• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.9
• /
• pp.700-704
• /
• 2009

In this paper, we propose a method for combining MR image with histopathology image of the prostate using image correction and multi-stage registration. Our method consists of four steps. First, the intensity of prostate bleeding area on T2-weighted MR image is substituted for that on T1-weighted MR image. And two or four tissue sections of the prostate in histopathology image are combined to produce a single prostate image by manual stitching. Second, rigid registration is performed to find the affine transformations that to optimize mutual information between MR and histopathology images. Third, the result of affine registration is deformed by the TPS warping. Finally, aligned images are visualized by the intensity intermixing. Experimental results show that the prostate tumor lesion can be properly located and clearly visualized within MR images for tissue characterization comparison and that the registration error between T2-weighted MR and histopathology image was 0.0815mm.

• Nuclear Engineering and Technology
• /
• v.51 no.6
• /
• pp.1610-1615
• /
• 2019

Positron emission tomography (PET)/magnetic resonance (MR) scanning has the advantage of less additional exposure to radiation than does PET/computed tomography (CT). In particular, MR based attenuation correction (MR AC) can greatly affect the image quality of PET and is frequently obtained using various MR sequences. Thus, the purpose of the current study was to quantitatively compare the image quality between MR non-AC (MR NAC) and MR AC in PET images with three MR sequences. Percent image uniformity (PIU), percent contrast recovery (PCR), and percent background variability (PBV) were estimated to evaluate the quality of PET images with MR AC. Based on the results of PIU, 15.2% increase in the average quality was observed for PET images with MR AC than for PET images with MR NAC. In addition, 28.6% and 71.1% improvement in the average results of PCR and PBV respectively, was observed for PET images with MR AC compared with that with MR NAC. Moreover, no significant difference was observed among the average values using three MR sequences. In conclusion, the current study demonstrated that PET with MR AC improved the image quality and can be help diagnosis in all MR sequence cases.

• Journal of Radiation Protection and Research
• /
• v.44 no.4
• /
• pp.149-155
• /
• 2019

Background: Magnetic resonance (MR) image guided radiation therapy system, enables real time MR guided radiotherapy (RT) without additional radiation exposure to patients during treatment. However, MR image lacks electron density information required for dose calculation. Image fusion algorithm with deformable registration between MR and computed tomography (CT) was developed to solve this issue. However, delivered dose may be different due to volumetric changes during image registration process. In this respect, synthetic CT generated from the MR image would provide more accurate information required for the real time RT. Materials and Methods: We analyzed 1,209 MR images from 16 patients who underwent MR guided RT. Structures were divided into five tissue types, air, lung, fat, soft tissue and bone, according to the Hounsfield unit of deformed CT. Using the deep learning model (U-NET model), synthetic CT images were generated from the MR images acquired during RT. This synthetic CT images were compared to deformed CT generated using the deformable registration. Pixel-to-pixel match was conducted to compare the synthetic and deformed CT images. Results and Discussion: In two test image sets, average pixel match rate per section was more than 70% (67.9 to 80.3% and 60.1 to 79%; synthetic CT pixel/deformed planning CT pixel) and the average pixel match rate in the entire patient image set was 69.8%. Conclusion: The synthetic CT generated from the MR images were comparable to deformed CT, suggesting possible use for real time RT. Deep learning model may further improve match rate of synthetic CT with larger MR imaging data.

• Journal of Ginseng Research
• /
• v.32 no.4
• /
• pp.332-336
• /
• 2008

Red ginseng is classified according to outer form and the quality of internal tissue, and red ginseng below third grade can't be sold. Also there are many differences in price according to grade. So if inferior Red ginseng is sold, economic loss and claim take place. This research is done conducted to investigate the possibility of the non-destruction internal tissue investigation of red ginseng. It is observed and compared that MR image after getting MR image agrees with real cutting side in 10-13% water content of red ginseng. The MR image can be obtained to see the internal section of red ginseng with equal condition of time, temperature and slice thickness in spin echo pulse sequence. The MR signal of red ginseng is very weak, because it contains low water density. So it takes about 30 minutes with the measurement of single point image (SPI). But the suitable time to distinguish internal tissues is about 9 seconds in TE (Echo Time) 2.23 ms, TR (Repetition Time) 150ms. The image to discriminate internal tissues in 9 seconds can be obtained when slice thickness is 10 mm with changes of 3, 5, 10 mm. The image obtained after 30 minutes' boiling of 55 degrees has clearer image than that of normal temperature. It is thought that MR signal is stronger through active motion of water particles as temperature increases. With this method MR image of red ginseng can be obtained and characteristics of internal tissues can be observed in such a short time.

• Journal of the Korean Society of Radiology
• /
• v.18 no.1
• /
• pp.37-44
• /
• 2024

Even though MR can reveal excellent soft-tissue contrast and functional information, CT is also required for electron density information for accurate dose calculation in Radiotherapy. For the fusion of MRI and CT images in RT treatment planning workflow, patients are normally scanned on both MRI and CT imaging modalities. Recently deep-learning-based generations of CT images from MR images became possible owing to machine learning technology. This eliminated CT scanning work. This study implemented a CycleGan deep-learning-based CT image generation from MR images. Three CT generators whose learning is based on T1- , T2- , or T1-&T2-weighted MR images were created, respectively. We found that the T1-weighted MR image-based generator can generate better than other CT generators when T1-weighted MR images are input. In contrast, a T2-weighted MR image-based generator can generate better than other CT generators do when T2-weighted MR images are input. The results say that the CT generator from MR images is just outside the practical clinics and the specific weight MR image-based machine-learning generator can generate better CT images than other sequence MR image-based generators do.

• Journal of Korean Neurosurgical Society
• /
• v.30 no.10
• /
• pp.1182-1186
• /
• 2001

Objectives : MR fluid-attenuated inversion recovery(FLAIR) image uses paired long inversion time and relaxation time that nulls the signal from CSF. With nulling of the CSF long echo time readout could be used to increase T2-weighting, hence improving the conspicuousness of most tissue lesions without the deleterious effects of CSF artifact seen on T2 weighted sequence. We examed the usefulness of FALIR image in the diagnosis of mild head injury. Methods : A total of 38 patients with mild head injury were examined by FLAIR image. We compared those images with CT scan and T1, T2-weighted images. Careful observation of MR images were done by two well-trained neuroradiologists. Each image was compared for conspicuousness and detectability of traumatic lesions might have shown abnormal signal intensities. The Wilcoxon signed ranks test was used for statistical evaluation. Results : The FLAIR image was significantly more sensitive than those of other images(p<0.001). T2 FFE(Fast Field Echo) image was more useful for detection of small petechial hemorrhages. Conclusion : FLAIR image is considered to be more sensitive than those of conventional MR images in the evaluation of mild head injuries.

• The Journal of Korean Society for Radiation Therapy
• /
• v.11 no.1
• /
• pp.43-48
• /
• 1999

Purpose : For practical application of the MR image for stereotactic radiosurgery, the target point achieved by acquisition of MR image in a relatively homogeneous phantom has to agree with the actual isocenter of irradiation in real radiosurgery and the amount of distortion of the MR image should be known. Materials and Methods : A dosimetric film with a random target point was inserted into a radish vertically and horizontally on axis Z and they were fixed with a headring. After image acquisition by stereotactic radiosurgery planning system, we achieved stereotactic coordinate of the target point and examined irradiation using the coordinate acquired as isocenter. After the irradiation, the film in the radish was developed and processed and the degree of coincidence between the target point marked on the film and the center of the radiation distribution. In order to measure the degree of distortion of the MR image in a different way, an acryl phantom was made and punctures were made at intervals of 1 cm and a drop of oil was dropped into it. Then, it was inserted into the radish vertically and horizontally on axis Z to acquire the MR image. Each coordinate was achieved and the estimation of distortion of MR image was made both in vertical and horizontal directions Results : The film from the radio was developed and for the one inserted vertically on axis Z, there was a good coincidence in the discrepancy between the target point marked on the film and the center of the radiation distribution. For the one inserted horizontally, the discrepancy between them was under 0.5 mm. As a result of estimating distortion of MR image using acryl, the discrepancy was under 0.45 mm in the case of the phantom inserted vertically on axis Z, and that of the one inserted horizontally was 1.4 mm. Conclusion : We were able to confirm good coincidence in homogeneous phantom in actual treatment position of radiosurgery using the MR image and the discrepancy measured in the analysis of distortion of the MR image did not exceed the permissible level. Therefore, it was evident the system of the hospital is suitable for radiosurgery using MR image.

• Journal of Biomedical Engineering Research
• /
• v.20 no.3
• /
• pp.283-290
• /
• 1999

Because MR(Magnetic Resonance) slice images have much information of functions about body organs, it is very effeclive for diagnoses lo analyze and visualize MR slice images. A visuahzation process is composed of medical image acquisition, preprocessmg, segmentation, inlerpolation, rendering. Segmentation and interpolation among thenl ,1re currenl hot topics because of MR slice image imperfections. This paper proposes a method for segmentalion, mlerpolation respectively and addresses 3 D-visualizmg of a head. We segmented head tissues uomg otructural knowledge of head studied by clinical experiments sequentially. We improved the dynamic elastic inlerpolation to Utilize in concave conlour. We compared the proposed segmentation method and the interpolation method with other methods.

• The Journal of Korean Academy of Prosthodontics
• /
• v.31 no.2
• /
• pp.249-270
• /
• 1993

For understanding of anatomy, physiology, and diseases of human TMJ, it is required to evaluate quantitatively the movement of the disc and condyle head of mandible. The histologic section of cadaver TMJ were examined, and the magnification of the MR image and its details of anatomy were evaluated. And then a quantitative analytic method, by comparing the Sectograph and the MR image of vital human TMJ, was proposed. For this study, 15 subjects(Male, 24~35years) were selected from a prosthodontic examination randomly, and each subject’s five interocclusal rubber registration records were made on the ICP, and 5, 10, 15, and 20mmjaw opening positions. All subjects were radiographed with a Denar Quint Sectograph Image System(Denar Corp., USA), and imaged with a MRP-20EX MR Image System(0.2T, Permanent Magnet Type, Hitachi Medical Corp., Japan) using an 100mm diameter bilateral type surface coil. These images were traced on the acetate tracing paper, and analyzed In this study, the findings led to the following conclusions. 1. In comparison of the histologic section of autopsy specimen with the MR image at the same section, the size(dimension) of MR image was 70% of the real one. It was possible to recognize the shape of articular disc, anterior and posterior attachments, and adjacent soft tissues, because of the excellent reproducibility of anatomical structure. 2. When we compared the amount of joint space on MR image with that of joint space on sectograph, the amount of joint space on sectograph was significantly greater than that of joint space on MR image, except at the top of condylar head. 3. The position of minimum joint space on sectograph at intercuspal position didn't coincide with the middle position of articular disc on MR image, and was approximately in the anterior third of posterior band of articular disc. 4. The amount of condylar movement on MR image at opening movement was greater than that of articular disc movement. From Intercuspal position to 5mm jaw-opening movement, the condylar movement showed hinge one, and over the range 5mm jaw-opening it suggested hinge & translatory one. 5. In terms of area variation of articular disc measured on MR image in sagittal plane, the area of posterior band increased with increasing the amount of Jaw opening, but the area of anterior band decreased conversely.

• Proceedings of the KOSOMBE Conference
• /
• v.1998 no.11
• /
• pp.60-61
• /
• 1998

In this paper, we propose an algorithm for automatic segmentation of 3-dimesional brain MR images. In order to segment 3-dimensional brain MR images, we start segmentation from a mid-sagittal brain MR image. Then the segmented mid-sagittal brain MR image is used as a mask that is applied to the remaining lateral slices. Then we apply preprocessing, which includes thresholding and region-labeling, to the lateral slices, resulting in simplified 3-D brain MR images. Finally, we remove remaining problematic regions in the 3-dimensional brain MR image using the connectivity-based thresholding segmentation algorithm. Experiments show satisfactory results.