Shinji Umeyama

Multidistance probe arrangement to eliminate artifacts in functional near-infrared spectroscopy

Functional near-infrared spectroscopy has the potential to easily detect cerebral functional hemodynamics. However, in practical fNIRS measurements, a subjects physical or systemic physiological activities often cause undesirable artifacts. Such activities can be evoked even by task execution. In this case, observed artifacts may correlate strongly with the task sequence, and it is difficult to eliminate them by conventional signal filtering techniques. We present a theoretical analysis and Monte Carlo simulations of layered media in which both scattering and absorption changes occur, and show that a multidistance probe arrangement is effective in removing artifacts and extracting functional hemodynamics. The probe arrangement is determined based on simulation results. Artifacts induced by nonfunctional tasks (body tilting, head nodding, and breath holding) are clearly observed when a conventional method is used; such artifacts are appreciably reduced by the proposed method. Signals evoked by single-sided finger movements are observed at both hemispheres when we use a conventional method. On the other hand, localized signals at the primary motor area are observed by the proposed method. A statistically significant increase in oxygenated hemoglobin and decrease in deoxygenated hemoglobin are simultaneously observed at the contralateral primary motor area.

Separation of fNIRS Signals into Functional and Systemic Components Based on Differences in Hemodynamic Modalities

In conventional functional near-infrared spectroscopy (fNIRS), systemic physiological fluctuations evoked by a bodys motion and psychophysiological changes often contaminate fNIRS signals. We propose a novel method for separating functional and systemic signals based on their hemodynamic differences. Considering their physiological origins, we assumed a negative and positive linear relationship between oxy- and deoxyhemoglobin changes of functional and systemic signals, respectively. Their coefficients are determined by an empirical procedure. The proposed method was compared to conventional and multi-distance NIRS. The results were as follows: (1) Nonfunctional tasks evoked substantial oxyhemoglobin changes, and comparatively smaller deoxyhemoglobin changes, in the same direction by conventional NIRS. The systemic components estimated by the proposed method were similar to the above finding. The estimated functional components were very small. (2) During finger-tapping tasks, laterality in the functional component was more distinctive using our proposed method than that by conventional fNIRS. The systemic component indicated task-evoked changes, regardless of the finger used to perform the task. (3) For all tasks, the functional components were highly coincident with signals estimated by multi-distance NIRS. These results strongly suggest that the functional component obtained by the proposed method originates in the cerebral cortical layer. We believe that the proposed method could improve the reliability of fNIRS measurements without any modification in commercially available instruments.

Cortical activity in multiple motor areas during sequential finger movements: an application of independent component analysis.

Multiple cortical regions such as the supplementary motor area (SMA), premotor cortex (PM), and primary motor cortex (M1) are involved in the sequential execution of hand movements, but it is unclear how these areas collaborate in the preparation and execution of ipsilateral and contralateral hand movements. In this study, we used right-handed subjects to examine the spatial distribution and temporal profiles of motor-related activity during visually cued sequential finger movements by applying independent component analysis (ICA) to event-related functional magnetic resonance imaging (fMRI) signals. The particular merit of the ICA method is that it allows brain activity in individual subjects to be elucidated without making a priori assumptions about the anatomical areas that are activated or the temporal profile of activity. By applying ICA, we found that (1) the SMA contributed to both the preparation and execution of movements of the right and left hand; (2) the left M1 and dorsal premotor cortex (PMd) contributed to both the preparation and execution of movements of the right and left hand, whereas the right M1 and PMd contributed mainly to the execution of movements of the left hand; (3) pre-SMA areas were activated in some subjects in concert with the posterior parietal and prefrontal cortex; and (4) fMRI signals over superficial cortical draining veins could be distinguished from cortical activation. We suggest that ICA is useful for categorizing distributed task-related activities in individual subjects into several spatially independent activities that represent functional units in motor control.

Monte Carlo study of global interference cancellation by multidistance measurement of near-infrared spectroscopy

The performance of near-infrared spectroscopy is sometimes degraded by the systemic physiological interference in the extracerebral layer. There is some systemic interference, which is highly correlated with the functional response evoked by a task execution. This kind of interference is difficult to remove by using ordinary techniques. A multidistance measurement method is one of the possible solutions for this problem. The multidistance measurement method requires estimation parameters derived from partial pathlength values of tissue layers to calculate an absorption coefficient change from a temporal absorbance change. Because partial path lengths are difficult to obtain, experimentally, we estimated them by a Monte Carlo simulation based on a five-layered slab model of a human adult head. Model parameters such as thickness and the transport scattering coefficient of each layer depend on a subject and a measurement position; thus, we assumed that these parameters obey normal distributions around standard parameter values. We determined the estimation parameters that provide a good separation performance in average for the model parameter distribution. The obtained weighting is robust to model parameter deviation and provides smaller errors on average compared to the parameters, which are determined without considering parameter distribution.

MICA: multimodal independent component analysis

We propose MICA (multimodal independent component analysis) that extends ICA (independent component analysis) to the case that there is a pair of information sources. MICA extracts statistically dependent pairs of features from the sources, where the components of feature vector extracted from each source are independent. Therefore, the cost function is constructed to optimize this degree of pairwise dependence as well as optimizing the cost function of ICA. We approximate the cost function by two dimensional Gram-Charlier expansion and propose a gradient descent algorithm derived by Amaris natural gradient The relation between MICA and traditional CCA (canonical correlation analysis) is similar to the relation between ICA and PCA (principal component analysis).

Face matching through information theoretical attention points and its applications to face detection and classification

This paper presents a face matching method through information theoretical attention points. The attention points are selected as the points where the outputs of Gabor filters applied to the contrast-filtered image (Gabor features) have rich information. The information value of Gabor features of the certain point is used as the weight and the weighed sum of the correlations is used as the similarity measure for the matching. To cope with the scale changes of a face, several images with different scales are generated by interpolation from the input image and the best match is searched. By using the attention points given from the information theoretical point of view, the matching becomes robust under various environments. This matching method is applied to face detection of a known person and face classification. The effectiveness of the proposed method is confirmed by experiments using the face images captured over years under the different environments.

New cross-talk measure of near-infrared spectroscopy and its application to wavelength combination optimization

In near-infrared spectroscopy, concentration changes in oxygenated and deoxygenated hemoglobin are calculated from the changes in the attenuation of the measurement light. This is done by solving a linear equation based on the modified Lambert-Beer law. To solve this equation, we need to know the partial optical pathlengths in the activated region in the brain. Because they are difficult to know, a wavelength-independent constant or a wavelength-dependent total optical pathlength has been substituted for these values in actual measurements. This kind of substitution inevitably produces errors, called cross-talk, when calculating concentration changes. In this paper, we propose a new cross-talk measure for dual and triple wavelength measurements, and analyze it over various wavelength combinations. The results indicate that constant substitution is not inferior to total path-length substitution in dual wavelength measurements, and that total path-length substitution is very effective for triple wavelength measurements.

Detection of an unstable and/or a weak probe contact in a multichannel functional near-infrared spectroscopy measurement

Abstract. Multichannel functional near-infrared spectroscopy measurements involve the placement of many probes on a subject’s head. A stable close contact between the probe and head surface is essential. We propose a way to detect two types of problematic probe contacts from the measurement data: an unstable contact whose light transmission easily fluctuates with body motion, and a weak contact whose light transmission is constantly small. An unstable contact causes large baseline fluctuation, whereas a weak contact causes large noise. Because absorbance changes caused by body motion and noise show different spectroscopic properties from the tissue hemoglobin absorption, they have a component orthogonal to the plane spanned by hemoglobin molar extinction coefficient vectors. We use this information to detect unstable and/or weak contacts. Probes are shared by different channels, and this sharing configuration is determined by the probe arrangement. Thus, the baseline fluctuation and noise of the channels are related to contact instability and weakness of the probe according to the probe arrangement. Unstable and/or weak probes are determined by solving an inverse problem of this relation. Problematic probes can be effectively determined using the proposed method.

A point pattern matching algorithm

This paper presents an algorithm for finding a matching between two point patterns given in m-dimensional Euclidean space. The point pattern matching algorithm here is formulated as a tree search procedure, and the search tree is pruned by a specified tolerance measure for matching. The proposed method is based on the locations of points and is invariant to rotation, translation, and scaling. Moreover, to guide the search, it is easy to use global geometric constraints, such as rotation or scaling parameters.

A multidistance probe arrangement NIRS for detecting absorption changes in cerebral gray matter layer

We provide theoretical validation of the brain-functional detection using multidistance probe arrangement based on Monte Carlo simulations of five-layered model in which both scattering and absorption changes occur. It shows that optimized multidistance probe arrangement can be effective in removing interferences by scattering and absorption changes in upper layers and extracting absorption change in the gray matter layer. Using newly designed probes and their holder system, both conventional and proposed fNIRS measurements were implemented with non-functional (body and head movements and respiratory change) and functional (finger opposition) tasks. Artifacts, even if it correlate with task sequence, were well reduced. Functional signals were well localized at lateralized cerebral functional area.