The following information was submitted:
Transactions: INTERNATIONAL JOURNAL of CIRCUITS, SYSTEMS and SIGNAL PROCESSING
Transactions ID Number: 20-684
Full Name: Takanori Koga
Position: Assistant Professor
Age: ON
Sex: Male
Address: Gakuendai, Shunan 745-8585
Country: JAPAN
Tel: +81-834-29-6319
Tel prefix:
Fax: +81-834-29-6319
E-mail address: koga@tokuyama.ac.jp
Other E-mails:
Title of the Paper: High-speed calculation for tissue characterization of coronary plaque by employing parallel computing techniques
Authors as they appear in the Paper: Takanori Koga, Shota Furukawa, Eiji Uchino and Noriaki Suetake
Email addresses of all the authors:
Number of paper pages: 8
Abstract: In recent years, remarkable progress can be seen in the field of computer-aided medical diagnosis of ischemic coronary arterial disease. Intravascular ultrasound (IVUS)-based tissue characterization of coronary plaque is a significant topic in this field. The authors have proposed the multiple k-nearest neighbor (MkNN) classifier for the tissue characterization of coronary plaque in an IVUS B-mode image. Although its characterization performance was highly evaluated, the calculation speed was too slow to use actually in medical practice. The purpose of this study is to accelerate the speed of MkNN classifier aiming for it to be actually used in the medical practice. Recently, some parallel computing techniques on central processing unit (CPU) or on graphics processing unit (GPU) have come into general usage. Especially, the general purpose computation technique on Graphics Processing Unit (GPGPU) has got into the limelight recently. In this study, the calculatio!
n speeds of the MkNN classifier are evaluated for cases of various implementations using the parallel computing techniques. By employing GPGPU technique, it has been confirmed that its speed has been drastically accelerated enough for the practical use.
Keywords: Acute coronary syndromes (ACS), Intravascular ultrasound (IVUS) method, Multiple k-nearest neighbor, Parallel computing, Pattern classification, Pixel classification, Tissue characterization, GPGPU
EXTENSION of the file: .pdf
Special (Invited) Session: Acceleration of the Speed of Tissue Characterization Algorithm for Coronary Plaque by Employing GPGPU Technique
Organizer of the Session: 653-154
How Did you learn about congress:
IP ADDRESS: 202.244.90.11
