3D MR image denoising using rough set and kernel PCA method

In this paper, we have presented a two stage method, using kernel principal component analysis (KPCA) and rough set theory (RST), for denoising volumetric MRI data. A rough set theory (RST) based clustering technique has been used for voxel based processing. The method groups similar voxels (3D cubes) using class and edge information derived from noisy input. Each clusters thus formed now represented via basis vector. These vectors now projected into kernel space and PCA is performed in the feature space. This work is motivated by idea that under Rician noise MRI data may be non-linear and kernel mapping will help to define linear separator between these clusters/basis vectors thus used for image denoising. We have further investigated various kernels for Rician noise for different noise levels. The best kernel is then selected on the performance basis over PSNR and structure similarity (SSIM) measures. The work has been compared with state-of-the-art methods under various measures for synthetic and real databases.     

3D MR microscopy with resolution 3.7 microm by 3.3 microm by 3.3 microm

The technique of magnetic resonance imaging microscopy holds promise of bringing the full capabilities of NMR to arbitrarily specified positions within spatially inhomogeneous systems, including biological cells, yet the possibilities are limited by the need for adequate sensitivity and spatial resolution. We report proton magnetic resonance images obtained by combining advances in receiver coil sensitivity, gradient strength, and pulse/gradient sequence design. We achieve resolution of 3.7 +/- 0.4 microm by 3.3 +/- 0.3 microm by 3.3 +/- 0.3 microm for a volume resolution approximately 40 femtoliters (corresponding to approximately 3 x 10(12) proton spins).     

Noninvasive 3D MR microscopy as a tool in pharmacological research: Application to a model of rheumatoid arthritis

Magnetic resonance microscopy (MRM) was applied to noninvasively image skeletal structures in the hindpaw of the live rat to characterize the progression of a heterologous type II collagen-induced arthritic process. Using a resonator, with optimized filling factor, three-dimensional (3D) gradient-echo images with voxel dimensions of 94 × 81 × 60 μm³ were acquired in 54.6 min. Three-dimensional MRM reduces the slice positioning problem, which is critical in longitudinal studies. Moreover, due to the much smaller slice thickness of images derived from 3D data sets, partial volume effects are less pronounced than in corresponding 2D images. Distinct pathomorphological changes associated with the collagen-induced arthritic process (e.g., increase of metatarsophalangeal joint space, and bone and cartilage erosion) could thus be analyzed under in vivo conditions.     

High resolution 3D x-ray diffraction microscopy

We have imaged a 2D buried Ni nanostructure at 8 nm resolution using coherent x-ray diffraction and the oversampling phasing method. By employing a 3D imaging reconstruction algorithm, for the first time we have experimentally determined the 3D structure of a noncrystalline nanostructured material at 50 nm resolution. The 2D and 3D imaging resolution is currently limited by the exposure time and the computing power, while the ultimate resolution is limited by the x-ray wavelengths. We believe these results pave the way for the development of atomic resolution 3D x-ray diffraction microscopy.     

Fringe projection based quantitative 3D microscopy

A fringe projection based quantitative three-dimensional microscopy (FP-3DM) is presented. The image formation for FP-3DM is formulated based on a concept of active micro stereovision. The problem of point correspondences in active stereo imaging can be solved with help of the phase measuring technique. A prototype of the FP-3DM is also established and calibration strategy for proposed FP-3DM is suggested. Some preliminary experiment results are also presented to verify this approach. The FP-3DM can provide quantitative 3D micro imaging especially for quantitative characterization of 3D microstructures.     

3D MR imaging of dental cavities-an in vitro study

A 3D spin-echo (3D SE) pulse sequence was used on a 4.7 T research MRI system to produce images of extracted human first molar tooth placed in CuSO4 water solution. The maximal resolution achieved was 35 x 63 x 300 microm3 in read and two phase directions, respectively. The high-intensity signal from water in solution together with the lack of signal from mineralized tooth tissue produce very good contrast allowing to visualize topography of outer and inner surfaces of the tooth. The 3D MR data were median filtered, binarized and then divided into separate segments corresponding to the inner tooth cavities and the hard tooth tissue. The topography of the root canals was visualized and the canals volume was calculated. The presented technique may be used for quantitative analysis of the root canal cavities shape and volume. The results of such an analysis may be applied for estimation of the quality of the impressional mapping methods in restorative dentistry or as an alternative non-impressional 3D mapping method.     

MR microscopy of micron scale structures

Magnetic resonance (MR) microscopy of up to 1–5-μm resolutions have been reported previously. The tested phantom structures, however, had widths one order of magnitude bigger than the reported resolutions, e.g., spherical beads or capillary tubes of tens-of-micron diameters or wall thicknesses have been imaged. In this study, we fabricated structures having a few micron widths on a silicon wafer and imaged them using our 1-μm-resolution MR microscopy at 14.1 T. Micron scale width structures were, for the first time, resolved by MR microscopy.     

Contrast-enhanced 3D MR voiding urethrography: preliminary results

Contrast-enhanced 3D MR voiding urethrography (CE 3D MRVU) was performed on 5 healthy volunteers and 18 patients with urethral disease. After intravenous injection of 0.3 ml/kg gadolinium, the images of the three consecutive acquisitions of the 3D MRVU technique were obtained during voiding. The raw data were reconstructed on all patients for visual analysis. The image quality of the volunteers was technically sufficient to demonstrate normal urethral anatomy. Contrast-enhanced 3D MR voiding urethrography of the urethral strictures was compared with conventional retrograde urethrography (n = 10) and urethroscopy (n = 12). The urethral pathologies including strictures and other obstructive causes of impaired urethral flow were correctly identified on CE 3D MRVU.     

Jet fuel toxicity: skin damage measured by 900-MHz MRI skin microscopy and visualization by 3D MR image processing

The toxicity of jet fuels was measured using noninvasive magnetic resonance microimaging (MRM) at 900-MHz magnetic field. The hypothesis was that MRM can visualize and measure the epidermis exfoliation and hair follicle size of rat skin tissue due to toxic skin irritation after skin exposure to jet fuels. High-resolution 900-MHz MRM was used to measure the change in size of hair follicle, epidermis thickening and dermis in the skin after jet fuel exposure. A number of imaging techniques utilized included magnetization transfer contrast (MTC), spin-lattice relaxation constant (T1-weighting), combination of T2-weighting with magnetic field inhomogeneity (T2*-weighting), magnetization transfer weighting, diffusion tensor weighting and chemical shift weighting. These techniques were used to obtain 2D slices and 3D multislice–multiecho images with high-contrast resolution and high magnetic resonance signal with better skin details. The segmented color-coded feature spaces after image processing of the epidermis and hair follicle structures were used to compare the toxic exposure to tetradecane, dodecane, hexadecane and JP-8 jet fuels. Jet fuel exposure caused skin damage (erythema) at high temperature in addition to chemical intoxication. Erythema scores of the skin were distinct for jet fuels. The multicontrast enhancement at optimized TE and TR parameters generated high MRM signal of different skin structures. The multiple contrast approach made visible details of skin structures by combining specific information achieved from each of the microimaging techniques. At short echo time, MRM images and digitized histological sections confirmed exfoliated epidermis, dermis thickening and hair follicle atrophy after exposure to jet fuels. MRM data showed correlation with the histopathology data for epidermis thickness (R²=0.9052, P<.0002) and hair root area (R²=0.88, P<.0002). The toxicity of jet fuels on skin structures was in the order of tetradecane>hexadecane>dodecane. The method showed a sensitivity of 87.5% and a specificity of 75%. By MR image processing, different color-coded skin structures were extracted and 3D shapes of the epidermis and hair follicle size were compared. In conclusion, high-resolution MRM measured the change in skin epidermis and hair follicle size due to toxicity of jet fuels. MRM offers a three-dimensional spatial visualization of the change in skin structures as a method of toxicity evaluation and for comparison of jet fuels.     

Three-dimensional snow images by MR microscopy

MR microscopy technique was introduced to visualize and quantify the three-dimensional structure of snowpack. Since the NMR signal from the ice was week, we looked at the air space instead filling with dodecane or aniline doped with iron acetylacetonate. Four types of snow were tested: ice spheres, large rounded poly crystals, small rounded mono-crystals and depth hoar crystals. A specific specimen-cooling system was developed to keep the temperature below 0 degrees C. In the experiments 0.5 to 2 h were necessary to accumulate the signals enough to obtain a 3D micro-image; the image matrix 128(3), voxel size (200 microm)3 or 256(3) (120 microm)3. Comparison with the 2D data using the conventional section plane method was also carried out and MR microscopy is proved to be a very useful method to visualize the microstructure of snowpack.     

Insight in the 3D morphology of silica-based nanotubes using electron microscopy

Amorphous silica-based nanotubes (SBNTs) were synthesized from phosphoryl triamide, OP(NH₂)₃, thiophosphoryl triamide, SP(NH₂)₃, and silicon tetrachloride, SiCl₄, at different temperatures and with varying amount of the starting material SiCl₄ using a recently developed template-free synthesis approach. Diameter and length of the SBNTs are tunable by varying the synthesis parameters. The 3D mesocrystals of the SBNTs were analyzed with focused ion beam sectioning and electron tomography in the transmission electron microscope showing the hollow tubular structure of the SBNTs. The reconstruction of a small SBNT assembly was achieved from a high-angle annular-dark field scanning transmission electron microscopy tilt series containing only thirteen images allowing analyzing beam sensitive material without altering the structure. The reconstruction revealed that the individual nanotubes are forming an interconnected array with an open channel structure.     

Laplacian based non-local means denoising of MR images with Rician noise

Magnetic Resonance (MR) image is often corrupted with a complex white Gaussian noise (Rician noise) which is signal dependent. Considering the special characteristics of Rician noise, we carry out nonlocal means denoising on squared magnitude images and compensate the introduced bias. In this paper, we propose an algorithm which not only preserves the edges and fine structures but also performs efficient denoising. For this purpose we have used a Laplacian of Gaussian (LoG) filter in conjunction with a nonlocal means filter (NLM). Further, to enhance the edges and to accelerate the filtering process, only a few similar patches have been preselected on the basis of closeness in edge and inverted mean values. Experiments have been conducted on both simulated and clinical data sets. The qualitative and quantitative measures demonstrate the efficacy of the proposed method.     

A general strategy for anisotropic diffusion in MR image denoising and enhancement

Anisotropic diffusion (AD) has proven to be very effective in the denoising of magnetic resonance (MR) images. The result of AD filtering is highly dependent on several parameters, especially the conductance parameter. However, there is no automatic method to select the optimal parameter values. This paper presents a general strategy for AD filtering of MR images using an automatic parameter selection method. The basic idea is to estimate the parameters through an optimization step on a synthetic image model, which is different from traditional analytical methods. This approach can be easily applied to more sophisticated diffusion models for better denoising results. We conducted a systematic study of parameter selection for the AD filter, including the dynamic parameter decreasing rate, the parameter selection range for different noise levels and the influence of the image contrast on parameter selection. The proposed approach was validated using both simulated and real MR images. The model image generated using our approach was shown to be highly suitable for the purpose of parameter optimization. The results confirm that our method outperforms most state-of-the-art methods in both quantitative measurement and visual evaluation. By testing on real images with different noise levels, we demonstrated that our method is sufficiently general to be applied to a variety of MR images.     

3D sensitivity encoded ellipsoidal MR spectroscopic imaging of gliomas at 3T

Purpose

The goal of this study was to implement time efficient data acquisition and reconstruction methods for 3D magnetic resonance spectroscopic imaging (MRSI) of gliomas at a field strength of 3T using parallel imaging techniques.

Methods

The point spread functions, signal to noise ratio (SNR), spatial resolution, metabolite intensity distributions and Cho:NAA ratio of 3D ellipsoidal, 3D sensitivity encoding (SENSE) and 3D combined ellipsoidal and SENSE (e-SENSE) k-space sampling schemes were compared with conventional k-space data acquisition methods.

Results

The 3D SENSE and e-SENSE methods resulted in similar spectral patterns as the conventional MRSI methods. The Cho:NAA ratios were highly correlated (P<.05 for SENSE and P<.001 for e-SENSE) with the ellipsoidal method and all methods exhibited significantly different spectral patterns in tumor regions compared to normal appearing white matter. The geometry factors ranged between 1.2 and 1.3 for both the SENSE and e-SENSE spectra. When corrected for these factors and for differences in data acquisition times, the empirical SNRs were similar to values expected based upon theoretical grounds. The effective spatial resolution of the SENSE spectra was estimated to be same as the corresponding fully sampled k-space data, while the spectra acquired with ellipsoidal and e-SENSE k-space samplings were estimated to have a 2.36–2.47-fold loss in spatial resolution due to the differences in their point spread functions.

Conclusion

The 3D SENSE method retained the same spatial resolution as full k-space sampling but with a 4-fold reduction in scan time and an acquisition time of 9.28 min. The 3D e-SENSE method had a similar spatial resolution as the corresponding ellipsoidal sampling with a scan time of 4:36 min. Both parallel imaging methods provided clinically interpretable spectra with volumetric coverage and adequate SNR for evaluating Cho, Cr and NAA.     

MR microscopy of water diffusion tensor in biological systems

Diffusion tensor imaging (DTI) of the control and mechanically injured spinal cord of a ratin vitro andin vivo are reported.In vitro experiments were done on a home-built 6.4 T magnetic resonance microscope. Results for formaline-fixed samples show significant differences in diffusion tensor components between gray matter and white matter of the control spinal cord. Moreover, it is shown that already 6–10 min after the injury DTI can detect changes in water diffusion in areas extending far beyond the region of primary tissue damage.In vivo experiments were performed using a 9.4 T Magnex magnet and Bruker Medspec imaging system. Good-quality DTI images, free from motion artifacts were obtained. Results from control samples confirm differences in water diffusion between white matter and gray matter, observedin vitro. In vivo experiments show that characteristic changes in water diffusion observedin vitro 6–10 min after injury are preserved 60–360 min after injury, without significant alteration during this time.     

Automatic localization of EEG electrode markers within 3D MR data

The electrical activity of the brain can be monitored using ElectroEncephaloGraphy (EEG). From the positions of the EEG electrodes, it is possible to localize focal brain activity. Thereby, the accuracy of the localization strongly depends on the accuracy with which the positions of the electrodes can be determined. In this work, we present an automatic, simple, and accurate scheme that detects EEG electrode markers from 3D MR data of the human head.     

Advances toward MR microscopy of single biological cells

Here we report further progress toward the goal of achieving proton magnetic resonance imaging microscopy with resolution approaching a few micrometers in all three dimensions. We obtain proton images of a phantom sample — a microcapillary containing water and 39 μm diameter polymer microspheres — with a resolution of a few micrometers (perhaps about 5 μm) in all three spatial dimensions.     

Direct MR arthrography of the shoulder: 2D vs. 3D gradient-echo imaging.

The aim of this study was to determine the value of a fat suppressed 3D gradient-echo sequence (GRE) data set in comparison to a 2D GRE sequence in direct MR arthrography of the shoulder. For this purpose we examined 50 consecutive patients with subacute or chronic disorders of the shoulder using a 1.5 T scanner: Transverse T1-weighted 2D (slice thickness 4 mm) and 3D GRE (slice thickness 1.5 mm reconstructed from 3 mm), oblique coronal T2- and T1-weighted turbo spin-echo (TSE) and sagittal T1-weighted TSE with fat saturation were applied. Visual image analysis of anatomical and pathological structures was performed by two independent observers. A correlation to surgical results was available in 21 patients. Transverse GRE sequences were well suited for analysis of the anterior/posterior labrum, the middle glenohumeral ligament, and cartilage. 3D GRE with fat suppression was slightly superior to 2D GRE without fat suppression in the evaluation of the anterior/posterior labrum, and the middle glenohumeral ligament, whereas for cartilage no significant differences were found between both sequences. Concerning pathological findings, in most of the cases 2D delivered the same results as 3D. In conclusion, a T1-weighted 3D GRE data set with fat saturation in transverse orientation may be useful for evaluation of the anterior/posterior labrum, and the middle glenohumeral ligament. However, similar measured slice thickness of 3 mm-even if interpolated to 1.5 mm-compared to a 2D sequence with 4 mm does not provide significant diagnostic advantages.     

Fluorescence microscopy with 3D resolution in the 100 nm range

We propose a combination of 4Pi and Theta microscopies to improve the resolution of fluorescence microscopy by using six microscope objectives in a configuration we call Multiple Objective Microscopy (MOM). A resolution in the 100 nm range is obtained in the three dimensions using low numerical aperture, long working distance objectives. The obtained results not only show that high resolution is not restricted to high numerical aperture objectives, but also open the possibility of exploring large volumes with a very good resolution.