• Journal of the Korean Physical Society
• /
• 제73권9호
• /
• pp.1393-1398
• /
• 2018

In conventional single-photon emission computed tomography (SPECT), only the photoelectric events in the detectors are used for image reconstruction. However, if the $^{131}I$ isotope, which emits high-energy radiations (364, 637, and 723 keV), is used in nuclear medicine, both photoelectric and Compton scattering events can be used for image reconstruction. The purpose of our work is to perform simulations for Compton SPECT by using the Geant4 application for tomographic emission (GATE). The performance of Compton SPECT is evaluated and compared with that of conventional SPECT. The Compton SPECT unit has an area of $12cm{\times}12cm$ with four gantry heads. Each head is composed of a 2-cm tungsten collimator and a $40{\times}40$ array of CdZnTe (CZT) crystals with a $3{\times}3mm^2$ area and a 6-mm thickness. Compton SPECT can use not only the photoelectric effect but also the Compton scattering effect for image reconstruction. The correct sequential order of the interactions used for image reconstruction is determined using the angular resolution measurement (ARM) method and the energies deposited in each detector. In all the results of simulations using spherical volume sources of various diameters, the reconstructed images of Compton SPECT show higher signal-to-noise ratios (SNRs) without degradation of the image resolution when compared to those of conventional SPECT because the effective count for image reconstruction is higher. For a Derenzo-like phantom, the reconstructed images for different modalities are compared by visual inspection and by using their projected histograms in the X-direction of the reconstructed images.

• 한국전자파학회논문지
• /
• 제27권7호
• /
• pp.646-652
• /
• 2016

본 논문에서는 전자파를 이용한 유방암 진단에서 유방의 조직형태에 따른 2차원 영상복원 결과를 논의하고 있다. 유방의 영상복원에 사용한 시스템은 16개의 송신/수신 안테나로 구성되어 있으며, 1,700 MHz를 사용하여 4가지의 유방조직형태(ED-, HD-, SC-, FT-type)에 대하여 영상복원을 수행하였다. 순방향 문제의 해석에는 모멘트법을 적용하였으며, 역문제 해석을 위한 최적화 알고리즘은 simplex 법을 사용하였다. 영상복원의 결과, ED형 및 HD형은 영상복원이 용이하지만, SC형 및 FT형의 영상복원에는 오차가 많이 포함되어 있어 복원이 쉽지 않음을 확인할 수 있었다.

• 한국추진공학회지
• /
• 제13권4호
• /
• pp.36-44
• /
• 2009

광학 선형 패터네이터 및 광학 토모그래피 기법을 사용하여 높은 주위 기체 압력 조건에서 분무단면을 측정하였다. 레이저 직선광이 분무 영역을 지나도록 하여 발생하는 Mie 산란 신호 및 직선광의 투과율을 측정하였으며, 이로부터 Beer-Lambert 법칙 및 수학적 단면 재구성 기법을 이용하여 분무 단면의 감쇠 계수 분포를 재구성하였다. 높은 주위 기체 압력 조건에서 광학적으로 밀한 분무가 발생하며, 그 결과 산란 신호의 감쇠 효과가 크게 증가한다. 따라서 미 산란 신호를 이용하는 광학 선형 패터네이터의 경우 감쇠 효과를 완화하는 데 한계를 보인다. 광학 토모그래피의 경우 분무를 통과하는 레이저광의 투과율 정보만을 이용하여 성공적으로 분무 단면을 재구성하였다.

• 한국분무공학회지
• /
• 제9권4호
• /
• pp.17-23
• /
• 2004

Spray characteristics of a twin-hole air shrouded nonle designed for gasoline injectors was investigated by using laser diffraction particle analyzer (LDPA) and tomography reconstruction- A confined spray chamber which is optically accessible through a pair of glass windows was made to simulate the fuel injection condition in intake manifold of gasoline engine. The measurement was applied to the twin hole injector with and without an air shroud. It demonstrates that for the case with an air shroud, fine atomization is achieved and there exists a large number of fine droplets between the region of the main spray streams, which conforms with the spray visualization. The drop size distribution was investigated as a function of elapse time after fuel injection. The distribution was greatly affected by the measurement position from the injector exit. Also, the spatially resolved spray volume fraction and Sauter Mean Diameter (SMD) from line-of-sight data of the LDPA are tomographically reconstructed by Convolution Fourier transformation under the steady injection condition.

• Smart Structures and Systems
• /
• 제17권1호
• /
• pp.17-29
• /
• 2016

Ultrasonic tomography is a powerful tool for identifying defects within an object or structure. But practical application of ultrasonic tomography to solids is often limited by time consuming transducer coupling. Air-coupled ultrasonic measurements may eliminate the coupling problem and allow for more rapid data collection and tomographic image construction. This research aims to integrate recent developments in air-coupled ultrasonic measurements with current tomography reconstruction routines to improve testing capability. The goal is to identify low velocity inclusions (air-filled voids and notches) within solids using constructed velocity images. Finite element analysis is used to simulate the experiment in order to determine efficient data collection schemes. Comparable air-coupled ultrasonic signals are then collected through homogeneous and isotropic solid (PVC polymer) samples. Volumetric (void) and planar (notch) inclusions within the samples are identified in the constructed velocity tomograms for a variety of transducer configurations. Although there is some distortion of the inclusions, the experimentally obtained tomograms accurately indicate their size and location. Reconstruction error values, defined as misidentification of the inclusion size and position, were in the range of 1.5-1.7%. Part 2 of this paper set will describe the application of this imaging technique to concrete that contains inclusions.

• Imaging Science in Dentistry
• /
• 제50권4호
• /
• pp.331-337
• /
• 2020

Purpose: As cone-beam computed tomography (CBCT) has become the most widely used 3-dimensional (3D) imaging modality in the dental field, storage space and costs for large-capacity data have become an important issue. Therefore, if 3D data can be stored at a clinically acceptable compression rate, the burden in terms of storage space and cost can be reduced and data can be managed more efficiently. In this study, a deep learning network for super-resolution was tested to restore compressed virtual CBCT images. Materials and Methods: Virtual CBCT image data were created with a publicly available online dataset (CQ500) of multidetector computed tomography images using CBCT reconstruction software (TIGRE). A very deep super-resolution (VDSR) network was trained to restore high-resolution virtual CBCT images from the low-resolution virtual CBCT images. Results: The images reconstructed by VDSR showed better image quality than bicubic interpolation in restored images at various scale ratios. The highest scale ratio with clinically acceptable reconstruction accuracy using VDSR was 2.1. Conclusion: VDSR showed promising restoration accuracy in this study. In the future, it will be necessary to experiment with new deep learning algorithms and large-scale data for clinical application of this technology.

• 한국의학물리학회:학술대회논문집
• /
• 한국의학물리학회 2002년도 Proceedings
• /
• pp.506-508
• /
• 2002

As an advancement of medical imaging modalities and analyzing software with multi-function, active researches to acquire high contrast and high resolution image being done. In recently, development of medical imaging modalities like as Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) is aiming to display anatomical structure more accuracy and faster. Thus, one of the important areas in CT today is the use of CT scanner for the quantitative evaluation of 3-D reconstruction images from 2-D tomographic images. In CT system, the effective slice thickness and the quality of 3-D reconstructed image will be influenced by imaging acquisition parameters (e.g. pitch and scan mode). In diagnosis and surgical planning, the accurate distance measurements of 3-D anatomical structures play an important role and the accuracy of distance measurements will depend on the acquisition parameters such as slice thickness, pitch, and scan mode. The skull phantom was scanned with SDCT for various acquisition parameters and acquisition slice thicknesses were 3 and 5 mm, and reconstruction intervals were 1, 2, and 3 mm to each pitch. 3-D visualizations and distance measurements were performed with PC based 3-D rendering and analyzing software. Results showed that the image quality and the measurement accuracy of 3-D SDCT images are independent to the reconstruction intervals and pitches.

• 대한의용생체공학회:의공학회지
• /
• 제24권5호
• /
• pp.473-480
• /
• 2003

반복법을 사용한 영상재구성은 기존의 여과역투사 방법에 비해 윌등한 장점을 가지고 있으므로 방출 전산화단층촬영에 있어 매우 중요한 역할을 해왔다. 그러나, 반복법을 사용한 영상재구성법은 데이터의 투사 및 역투사 계산의 반복으로 이루어져 있으므로 한 영상을 재구성하는데 필요한 계산량은 알고리즘의 구현을 위해 사용된 투사기 및 역투사기의 성능에 크게 좌우된다. 본 연구에서는 투사기 및 역투사기를 구현하기 위한 대표적 방법들의 성능을 정량적으로 비교한다. 각 구현방법에 있어서 투사-역투사 계산에 필요한 연산량을 줄이기 위해 이전에 계산된 결과들을 재사용함으로써 불필요하게 반복되는 연산들을 최소화하도록 하는 방법에 고찰한다. 실험결과에 의하면 선추적법은 연산속도 면에서 다른 방법에 비해 우수할 뿐 아니라, 정확성이 높은 개선된 영상을 제공하는 것으로 나타났다.

• 대한의용생체공학회:의공학회지
• /
• 제9권2호
• /
• pp.153-158
• /
• 1988

A three-dimensional surface is reconstructed from contour information as identified on two-dimensional computed tomographic slices. Gradient operator with curvature constraint would be applied to extract the contour automatically, and backtracking is also adopted to reduce the tracking error. The surface between the consecutive slice is efficiently reconstructed using a triangular surface tiles. Hidden surface elimination, shading and parallel projection of the reconstructed surface are provied on the display screen.

• 한국음향학회:학술대회논문집
• /
• 한국음향학회 1998년도 학술발표대회 논문집 제17권 2호
• /
• pp.239-242
• /
• 1998

본 논문에서는 횡파를 이용한 STAM 시스템을 이용하여 다층구조물에 대한 토모그라픽 영상을 복원함으로써 토모그라픽 영상의 분해능을 개선시키는 방법에 관하여 연구하였다. 먼저, STAM 시스템의 수조에서 초음파가 시료로 입사될 때 입사각도에 따라 발생되는 종파와 횡파에 대한 모드변환에 관하여 고찰하였다. 실험을 위하여 시료로 각 층이 서로 다른 패턴을 갖는 2층의 알루미늄을 가공하여 다층구조물을 제작하였으며, 시료에 횡파를 여기시키기 위하여 초음파변환기의 입사각을 18$^{\circ}$로 사각 입사시킬 수 있는 웨지를 가공하여 수조를 제작하였다. 본 실험에서는 동작주파수가 10MHz인 STAM 시스템을 이용하여 4중 프로젝션에 대한 토모그라픽 데이터를 영상처리하여 토모그라픽 영상을 복원하였으며 이를 기존의 종파를 이용한 토모그라픽 영상과 비교분석한 결과, 콘트라스트와 축방향 분해능이 향상되고 토모그라픽 영상을 복원할 수 있었다.