Chuchart Pintavirooj

Penalized-likelihood reconstruction for metal artifact reduction in cone-beam CT

CT reconstruction from metal-embedded data usually produces streak artifacts that reduce the quality of the reconstructed images. In this paper, we propose a new technique for metal artifact reduction in cone-beam CT based on statistical reconstruction. First, the metal objects are segmented in the reconstructed images and then reprojected to obtain the measurement data of the metal objects using cone-beam reconstruction. The original measurement data in the metal area are corrected through cubic interpolation. y, the corrected measurement data are reconstructed with the penalized likelihood using the modified convex algorithm. The simulation results show that the reconstructed images of the metal object using the proposed metal artifact reduction technique are superior to conventional filtered backprojection reconstruction.

Real-time multianalyte biosensors based on interference-free multichannel monolithic quartz crystal microbalance

In this work, we design, fabricate and characterize a new interference-free multichannel monolithic quartz crystal microbalance (MQCM) platform for bio-sensing applications. Firstly, interference due to thickness-shear vibration mode coupling between channels in MQCM array is effectively suppressed by interposing a polydimethylsiloxane wall between adjacent QCM electrodes on a quartz substrate to form inverted-mesa-like structure. In addition, the electrical coupling due to the electrical impedance of solution is diminished by extending the flow path between them with an extended-design flow channel. The electrical testing results show that individual QCM signal is unaffected by those of adjacent channels under liquid loading, signifying the achievement of interference-free MQCM. The MQCM is applied for multi-analyte biosensing of IgG and HSA. The anti-IgG and anti-HSA are separately immobilized on two adjacent QCM electrodes, which are subsequently blocked with BSA to avoid unspecific binding. The MQCM biosensors are tested with single- and double-analyte solutions under continuous flow of buffer. The IgG and HSA QCM sensors only show frequency shift responses to their corresponding analytes and there are very small cross frequency shifts due to remnant unspecific binding. Moreover, MQCM sensors show approximately linear frequency shift response with analyte concentration. Therefore, the developed MQCM platform is promising for real-time interference-free label-free detection and quantification of multiple bio-analytes.

White Blood Cell Segmentation by Distance Mapping Active Contour

White blood cell segmentations are an important research issue in Hematology and related study field. Our research proposes a segmentation of nucleus and cytoplasm of peripheral white blood cell from color image slides. The segmentation is started by using a dilated perimeter of nucleus convex hull which propagated into a surrounding region in order to setup a color reference table of cytoplasm. Primary cytoplasm region was then estimated roughly. Distance mapping was applied to this primary area and used to create a gradient vector flow. The active contouring technique was then implemented according to the vector field and finally segmented the WBC boundary. The obtained segmentation outputs show that active contour which guided by the distance mapping from a surrounding area is able to extract nucleus and cytoplasm region efficiently.

Real time eye tracking using initial centroid and gradient analysis technique

Robust and accurate algorithm in real time eye tracking system has been a fundamental and challenging problem for computer vision and biomedical engineering area. This paper proposes a new method to estimate eye-motion position and direction based on initial centroid and gradient analysis technique. The proposed method was validated by tracking eye position within high and low occlusion conditions. Experimental results show that the proposed method improves the accuracy of tracking pupil position compared to conventional window-matching method. The proposed method yields 76.07% and 87,89% accuracies while the conventional window-matching performs 44.56% and 67.89% accuracies within high and low occlusion conditions, respectively.

Gaussian curvature-based geometric invariance

In this paper we derive a novel geometric invariance on surfaces that it is preserved under affine and weak perspective transformations, and it is local, intrinsic and computed from the differential geometry of the surface. Our 3D shape features are based on the Gaussian curvature and Mean curvature. When a surface undergoes an affine transformation, the shape features are the affine transformed shape features of the original surface, i.e., they are preserved and hence can be used for shape matching. We have tested robustness of the shape feature on the 3D facial data for various linear geometric transformations. The results show that our purposed shape feature is suitable for further application to 3D face identification due to its robustness to geometric transformation.

Rapid Simultaneous Algebraic Reconstruction Technique (SART) for Cone-Beam Geometry on Clustering System

An important problem in image processing is to construct a cross section of an object from several images of its transaxial projection. However, the time consuming and the complexity of the reconstruction process are the crucial problems. In addition, the reconstruction process requires very high performance of the computer. Therefore, in this paper, a concept of parallel programming method is employed to speed up a 3D simultaneous algebraic reconstruction technique. The scheduling process in the clustering system is improved. The appropriate amount of work is distributed to each computer (node) in the clustering system using centralized dynamic load balancing and a work-pool scheduling scheme, In this scheme round-robin algorithm for selecting a process is exercised. Our proposed system works successfully with decreasing the reconstruction time up to 78% percent referred to the normal image reconstruction form projection performed on a single computer.

Ultrasonic refractive index and sound velocity tomography

Ultrasonic computed tomography (UCT) is one of the methods capable of tissue characterization. It is expected to provide not only a quantitative image but also diagnostic information. As in X-ray CT, UCT requires a projection data to reconstruct a cross-sectional image. The projection data of ultrasonic tomography is based on measurement the time delay, which is time difference between ultrasound traverse with and without object. In this paper, we investigate two different types of quantitative UCT image, refractive-index and sound-velocity image. We purpose the new method of measurement the time delay by converting the received ultrasonic pulse to frequency domain and measuring the phase shift of the center frequency of the broadband pulse. The method seems more robust to noise. Two image reconstruction techniques are used for comparison purpose including traditional filtered back-projection and simultaneous algebraic reconstruction technique (SART).

2D Ultrasonic Reflection Tomography by Linear Array Transducer and Wave Reflector

In general, it was known that when the series of an ultrasonic broadband pulse travel through living organ or soft tissue, their reflected projection data can be used to reconstruct a novel tomographic image. Whereas, a data obtain from an ultrasonic tomography in transmission mode is not able to be used for reconstruction a characteristic image of soft tissue which naturally shadowed by a bone or hard tissue. An ultrasonic tomography in reflection mode is realized to be a solution for this problem. However the implementation to identify the projection by using an integrated attenuation coefficient of tissue from the pulsed wave ray path in reflection mode is still more complicated. In this paper, we propose a simulation in 2D of a reflection-mode ultrasonic tomographic system by using the linear array transducer and also a frequency centroid shift method to calculate the projection data. The imaging setup system was enhanced by wave reflecting plate in order to increase an echo wave path. The ultrasound wave paths of our simulation system were intend to focus on the fan beam trajectory which is rather close to a real propagation of the ultrasound wave travelling through soft tissue and ART was used finally as an image reconstruction algorithm. The obtained output shows that our system is more practical and possible toward for using in a clinical trial.

Comparison Study of Muscular-Contraction Classification Between Independent Component Analysis and Artificial Neural Network

We developed a multi-channel electromyogram acquisition system using PSOC microcontroller to acquire multi-channel EMG signals. An array of 4 times 4 surface electrodes was used to record the EMG signal. The obtained signals were classified by a back-propagation-type artificial neural network. B-spline interpolation technique has been utilized to map the EMG signal on the muscle surface. The topological mapping of the EMG is then analyzed to classify the pattern of muscle contraction using independent component analysis. The proposed system was successfully demonstrated to record EMG data and its surface mapping. The comparison study of muscular contraction classification using independent component analysis and artificial neural network demonstrates shows that performance of ANN classification is as comparable as that of the ICA. The computational time of ANN is also less than that of the ICA.

Using intrinsic surface geometry and absolute invariants for 3D face alignment and registration

In this paper we derive novel surface fiducial points that are computed from the differential geometry of the surface. The fiducial intrinsic points are intrinsic, local, and relative invariants, i.e., they are preserved under similarity, affine, and nonlinear transformations that are piecewise affine. As the fiducial points are computed from high order surface shape derivatives, their sensitivity to any noise either due to measurement error or local distortion is high. To reduce these effects, we use a B-Spline curve/surface representation that smoothes out the curve/surface prior to the computation of these intrinsic invariant points. The fiducial points are used in a non-iterative geometric-based method for 3D shape matching and registration of human faces. The matching is achieved by establishing correspondences between fiducial points after a sorting based on a set of absolute local affine invariants derived from them. The performance of the matching based on these fiducial points although shown for face alignment, the overall approach is generic to the alignment a variety of objects for which these intrinsic fiducial points exist and for scenarios where the classes of nonlinear transformations are piecewise affine.