Discrete wavelet and modified PCA decompositions for postural stability analysis in biometric applications

The aim of this study is to compare the Discrete wavelet decomposition and the modified Principal Analysis Component (PCA) decomposition to analyze the stabilogram for the purpose to provide a new insight about human postural stability. Discrete wavelet analysis is used to decompose the stabilogram into several timescale components (i.e. detail wavelet coefficients and approximation wavelet coefficients). Whereas, the modified PCA decomposition is applied to decompose the stabilogram into three components, namely: trend, rambling and trembling. Based on the modified PCA analysis, the trace of analytic trembling and rambling in the complex plan highlights a unique rotation center. The same property is found when considering the detail wavelet coefficients. Based on this property, the area of the circle in which 95% of the trace’s data points are located, is extracted to provide important information about the postural equilibrium status of healthy subjects (average age 31 ± 11 years). Based on experimental results, this parameter seems to be a valuable parameter in order to highlight the effect of visual entries, stabilogram direction, gender and age on the postural stability. Obtained results show also that wavelets and the modified PCA decomposition can discriminate the subjects by gender which is particularly interesting in biometric applications and human stability simulation. Moreover, both techniques highlight the fact that male are less stable than female and the fact that there is no correlation between human stability and his age (under 60).

Fast Encoding-Decoding of 3D Hyperspectral Images Using a Non-Supervised Multimodal Compression Scheme

We introduce in this paper an extension of the Multimodal Compression technique (MC) for the purpose of coding hyperspectral image sequences. The main idea requires few steps, namely: (1) reducing the size of the sequence by inserting smooth images containing less information into the remaining images of the same sequence, (2) then coding the new compacted sequence using 3D-SPIHT algorithm. In this new scheme, called MC-3D-SPIHT, the insertion is achieved only in the contour of each image, according to a non-supervised way, so that one can preserve the Region of Interest (ROI) quality. For this purpose, a mixing function is employed. After the decoding process, inserted images are extracted by a separation function and the original sequence is reconstructed. By considering data from AVIRIS database, we will show how one decrease significantly the computing time for both coding and decoding.

A New Useful Biometrics Tool Based on 3D Brain Human Geometrical Characterizations

Previous clinical research studies consider the existence of differences among normal individual cerebral cortex and show that some features such as cerebral sulci are unique for each individual so that human brains are anatomically different and unique to each individual. In this work, we highlight the idea which consists in using medical data, such as brain MRI images for the purpose of individual identification or verification. In other words, we raise the following question: can one identify individuals using their brain geometry characteristics? Our aim is to validate the feasibility of this new biometrics tool based on human brain characterization. The proposed approach differs from existent biometrics modalities (e.g. fingerprint, hand, etc.) in the sense that brain features cannot be modified by individuals as it is the case when using fake fingerprints, or fake hands. In this work, we consider volumetric Magnetic Resonance Images (MRI) from which brain shapes are extracted using a 3D level-sets segmentation approach. Afterwards, geometrical descriptors are extracted from the 3D brain volume and from a projected version which provide specific features such as the isoperimetric ratio, the cortical surface curvature and the Gyrification index. Evaluations performed on a set of MRI images obtained from (MeDEISA) database show that it is possible to distinguish between individuals using their brain characteristics. Preliminary results are particularly encouraging.

Characterization of a human brain cortical surface mesh using discrete curvature classification and digital elevation model

In this paper we present a novel approach for brain surfacec characterization based on convexity and concavity analysis of cortical surface mesh. Initially, volumetric Magnetic Resonance Images (MRI) data is processed to generate a discrete representation of cortical surface using low-level segmentation tools and Level-Sets method. Afterward, pipeline procedure for brain characterization/labeling is developed. The first characterization method is based on discrete curvature classification. This is consists on estimating curvature information at each vertex in the cortical surface mesh. The second method is based on transforming the brain surface mesh into Digital Elevation Model (DEM), where each vertex is designed by its space coordinates and geometric measures related to a reference surface. In other word, it consists on analyzing the cortical surface as a topological map or an elevation map where the ridge or crest lines represent cortical gyri and valley lines represents sulci. The experimental results have shown the importance of these characterization methods for the detection of significant details related to the cortical surface.

Multimodal compression applied to biomedical data

In this paper, we introduce a novel approach to compress jointly a medical image and a multichannel bio-signals (e.g. ECG, EEG). This technique is based on the idea of Multimodal Compression (MC) which requires only one codec instead of multiple codecs. Objectively, biosignal samples are merged in the spatial domain of the image using a specific mixing function. Afterwards, the whole mixture is compressed using JPEG 2000. The spatial mixing function inserts samples in low-frequency regions, defined using a set of operations, including down-sampling, interpolation, and quad-tree decomposition. The decoding is achieved by inverting the process using a separation function. Results show that this technique allows better performances in terms of Compression Ratio (CR) compared to approaches which encode separately modalities. The reconstruction quality is evaluated on a set of test data using the PSNR (Peak Signal Noise Ratio) and the PRD (Percent Root Mean Square Difference), respectively for the image and biosignals.