• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.05a
• /
• pp.703-706
• /
• 2004

나선형 CT 혈관촬영에서 획득한 영상의 분석를 통해서 폐색전증이 의심되는 부위를 자동으로 검출하는 방법으로, 연구 대상은 20명의 환자를 대상으로 분석하였으며 CT 검사 후 방사선과 의사가 정상소견을 받은 환자 5명과 폐색전증이 있는 판독소견을 가진 15명을 대상으로 비교 분석하였다. CT 검사하는 동안에 조영제를 투입하면, 폐색전증이 발생한 부위는 조영제 양과 분포가 불균등하여 명암값이 낮게 검출된다. 검출방법으로는 전처리 작업으로 폐영역만을 분할하고, 분할된 폐영역에서 혈관을 찾기 위해 모폴로지기법를 적용하여 세선화(thinning) 작업을 진행한다. 다음 공정으로는 경계선을 찾아 local watershed를 적용하여 혈관을 검출하고, 검출된 혈관내에서 원형모델을 적용하여 모폴로지(morphology)을 통해 국소 부위의 미세한 농도변화를 인지하여 색전이 발생한 영역을 자동검출하였다. 본 논문의 자동검출시스템에서는 색전증이 있는 경우에 true positive의 발생빈도는 case 당 4.5개가 검출되었다. 정상인의 경우에도 혈류의 흐름, 혈류의 분기점, 노이즈로 인한 false positive의 빈도는 case 당 2.6개가 발생하여 전체적으로 false positive는 5.2개가 검출되었다. 본 논문은 false positive의 비율이 높게 검출되었지만 폐영역 CT 검사의 컴퓨터지원진단시스템(computer aided diagnosis)의 향후 연구과제에 방향을 제시할 수 있을 것이라 사료된다.

• Proceedings of the Korea Contents Association Conference
• /
• 2011.05a
• /
• pp.525-526
• /
• 2011

복부 MRI 영상에서 간의 인식은 간에 존재하는 질병을 파악하는 것뿐만 아니라 간에 대한 치료 방법이나 수술 방법을 결정하는 중요한 정보를 제공한다. 따라서 본 논문은 복부 MRI 영상에서 이러한 간 영역과 간의 혈관을 자동으로 검출하는 알고리즘을 제안한다. 제안 알고리즘은 1단계로 Region Growing의 개선된 방법으로 초기 설정된 간 영역의 외곽선 정보를 이용하여 씨앗점을 설정해 간 영역을 초기 분할한다. 2단계로 분할된 영역 경계선의 지역적 최소값을 이용하여 경계선의 보정 작업을 수행한다. 이후 3단계로 추출된 간 영역을 기반으로 영상 개선 후, 인페인팅(In-painting) 기법을 활용하여 간 영역 내부의 혈관 부분을 검출하게 된다. 제안 알고리즘의 평가는 전문의의 수작업 결과와 비교하였고, 결과 82.5%의 평균 정확도를 보여 제안 알고리즘의 효과적인 간혈관 검출을 확인하였다. 향후 본 연구의 확장으로 검출된 혈관 영역의 3D Volume Rendering 연구를 수행할 예정이다.

• Journal of Korea Multimedia Society
• /
• v.12 no.8
• /
• pp.1082-1088
• /
• 2009

Fundus images are highly useful in evaluating patients' retinal conditions in diagnosing eye diseases. In particular, vessel regions are essential in diagnosing diabetes and hypertension. In this paper, we used top-hat filter to compensate for non-uniform background. Image contrast was enhanced by using contrast limited adaptive histogram equalization (CLAHE) method. Hessian matrix was next applied to detect vessel regions. Results indicate that our method is 1.3% more accurate than matched filter method. Our proposed method is expected to contribute to diagnosing eye diseases.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.397-398
• /
• 2019

망막혈관 영상에서(retinal image) 혈관의 모양 또는 생성변화를 효과적으로 검진하기 위해서 망막혈관을 자동적으로 분리하는 영상분할 기법의 개발은 매우 중요한 사안이다. 이를 위해서 주로 망막혈관영상의 잡음을 억제하고 또한 혈관의 명암대비도(contrast)를 증가시키는 전처리 과정을 거쳐서 혈관의 국부적인 화소값의 변화, 방향성을 판별하여 혈관을 자동적으로 검출하는 방법들이 제시되어왔으며 최근에는 합성곱 신경망(CNN) 딥러닝 학습모델을 활용한 망막혈관 분리 알고리즘들이 제시되고 있다.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.9
• /
• pp.99-109
• /
• 2016

This paper presents a method for generating spatiotemporal images to detect the leukocyte motions in a microvessel. By using the constraint that the leukocytes move along the contour line of a blood vessel wall, the method detects leukocyte motions and then generates spatiotemporal images. the translational motion by a movement in vivo is removed first by the template matching method. Next, a blood vessel region is detected by the automatic threshold selection method to binarize the temporal variance image, then a blood vessel wall's contour is expressed by B-spline function. With the detected blood vessel wall's contour as an initial curve, the plasma layer of the best accurate position is determined to be the spatial axis by snake. Finally, the spatiotemporal images are generated. The experimental results show the spatiotemporal images are generated effectively through the comparison of each step of three image sequences.

• Progress in Medical Physics
• /
• v.17 no.1
• /
• pp.32-39
• /
• 2006

To assess the ability of an extravasation detection accessory (EDA) system to detect clinically important extravascular injection of iodinated contrast material delivered with an automated power injector. Fifty patients referred for contrast material enhanced body computed tomography studied in a prospective, observation study in which the EDA system was used to identify and interrupt any injection associated with clinically Important extravasation. The presence or absence of extravasation was definitively established with multi-detector CT at the injection site (injection rate, $2.0{\sim}2.5$ mL/sec). There were two true positive, extravasation volumes $22{\sim}25$ mL. The EDA system had sensitivity in the detection of clinically important extravasation. The EDA system is easy to use, safe, and accurate In the monitoring of intravenous injections for extravasation, which may prove especially useful in CT applications.

• Journal of the Institute of Convergence Signal Processing
• /
• v.15 no.4
• /
• pp.128-134
• /
• 2014

In this paper, we propose a detection method of the leukocyte motions in a microvessel by using spatiotemporal image analysis. The leukocyte motions that adhere to blood vessel walls can be visualized to move along the blood vessel wall's contours in a sequence of images. In this proposal method, we use the constraint that the leukocytes move along the blood vessel wall's contours and detect the leukocyte motions by using the spatiotemporal image analysis method. The generated spatiotemporal image is processed by a special-purpose orientation-selective filter and then subsequent grouping processes are done. The subsequent grouping processes select and group the leukocyte trace segments among all the segments obtained by simple thresholding and skeletonizing operations. Experimental results show that the proposed method can stably detect the leukocyte motions even when multiple leukocyte traces intersect each other.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.9
• /
• pp.2013-2018
• /
• 2011

Coronary angioplasty and coronary artery bypass graft, both are for the treatment of myocardial infarction widely used methods. For these procedures, there are especially difficulties in stenosis of blood vessels to diagnose accurately. To remedy this problem, by several researchers by using edge detection to detect stenosis of blood vessels has been studying. However, the results of using these methods vary defend on the vascular structure and the quality of the image. In this study, to improve these problems, the new algorithm is proposed. The proposed algorithm consists of methods to detect bifurcation of blood vessels and its ending point by using multi sampling, threshold and fuzzy algorithm. To evaluate the performance of the proposed algorithm, angiography was used for the different results of the blood vessels of the proposed algorithm, and the result was effective in detecting bifurcation of blood vessels and its ending point.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.9
• /
• pp.137-142
• /
• 2016

When injecting intravenously of CT inspection, the effusion of the contrast meium can induce the tissue damage with the blood vessel outside. We detect extravasation which is generated in the course where we inject the contrast medium into the blood vessel. And we use the bolus tracking system for the detection of that. By using MPR and VR images, moreover we detected the extravasation in order to prevent the tissue damage. In order to detect the effusion of the contrast medium, we used 16-MDCT and 64-MDCT. Three dimensional images about the outflow of the blood vessel can provide the treatment information which is important in the patient treatment. Moreover we applied the image processing technique in order to improve sharpness between contrast media and organization. And sharpness and contrast was improved.

• The KIPS Transactions:PartB
• /
• v.12B no.5 s.101
• /
• pp.543-552
• /
• 2005

This paper proposes the extraction of coronary arteries based on DSA(Digital Subtraction Angiography) through a texture analysis of background in the angiography. DSA is a well established modality for the visualization of coronary arteries. DSA involves the subtraction of a mask image - an image of the heart before injection of contrast medium - from live image. However, this technique is sensitive to the movement of background and can result to a wrong detection by the variance of background gray-level intensity between two images. Therefore, this paper solves a structural problem resulted from a background movement bV selecting an image which has the least difference of movement through an analysis of the similarity of background texture and proposes a method to extract only the blood vessel efficiently through local gray-level correction of the selected image. Using the coronary angiogram of 5 patients clinical data, we proved that the proposed method has the lower false-detection rate, approximately $2\%$, and the higher accuracy than the existing methods.