration, storage, analysis and performs segmentation, filtering, and recognition of complex real-time complex and multidimensional medical images or videos. The proposed architecture is highly reliable concerning cost, performance, and power. ViPS is designed and evaluated on a Xilinx Virtex-7 FPGA VC707 Evaluation Kit. The performance of ViPS is compared with the Heterogeneous Multi-Processing Odroid XU3 board and GPU Jetson TK1 Embedded Development Kit with 192 CUDA cores based graphic systems. In comparison to the Heterogeneous Multi-core and GPU based Graphics systems, the results show that ViPS improves 2.4 and 1.4 times system performance respectively, for iridology application. While executing real-time complex images reconstruction at 2× and 1.25× of higher frame rate, the ViPS achieves 15.2× and 5.26× of performance improvement while running various image processing algorithms. The ViPS gets 3.01× and 1.13× of speedup for video processing algorithms and draws 1.55 and 2.27 times less dynamic power.

Introduction

Graphics software programs are growing and are desirable in a vast spectrum ranging from medical science to the gaming technology because they can generate realistic images and enable graphics effects to produce different viewpoints and visual clues to the user. Different software engines [1], [2], [3] are introduced which provide an efficient means of data reuse, analysis and processing for a set of related products. Those software solutions do provide flexibility and re-programmability, but graphics performance is limited by the computation power of graphics devices.

High Performance Computing (HPC) graphics system architectures are now being used in medical imaging for early diagnosis, treatment planning, managing the physical form of the patients and monitoring disease progression. As the biomedical industry tries to achieve the low treatment cost and earlier prediction of disease, the medical imaging equipment takes a critical role in health care. To satisfy the industry demands, the bio-medical enterprise is moving towards high-performance computing designs. As the need for such equipment increases, application specific and high-performance architecture is needed to operate medical applications.

Aligning the medical information collected from various medical types of equipment and to display them using different visual approaches gives more detail knowledge of understanding a disease state and reason. The visual representation uses alignment and registration of the complex and multi-dimensional images. The alignment and registration of complex images having sparse data and control flow are a cumbersome process. Typically a medical machine (e.g. Radiological Imaging) spend 75% of processing time in alignment and registration. For example, a Computed Tomography (CT) imaging application aligns three-dimensional space images with isotropic resolution. While performing image acquisitions, the applications have to keep the anatomical and disease compositions.

Digital Signal Processors (DSPs), Single Instruction Multiple Data (SIMD) Processor, Application Specific Instruction Set Processor (ASIP), Field Programmable Gate Arrays (FPGAs) are also used for medical image processing. Graphical Processing Units (GPUs) are specialized processors with dedicated memory and multiple Stream Multiprocessor (SM) having SIMD support that conventionally performed floating point operations required for rendering graphics. The GPUs are widely used in medical imaging over the last few years, and the GPU programming tools are dramatically evolved and become capable of executing the different scientific algorithm. On the other side, GPUs computing architecture became more powerful; it allows to perform complex and compute intensive techniques which give better results to medical diagnosis and conclusion. Many medical visualization and image-processing techniques use the same algorithms of the disciplines, such as pattern recognition, it allows medical scientist to execute existing algorithms/techniques on GPU architectures for medical imaging.

The medical statistic [4], [5] shows that the early-stage disease prediction e.g. breast, colorectal and lung cancers, etc., can save lives. This requires improving disease identification screening techniques to create a high-quality, three-dimensional reconstruction based on tomographic images. With the increase of medical scanner technology complexity, it requires a high-performance computing architecture for real-time application processing. Existing computing architectures (GPUs, DSPs, Multi-cores, etc.) have hardware design and programming limitation for medical imaging applications. Therefore, a reliable programmable and high-performance application specific architecture is required for real-time image reconstruction applications.

In this paper, we intend to develop a low-power, low-cost, easy to use and high-performance medical graphics architecture called visual processing system (ViPS). ViPS provides a high-performance FPGA-based design which takes complex image/video data having 1D/2D/3D data structures from medical imaging interfaces or stored in the memory, manages them in an on-chip specialized memory and process them using specialized hardware accelerators or multi-core system. The ViPS Medical Application Programming Model reduces the programming effort involved in the manually arrangement of data transfer requests, memory management, input/output peripheral management and achieve the required performance of the medical imaging applications. ViPS provides low-cost and efficient control mechanism that arranges many imaging peripherals and interface with integrated processing units. Specialized medical hardware accelerators are integrated into the design as they consume low power and delivers high-performance. ViPS supports multi-peripherals (camera, display) and processing cores without the backing of an Operating System (OS). While comparing results with the Heterogeneous Multi-core and GPU-based Graphics systems, it shows that ViPS improves 2.4 and 1.4 times system performance respectively while executing the iridology application. The ViPS performs real-time images reconstruction at 2× and 1.25× of higher frame rate and achieves 14.6× to 4.8× of speedup while executing different image processing applications. The ViPS gets 3.01× and 1.13× of speedup for video processing and draws 1.55 and 2.27 times less dynamic power.

The rest of this paper is organized as follows: Section 2 discusses the literature review and Section 3 describes ViPS system. Finally, Section 5 presents the results and Section 6 provides the conclusions.

Section snippets

Related work

Imaging applications for clinical procedures and disease analysis require a high degree of performance and accuracy. Many image processing toolkits and hardware architectures exist for the medical imaging application, but to the best of our knowledge a scalable, portable and high-performance system architecture is required which can be programmed easily.

Lbanez [6] developed an open-source medical imaging toolkit called the Insight Toolkit (ITK). The ITK supports a number multi-platform system

Advanced visual processing system architecture

The section discusses the advanced visual processing system (ViPS) architecture. The section is further categorized into five subsections: Working of ViPS, the Registration System, Memory System, Processing System and Medical Application Toolkit.

Experimental framework

In this section, we describe the FGPA based visual processing system. An Odroid Heterogeneous Multi-core and Jetson TK1 GPU based Graphics Systems are used as baseline systems. The section is further categorized into the Heterogeneous Multi-Core based Graphics System, the GPU based Graphics System, the FPGA based ViPS and the Test Algorithms.

Results and discussion

In this section, we analyze the results of different experiments conducted on ViPS. To evaluate the performance of ViPS, the results are compared with the Heterogeneous Multi-core Graphics System and the GPU based Graphics System. The architectures are connected with different medical image sensing devices. The experimentation is classified into six subsections: System Performance, Image Reconstruction, Image Processing, Video Processing, Dynamic Power Consumption and Programmability,

Conclusion

In this paper, we have proposed and developed a visual processing system (ViPS) for medical imaging applications. The system can register multi-dimensional high resolution and complex images, and executes different processing algorithm in hardware and software. The ViPS system provides efficient data access from medical image equipment that eliminates the on-chip/off-chip bus delays for arranging and gathering data. The data structure of medical image or video is specified in the program memory

References (33)

  • G. Pratx et al.

    3-D tomographic image reconstruction from randomly ordered lines with CUDA

    W. mei Hwu (Morgan Kaufmann, San Francisco, CA)

    (2011)
  • OGRE: Object-Oriented Graphics Rendering Engine....
  • A. Alzaabi

    Tcct: a GUI table comparison computer tool

    Emerging Trends in Computing, Informatics, Systems Sciences, and Engineering

    (2013)
  • Alzaabi Ali

    Tcct: a GUI table comparison computer tool

    Emerging Trends in Computing, Informatics, Systems Sciences, and Engineering

    (2013)
  • R. Siegel et al.

    Colorectal cancer statistics

    CA: Cancer J. Clin.

    (2014)
  • R.L. Siegel et al.

    Cancer statistics, 2016

    CA: Cancer J. Clin.

    (2016)
  • L. Ibanez et al.

    The ITK Software Guide

    (2003)
  • W.J. Schroeder et al.

    Visualizing with VTK: a tutorial

    Comput. Graphics Appl., IEEE

    (2000)
  • I. Wolf et al.

    The medical imaging interaction toolkit (MITK): a toolkit facilitating the creation of interactive software by extending VTK and ITK

    Medical Imaging 2004

    (2004)
  • R. Bakalash et al.

    Medicube: A 3D medical imaging architecture

    Comput. Graphics

    (1989)
  • J.-S. Lee et al.

    An advanced video camera system with robust AF, AE, and AWB control

    IEEE Trans. Consum. Electron.

    (2001)
  • J.-S. Lee et al.

    An advanced video camera system with robust AF, AE, and AWB control

    IEEE Trans. Consum. Electron.

    (2001)
  • D. Jinghong et al.

    Development of image processing system based on DSP and FPGA

    Electronic Measurement and Instruments, 2007. ICEMI’07

    (2007)
  • J.D. Owens et al.

    A survey of general-purpose computation on graphics hardware

    Computer Graphics Forum, vol. 26

    (2007)
  • R.M. Jiang et al.

    FPGA implementation of 3d discrete wavelet transform for real-time medical imaging

    18th European Conference on Circuit Theory and Design, 2007. ECCTD 2007, IEEE

    (2007)
  • S. Coric et al.

    Parallel-beam backprojection: an FPGA implementation optimized for medical imaging

    Proceedings of the 2002 ACM/SIGDA Tenth International Symposium on Field-Programmable Gate Arrays

    (2002)
  • Cited by (0)

    View full text