Using Reconstructed IVUS Images for Coronary Plaque Classification

Abstract

Coronary plaque rupture is one of the principal causes of sudden death in western societies. Reliable diagnostic of the different plaque types are of great interest for the medical community the predicting their evolution and applying an effective treatment. To achieve this, a tissue classification must be performed. Intravascular Ultrasound (IVUS) represents a technique to explore the vessel walls and to observe its histological properties. In this paper, a method to reconstruct IVUS images from the raw Radio Frequency (RF) data coming from ultrasound catheter is proposed. This framework offers a normalization scheme to compare accurately different patient studies. The automatic tissue classification is based on texture analysis and Adapting Boosting (Adaboost) learning technique combined with Error Correcting Output Codes (ECOC). In this study, 9 in-vivo cases are reconstructed with 7 different parameter set. This method improves the classification rate based on images, yielding a 91% of well-detected tissue using the best parameter set. It also reduces the inter-patient variability compared with the analysis of DICOM images, which are obtained from the commercial equipment.

HoliMAb: A holistic approach for Media–Adventitia border detection in intravascular ultrasound

“…The description of tissue properties is obtained extracting a set of textural features from the image in polar coordinates. This set is obtained from previous studies on lumen segmentation (Ciompi et al, 2009) and plaque characterization in IVUS (Ciompi et al, 2010a;Caballero et al, 2007).…”

Fusing in-vitro and in-vivo intravascular ultrasound data for plaque characterization

“…• Band pass filtering: a Butterworth Band Pass (BP) filter of order 10 and frequencies f L = 20 MHz, f H = 60 MHz • Envelope recovering: computed as the absolute value of the Hilbert transform of the RF signal • Logarithmic compression: it improves the visualization of areas with a low signal intensity, as in [24,28] • Digital Development Process: a non-linear radial adjustment of the gain and edge-emphasis process to enhance the tissue visualization [9,24] At the end of this process, a polar representation of an IVUS image is obtained (Fig. 3a).…”

“…Then, we extract a wide set of textural features by applying texture descriptors to the obtained image. As in [9] we apply a bank of 16 Gabor filters [29] and three configurations of Local Binary Patterns [30], thus obtaining 19 textural features. Furthermore, due to the high echo-reflectivity property of calcified tissue, we decided to add two features, called (1) shadow and (2) relative shadow, computed by (1) accumulating the mean grey level value along each radial direction and by (2) multiplying the obtained value by the radial depth in each considered position.…”

“…Several IVUS-based approaches for the automatic assessment of plaque composition have been proposed, mainly based on image texture analysis [7][8][9] as well as processing the raw Radio Frequency (RF) signals [10][11][12][13][14][15]. Some approaches use AutoRegressive Models (ARM) [10][11][12] or the FFT analysis [13,15] to compute the power spectrum from which spectral features are extracted, while others perform a direct measure of the Acoustic Impedance by Plane Wave Born Approximation [14].…”

Classification of Intravascular Ultrasound Images Based on Non-Negative Matrix Factorization Features and Maximum Likelihood Classifier

“…Coronary plaque classification using IVUS image is discussed in [3]. Texture features like LBP, Gabor filters and co-occurrence matrix are extracted.…”