Masaaki Shinohara

Accuracy of deconvolution analysis based on singular value decomposition for quantification of cerebral blood flow using dynamic susceptibility contrast-enhanced magnetic resonance imaging

Deconvolution analysis (DA) based on singular value decomposition (SVD) has been widely accepted for quantification of cerebral blood flow (CBF) using dynamic susceptibility contrast-enhanced magnetic resonance imaging (DSC-MRI). When using this method, the elements in the diagonal matrix obtained by SVD are set to zero when they are smaller than the threshold value given beforehand. In the present study, we investigated the effect of the threshold value on the accuracy of the CBF values obtained by this method using computer simulations. We also investigated the threshold value giving the CBF closest to the assumed value (optimal threshold value) under various conditions. The CBF values obtained by this method largely depended on the threshold value. Both the mean and the standard deviation of the estimated CBF values decreased with increasing threshold value. The optimal threshold value decreased with increasing signal-to-noise ratio and CBF, and increased with increasing cerebral blood volume. Although delay and dispersion in the arterial input function also affected the relationship between the estimated CBF and threshold values, the optimal threshold value tended to be nearly constant. In conclusion, our results suggest that the threshold value should be carefully considered when quantifying CBF in terms of absolute values using DSC-MRI for DA based on SVD. We believe that this study will be helpful in selecting the threshold value in SVD.

An anisotropic diffusion method for denoising dynamic susceptibility contrast-enhanced magnetic resonance images

The purpose of this study was to present an application of a novel denoising technique for improving the accuracy of cerebral blood flow (CBF) images generated from dynamic susceptibility contrast-enhanced magnetic resonance imaging (DSC-MRI). The method presented in this study was based on anisotropic diffusion (AD). The usefulness of this method was firstly investigated using computer simulations. We applied this method to patient data acquired using a 1.5 T MR system. After a bolus injection of Gd-DTPA, we obtained 40-50 dynamic images with a 1.32-2.08 s time resolution in 4-6 slices. The dynamic images were processed using the AD method, and then the CBF images were generated using pixel-by-pixel deconvolution analysis. For comparison, the CBF images were also generated with or without processing the dynamic images using a median or Gaussian filter. In simulation studies, the standard deviation of the CBF values obtained after processing by the AD method was smaller than that of the CBF values obtained without any processing, while the mean value agreed well with the true CBF value. Although the median and Gaussian filters also reduced image noise, the mean CBF values were considerably underestimated compared with the true values. Clinical studies also suggested that the AD method was capable of reducing the image noise while preserving the quantitative accuracy of CBF images. In conclusion, the AD method appears useful for denoising DSC-MRI, which will make the CBF images generated from DSC-MRI more reliable.

Wavelet de-noising in digital chest radiographs by generalized cross validation

Unlike the “universal threshold” [1], the GCV (generalized cross validation) method [2] has an advantage that the noise-power information is not needed for determining the optimal threshold for wavelet de-noising. The GCV value behaves like MSE (mean square error) with change of threshold. We investigated whether this GCV method would be effective to noise reduction of chest x-ray digital images.

Automatic extraction of cerebral blood vessel and aneurysm from magnetic resonance angiography

Region growing algorithm has been often used for extracting blood vessels from magnetic resonance angiography (MRA) data. The purpose of this study was to develop an objective method with which cerebral blood vessels and aneurysms are automatically extracted from MRA data.

Fast image generation of cerebral perfusion parameters using multi-detector row CT and deconvolution analysis

Patients with acute cerebral ischemia may be treated by means of thrombolytic therapy within 4–6 hours after symptom onset. As a result, clinical examination and diagnostic imaging must be performed within a short period of time. Then, this study was undertaken to develop a fast method for generation of the maps of cerebral blood flow (CBF), cerebral blood volume (CBV) and mean transit time (MTT) using a multi-detector row CT and deconvolution analysis based on singular value decomposition (DA-SVD).

Accuracy of cerebral blood flow obtained from dynamic susceptibility contrast-enhanced MRI using deconvolution analysis based on singular value decomposition

Based on the indicator dilution theory, cerebral blood flow (CBF) can be quantified from dynamic susceptibility contrast-enhanced magnetic resonance imaging (DSC-MRI) by deconvolution between the arterial input function (AIF) and the concentration curve in the brain [1]. There are several methods for performing deconvolution, and an algebraic approach based on singular value decomposition (SVD) has been shown to be one of the most reliable options. The purpose of this study was to investigate the effect of the threshold value for SVD on the accuracy of CBF quantification from DSC-MRI under various conditions.

Optimization of an anisotropic diffusion method for medical image processing

An anisotropic diffusion (AD) method is based on a new concept for image processing, in which smoothing is formulated as a diffusive process and is suppressed or stopped at boundaries by selecting locally adaptive diffusion strengths [1]. It is known that this method is capable of reducing image noise, while preserving the spatial resolution of images. However, the results depend on the parameters used in the diffusion function (DF) adopted in the AD method. Then, this study was undertaken to optimize the parameters used in DF by computer simulations.