• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.29-40
• /
• 2022

In this paper, we studied a system that detects and analyzes the pathological features of diabetic retinopathy using Mask R-CNN and a Random Forest classifier. Those are one of the deep learning techniques and automatically diagnoses diabetic retinopathy. Diabetic retinopathy can be diagnosed through fundus images taken with special equipment. Brightness, color tone, and contrast may vary depending on the device. Research and development of an automatic diagnosis system using artificial intelligence to help ophthalmologists make medical judgments possible. This system detects pathological features such as microvascular perfusion and retinal hemorrhage using the Mask R-CNN technique. It also diagnoses normal and abnormal conditions of the eye by using a Random Forest classifier after pre-processing. In order to improve the detection performance of the Mask R-CNN algorithm, image augmentation was performed and learning procedure was conducted. Dice similarity coefficients and mean accuracy were used as evaluation indicators to measure detection accuracy. The Faster R-CNN method was used as a control group, and the detection performance of the Mask R-CNN method through this study showed an average of 90% accuracy through Dice coefficients. In the case of mean accuracy it showed 91% accuracy. When diabetic retinopathy was diagnosed by learning a Random Forest classifier based on the detected pathological symptoms, the accuracy was 99%.

• Journal of IKEEE
• /
• v.12 no.3
• /
• pp.158-166
• /
• 2008

Medical ultrasound has benefits in mobility and safety than any other medical techniques such as X-ray, CT and MRI but has speckle noise which decrease the ability of an observer to distinguish the fine details in diagnostic examination. But simple removing of speckle often causes losing boundary information. Then, in this paper, we presented a novel neuro-fuzzy method which could remove speckle efficiently without loss of boundary information. Proposed method consists of image clustering by fuzzy algorithm and image processingby neural networks which was learned by back propagation. From the experiments for simulation image and real ultrasound image, we could verify the proposed method.

• Journal of the Korean Society of Radiology
• /
• v.3 no.3
• /
• pp.13-21
• /
• 2009

PET/CT is a machine for imaging in vivo functions or metabolic activities after the administration of radiopharmaceuticals labeled with radioisotope emitting positrons in the body. Recently the number of PET/CT installed in Korean medical institutions is increasing rapidly. In response, the number of PET/CT tests to be used in the diagnosis and treatment of tumors is also increasing every year, and this is increasing the necessity for developing the methods of PET/CT performance evaluation and quality control. Among the test items for the performance evaluation and quality control of PET/CT suggested in NU 2-2007, this study examined spatial resolution test, sensitivity test, image quality, attenuation accuracy & scatter correction test, scatter fraction, count losses and randoms test and accuracy( correction for count losses and randoms).

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10a
• /
• pp.139-142
• /
• 2006

본 논문에서는 햅틱 피드백과 스테레오 비쥬얼 큐를 혼합한 다중 감각 기반의 지능형 3차원 형상 분석 방법을 소개한다. 지능형 형상 분석 방법은 3차원 모델의 구조에 대한 보다 상세한 정보를 제공한다. 특히 의료 분야에 사용될 경우 전문가의 개입을 최소화하여 질병 진단 및 치료 등에 사용될 수 있다. 본 연구에서는, MRI나 CT 영상으로부터 생성된 3차원 매개변수형 모델을 이용하여 유사 모델 집단을 대표하는 통계 형상을 구축한 후, SVM (Support Vector Machine) 학습 알고리즘을 이용하여 두 집단간 형상 차이를 분석한다. 3차원 형상에 대한 신속한 시각적 이해와 직관적 조작감은 물체 표면의 형상 변화를 분석하는데 효과적으로 사용될 수 있다. 본 논문에서는 물체 조작 및 관찰 등의 작업을 수행할 때, 햅틱 피드백과 스테레오 비쥬얼 큐를 혼합한 인터랙션 기법을 사용하여 공간감과 깊이감을 향상시켜 형상 분석 결과를 효과적으로 분석한다. 본 연구에서는 해마, 관상 동맥, 뇌와 같은 인체 장기를 실험 데이터로 사용하여 제안한 SVM 기반의 분석 방법과 인터랙션 환경의 성능을 평가한다. 본 연구에서 구현한 SVM 기반 이진 분류기는 두 집단간 형상 차이를 효과적으로 분석하며, 또한 다중 감각 인터랙션은 사용자가 분석 결과를 관찰하고 카메라 및 형상을 효율적으로 조작하는 데 도움을 준다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.4
• /
• pp.81-87
• /
• 2016

With the rapid growth of the up-to-date smartphone and wireless network technologies, huge and various types of smart applications using these technologies are actively developed recently. Especially, multiplex types of smart applications using smartphone are developed and diffused with the rapid development of the ubiquitous sensor network technology using various sensors and the mobile computing technology that enables us to get network services at any time in any places. In this paper, we design a smart application that can accurately diagnose and process the current state of the local environment, objects, and persons remotely based on the context information such as local ecology, circumstances, medical or healthcare records and realtime sound or motion pictures using up-to-date samrtphone technology on the USN based mobile computing environment.

• Journal of the Korea Convergence Society
• /
• v.7 no.4
• /
• pp.1-7
• /
• 2016

Structural MRI(sMRI) imaging is used to extract morphometric features after Grey Matter (GM), White Matter (WM) for several univariate and multivariate method, and Cerebro-spinal Fluid (CSF) segmentation. A new approach is applied for the diagnosis of very mild to mild AD. We propose the classification method of Alzheimer disease patients from normal controls by combining morphometric features and Gaussian Mixture Models parameters along with MMSE (Mini Mental State Examination) score. The combined features are fed into Multi-kernel SVM classifier after getting rid of curse of dimensionality using principal component analysis. The experimenral results of the proposed diagnosis method yield up to 96% stratification accuracy with Multi-kernel SVM along with high sensitivity and specificity above 90%.

• Journal of the Society of Cosmetic Scientists of Korea
• /
• v.43 no.3
• /
• pp.247-254
• /
• 2017

Optical Coherence Tomography (OCT) is a non-invasive imaging method that utilizes the optical scattering and interference for visualizing the surface as well as cross-sectional structures of tissue. OCT has been used for diagnosing diseases in early stage in various medical fields, but an application in cosmetics is still at early stage. In this study, OCT was adopted to evaluate skin texture and wrinkle. Results showed similar patterns of evaluation with PRIMOS in the assessment using replica. In addition, OCT produced smaller errors at different angles compared to the PRIMOS in the assessment using 3-dimensional models of wrinkles. The resolution of the image was also high enough to differentiate the images of before and after the application of makeup products. Possible use of OCT in the evaluation of fine wrinkle assessment was studied in this research. Further development of methods is necessary to provide more evidences of the effectiveness.

• 한국노년학
• /
• v.30 no.4
• /
• pp.1197-1211
• /
• 2010

This research was concucted to present a model of advance directives(AD) when a patient, who is in consciousness, shows a preference for an end of life care as an act of preparing for an uncertain situation that may arise in the forseeable future. The subjects of the research are 383 doctors/nurese and adults, who live in six cities and provinces, to investigate the status of AD, attitude regarding a meaningless life-prolonging treatment, and moreover, an understanding of and a preference for AD. The research was done by the well-structured questionnaire. Also, SPSS 14.0 is used to analyse the collected data, focused on frequency analysis, avearage and standard deviation, X2 test. As the results of the study, the most of the surveyed doctors/nurese knew DNR orders and AD and a few of them used DNR orders and AD practically. Also, the result shows that there is a negative conception of meaningless life-prolonging treatment among the responents, in addition, most of them agreed upon the idea of introducing AD to Korea, filling it out and making it legally effective. As a method of making AD out, the respondents wanted to use a form that mixed living will with an Power of Attorney in a document. Also, considering the appropriate time, respondents prefered when they are diagnosed with terminal illness. At the moment, the introductory model for AD, which is suitable for the Korean culture and current situation is presented based on the result of this research. In the future, other researches should deal with specific measures that can lead to a social consensus to adopt AD in Korea.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2015.08a
• /
• pp.229.1-229.1
• /
• 2015

생체 시료인 세포나 조직을 분석을 위해 임의로 파괴하거나 훼손하지 않은 본래의 상태에서 세포에 존재하는 다양한 생체분자 물질의 질량과 조성을 분석하고 영상화할 수 있는 대기압 표면 질량분석 이미징 기술을 개발했다. 생체 시료의 표면을 질량 분석을 하기 위해서는 대기압 분위기에서 시료에 열적 손상이 없는 조건으로 시편의 이온화 및 탈착 과정이 이루어지게 하기 위해 저온 대기압 탈착/이온화원으로 저온대기압 플라즈마 젯과 펨토초 적외선 레이저를 결합하여 대기압 이온화원을 제작하였다. 기존에 잘 알려진 저온 대기압 플라즈마 젯 소자는 유리관에 방전기체를 흘려주고 전극에 고전압을 인가하는 방식으로 제작했으며, 또 다른 대기압 이온화원으로서 근적외선 대역의 고출력 펨토초 레이저 빔을 현미경용 대물렌즈로 집속하여 생체시료에 조사시켰다. 수백 나노미터에서 수 마이크로미터 수준으로 빔을 집속할 수 있는 펨토초 레이저는 금나노로드의 도움으로 생체 시료를 매우 작은 수준으로 탈착하는 데 주로 사용하며, 수십 마이크로미터에서 수 밀리미터 정도의 크기를 가지는 저온 대기압 플라즈마 젯은 탈착된 물질을 이온화시키는데 사용하여, 이 두 가지 이온화원을 결합하여 이온화원으로 사용한다. 시료에서 발생한 이온을 질량분석기 입구까지 잘 끌고 갈 수 있도록 이온 전달관을 설계하고 보조펌프를 장착 사용한다. 이렇게 자체 개발한 대기압 이온화원을 상용 질량분석기기와 결합하여 대기압 분위기에서 시료의 표면을 질량분석할 수 있는 시스템과 측정 기술을 개발했다. 현미경 스테이지에 정밀 2-D 자동 스캐닝 스테이지를 장착하여 질량분석 정보에 공간 정보를 더할 수 있는 질량분석 이미징 기술 방법을 개발하여 생체 시편의 질량분석 이미징을 얻었다. 수분을 포함하는 생채시료로부터 단백질, 지질, 대사물질을 직접 분리하여 분석하는 이 새로운 질량분석법은 기존의 분석법에 비해 훨씬 더 많은 생체분자 정보를 얻을 수 있으며 공간정보를 더해 영상화할 수 있는 큰 장점이 있다. 대기압 표면 질량분석 기술은 생체시료를 파괴해서 용액화할 필요도 없으며, 진공 챔버에 넣기 위해 필요한 복잡한 전처리 과정 단계를 간략화 할 수 있으며 최종적으로는 살아있는 세포나 생체 조직도 정량 분석이 가능하여 생명과학 및 의료진단 분야에서 응용할 수 있는 분야는 무궁무진할 것이다.

• Journal of Intelligence and Information Systems
• /
• v.11 no.1
• /
• pp.17-33
• /
• 2005

This paper presents the proposed a classifier of liver cirrhotic step using MR(magnetic resonance) imaging and hierarchical neural network. The data sets for classification of each stage, which were normal, 1type, 2type and 3type, were analysis in the number of data was 231. We extracted liver region and nodule region from T1-weight MR liver image. Then objective interpretation classifier of liver cirrhotic steps. Liver cirrhosis classifier implemented using hierarchical neural network which gray-level analysis and texture feature descriptors to distinguish normal liver and 3 types of liver cirrhosis. Then proposed Neural network classifier learned through error back-propagation algorithm. A classifying result shows that recognition rate of normal is $100\%$, 1type is $82.8\%$, 2type is $87.1\%$, 3type is $84.2\%$. The recognition ratio very high, when compared between the result of obtained quantified data to that of doctors decision data and neural network classifier value. If enough data is offered and other parameter is considered this paper according to we expected that neural network as well as human experts and could be useful as clinical decision support tool for liver cirrhosis patients.