Self-organizing maps
Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
Hi-index | 0.00 |
Tissue characterization of plaque is very important for a diagnosis of the acute coronary syndromes. The conventional methods however can not perform a good characterization, because the features employed for characterization are not particular for each tissue. In this paper, we propose a novel tissue characterization method by using an adaptive subspace self-organizing map (ASSOM). ASSOM can produce various features from an intravascular ultrasound (IVUS) radiofrequency signal. The acquired features are suitable for the tissue characterization, because the overlap of the distributions of the acquired features is much smaller than the one, e.g., by an integrated backscatter (IB) analysis or by a traditional Fourier spectrum analysis. A tissue is characterized by using the statistical information of the features acquired by ASSOM. Through the application to the tissue characterization of the real IVUS signal, the performance of the proposed method has been verified by comparing it with the conventional methods.