• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.432-435
• /
• 2019

Convolutional Neural Networks (CNNs) have recently been gaining popularity in the medical image analysis field because of their image segmentation capabilities. In this paper, we present a CNN that performs automated brain tumor segmentations of sparsely annotated 3D Magnetic Resonance Imaging (MRI) scans. Our CNN is based on 3D U-net architecture, and it includes separate Dilated and Depth-wise Convolutions. It is fully-trained on the BraTS 2018 data set, and it produces more accurate results even when compared to the winners of the BraTS 2017 competition despite having a significantly smaller amount of parameters.

• The KIPS Transactions:PartB
• /
• v.18B no.3
• /
• pp.131-138
• /
• 2011

In this paper, we compare our proposed method with previous methods for the volumetric image segmentation using level set. In order to obtain an exact segmentation, the region and boundary information of image object are used in our proposed speed function. The boundary information is defined by the gradient vector flow obtained from the gradient images and the region information is defined by Gaussian distribution information of pixel intensity in a region-of-interest for image segmentation. Also the regular term is used to remove the noise around surface. We show various experimental results of real medical volume images to verify the superiority of proposed method.

• Proceedings of the KOSOMBE Conference
• /
• v.1997 no.05
• /
• pp.119-122
• /
• 1997

In this paper, we represented the variation of heart cavity area in the space domain by 3-d rendering. We arranged the 2-d sequence of ultrasonic image acquired in the time domain as volumetric data, and extracted heart cavity region from 3-d data. For the segmentation of 3-d volume data, we extracted the cavity region using the method of expanding the cavity region that is same statistical property. By shading which is using light and object normal vector, we visualized the volume data on image plane.

• Korean Journal of Radiology
• /
• v.22 no.3
• /
• pp.435-441
• /
• 2021

Objective: To evaluate the usefulness of the ventricular volume percentage quantified using three-dimensional (3D) brain computed tomography (CT) data for interpreting serial changes in hydrocephalus. Materials and Methods: Intracranial and ventricular volumes were quantified using the semiautomatic 3D threshold-based segmentation approach for 113 brain CT examinations (age at brain CT examination ≤ 18 years) in 38 patients with hydrocephalus. Changes in ventricular volume percentage were calculated using 75 serial brain CT pairs (time interval 173.6 ± 234.9 days) and compared with the conventional assessment of changes in hydrocephalus (increased, unchanged, or decreased). A cut-off value for the diagnosis of no change in hydrocephalus was calculated using receiver operating characteristic curve analysis. The reproducibility of the volumetric measurements was assessed using the intraclass correlation coefficient on a subset of 20 brain CT examinations. Results: Mean intracranial volume, ventricular volume, and ventricular volume percentage were 1284.6 ± 297.1 cm³, 249.0 ± 150.8 cm³, and 19.9 ± 12.8%, respectively. The volumetric measurements were highly reproducible (intraclass correlation coefficient = 1.0). Serial changes (0.8 ± 0.6%) in ventricular volume percentage in the unchanged group (n = 28) were significantly smaller than those in the increased and decreased groups (6.8 ± 4.3% and 5.6 ± 4.2%, respectively; p = 0.001 and p < 0.001, respectively; n = 11 and n = 36, respectively). The ventricular volume percentage was an excellent parameter for evaluating the degree of hydrocephalus (area under the receiver operating characteristic curve = 0.975; 95% confidence interval, 0.948-1.000; p < 0.001). With a cut-off value of 2.4%, the diagnosis of unchanged hydrocephalus could be made with 83.0% sensitivity and 100.0% specificity. Conclusion: The ventricular volume percentage quantified using 3D brain CT data is useful for interpreting serial changes in hydrocephalus.

• Proceedings of the KOSOMBE Conference
• /
• v.1996 no.11
• /
• pp.301-303
• /
• 1996

A platform is developed for fast and effective functional mapping of human brain, which can allow semi-automatically the whole processes of an image segmentation, a fusion of MR and PET images, and 3-D rendering of volumetric data, including DICOM-based image transfers from PACS archiver within a short period of time.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.1
• /
• pp.191-197
• /
• 2013

Conventional CT and MRI scans produce cross-section slices of body that are viewed sequentially by radiologists who must imagine or extrapolate from these views what the 3 dimensional anatomy should be. By using sophisticated algorithm and high performance computing, these cross-sections may be rendered as direct 3D representations of human anatomy. The 2D medical image analysis forced to use time-consuming, subjective, error-prone manual techniques, such as slice tracing and region painting, for extracting regions of interest. To overcome the drawbacks of 2D medical image analysis, combining with medical image processing, 3D visualization is essential for extracting anatomical structures and making measurements. We used the gray-level thresholding, region growing, contour following, deformable model to segment human organ and used the feature vectors from texture analysis to detect harmful cancer. We used the perspective projection and marching cube algorithm to render the surface from volumetric MR and CT image data. The 3D visualization of human anatomy and segmented human organ provides valuable benefits for radiation treatment planning, surgical planning, surgery simulation, image guided surgery and interventional imaging applications.

• Investigative Magnetic Resonance Imaging
• /
• v.19 no.2
• /
• pp.67-75
• /
• 2015

Purpose: To investigate the value of image post-processing software (FreeSurfer, IBASPM [individual brain atlases using statistical parametric mapping software]) and inversion time (TI) in volumetric analyses of the hippocampus and to identify differences in comparison with manual tracing. Materials and Methods: Brain images from 12 normal adults were acquired using magnetization prepared rapid acquisition gradient echo (MPRAGE) with a slice thickness of 1.3 mm and TI of 800, 900, 1000, and 1100 ms. Hippocampal volumes were measured using FreeSurfer, IBASPM and manual tracing. Statistical differences were examined using correlation analyses accounting for spatial interpretations percent volume overlap and percent volume difference. Results: FreeSurfer revealed a maximum percent volume overlap and maximum percent volume difference at TI = 800 ms ($77.1{\pm}2.9%$) and TI = 1100 ms ($13.1{\pm}2.1%$), respectively. The respective values for IBASPM were TI = 1100 ms ($55.3{\pm}9.1%$) and TI = 800 ms ($43.1{\pm}10.7%$). FreeSurfer presented a higher correlation than IBASPM but it was not statistically significant. Conclusion: FreeSurfer performed better in volumetric determination than IBASPM. Given the subjective nature of manual tracing, automated image acquisition and analysis image is accurate and preferable.

• Korean Journal of Radiology
• /
• v.21 no.11
• /
• pp.1256-1264
• /
• 2020

Objective: Lung segmentation using volumetric quantitative computed tomography (CT) analysis may help predict outcomes of patients with coronavirus disease (COVID-19). The aim of this study was to investigate the relationship between CT volumetric quantitative analysis and prognosis in patients with COVID-19. Materials and Methods: CT images from patients diagnosed with COVID-19 from February 18 to April 15, 2020 were retrospectively analyzed. CT with a negative finding, failure of quantitative analysis, or poor image quality was excluded. CT volumetric quantitative analysis was performed by automated volumetric methods. Patients were stratified into two risk groups according to CURB-65: mild (score of 0-1) and severe (2-5) pneumonia. Outcomes were evaluated according to the critical event-free survival (CEFS). The critical events were defined as mechanical ventilator care, ICU admission, or death. Multivariable Cox proportional hazards analyses were used to evaluate the relationship between the variables and prognosis. Results: Eighty-two patients (mean age, 63.1 ± 14.5 years; 42 females) were included. In the total cohort, male sex (hazard ratio [HR], 9.264; 95% confidence interval [CI], 2.021-42.457; p = 0.004), C-reactive protein (CRP) (HR, 1.080 per mg/dL; 95% CI, 1.010-1.156; p = 0.025), and COVID-affected lung proportion (CALP) (HR, 1.067 per percentage; 95% CI, 1.033-1.101; p < 0.001) were significantly associated with CEFS. CRP (HR, 1.164 per mg/dL; 95% CI, 1.006-1.347; p = 0.041) was independently associated with CEFS in the mild pneumonia group (n = 54). Normally aerated lung proportion (NALP) (HR, 0.872 per percentage; 95% CI, 0.794-0.957; p = 0.004) and NALP volume (NALPV) (HR, 1.002 per mL; 95% CI, 1.000-1.004; p = 0.019) were associated with a lower risk of critical events in the severe pneumonia group (n = 28). Conclusion: CRP in the mild pneumonia group; NALP and NALPV in the severe pneumonia group; and sex, CRP, and CALP in the total cohort were independently associated with CEFS in patients with COVID-19.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.592-594
• /
• 2000

변형모델(deformable model)은 볼륨의료영상(volumetric medical image)으로부터 복잡한 인체기관의 3차원적 경계를 분할해내기 위해 효과적인 방법을 제공한다. 그러나, 기존 변형모델은 초기와 의존성, 오목한 경계(concavity) 분할의 비적합성, 그리고 모델내 요소간 자체교차(self-intersection)의 제한점을 가지고 있었다. 본 연구에서는 이러한 제한점을 극복하고, 오목한 구조를 포함하는 복잡한 인체기관의 경계를 분할하기에 적합한 새로운 변형모델을 제안하였다. 제안한 변형모델은 볼륨영상 피라미드(pyramid)를 기반으로 다해상도(multiresolution)의 모델 정제화(refinement)를 수행한다. 다해상도 모델 정제화는 전역적 시셈플링(global resampling) 및 지역적 리샘플링(local resampling)를 통하여 저해상도의 모델로부터 점차 고해상도의 모델로 이동하면서 객체의 경계를 계층적으로 분할해가는 방법이다. 다해상도 모델에 의한 계층적 경계 분할은 초기화 조건에의 의존성을 극복할 수 있게할 뿐 아니라, 빠른 속도로 원하는 객체의 경계에 수렴할 수 있게 한다. 또한 지역적 리샘플링은 모델 구성요소의 정규화를 수행함으로써 객체의 오목한 부분을 성공적으로 분할할 수 있게 한다. 그리고, 제안 모델은 기존 변형모델에서 포함하는 내부 힘(internal force)과 외부 힘(external force)외에 자체교차방지 힘(non-self-intersection force)을 추가함으로서 효과적으로 모델내의 자체교차를 방지할 수 있게 하였다.