• 대한전기학회논문지:시스템및제어부문D
• /
• 제52권1호
• /
• pp.64-75
• /
• 2003

Medical image segmentation is the process by which an original image is partitioned into some homogeneous regions like bones, soft tissues, etc. This study demonstrates an automatic medical image segmentation technique based on independent component analysis. Independent component analysis is a generalization of principal component analysis which encodes the higher-order dependencies in the input in addition to the correlations. It extracts statistically independent components from input data. Use of automatic medical image segmentation technique using independent component analysis under the assumption that medical image consists of some statistically independent parts leads to a method that allows for more accurate segmentation of bones from CT data. The result of automatic segmentation using independent component analysis with square test data was evaluated using probability of error(PE) and ultimate measurement accuracy(UMA) value. It was also compared to a general segmentation method using threshold based on sensitivity(True Positive Rate), specificity(False Positive Rate) and mislabelling rate. The evaluation result was done statistical Paired-t test. Most of the results show that the automatic segmentation using independent component analysis has better result than general segmentation using threshold.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권7호
• /
• pp.1706-1725
• /
• 2024

The unique U-shaped structure of U-Net network makes it achieve good performance in image segmentation. This network is a lightweight network with a small number of parameters for small image segmentation datasets. However, when the medical image to be segmented contains a lot of detailed information, the segmentation results cannot fully meet the actual requirements. In order to achieve higher accuracy of medical image segmentation, a novel improved U-Net network architecture called multi-scale encoder-decoder U-Net+ (MEDU-Net+) is proposed in this paper. We design the GoogLeNet for achieving more information at the encoder of the proposed MEDU-Net+, and present the multi-scale feature extraction for fusing semantic information of different scales in the encoder and decoder. Meanwhile, we also introduce the layer-by-layer skip connection to connect the information of each layer, so that there is no need to encode the last layer and return the information. The proposed MEDU-Net+ divides the unknown depth network into each part of deconvolution layer to replace the direct connection of the encoder and decoder in U-Net. In addition, a new combined loss function is proposed to extract more edge information by combining the advantages of the generalized dice and the focal loss functions. Finally, we validate our proposed MEDU-Net+ MEDU-Net+ and other classic medical image segmentation networks on three medical image datasets. The experimental results show that our proposed MEDU-Net+ has prominent superior performance compared with other medical image segmentation networks.

• 융복합기술연구소 논문집
• /
• 제11권1호
• /
• pp.1-5
• /
• 2021

Deep learning based methods achieve state-of-the-art accuracy, however, they typically rely on supervised training with large labeled datasets. It is known in many medical applications that labeling medical images requires significant expertise and much time, and typical hand-tuned approaches for data augmentation fail to capture the complex variations in such images. This paper proposes a 3D image augmentation method to overcome these difficulties. It allows us to enrich diversity of training data samples that is essential in medical image segmentation tasks, thus reducing the data overfitting problem caused by the fact the scale of medical image dataset is typically smaller. Our numerical experiments demonstrate that the proposed approach provides significant improvements over state-of-the-art methods for 3D medical image segmentation.

• 전자공학회논문지
• /
• 제50권8호
• /
• pp.225-231
• /
• 2013

의료영상분할은 다양한 의료영상처리를 수행하기에 앞서 먼저 수행되어야 하는 영상처리 기술이다. 그래서 빠르고 정확한 의료영상분할이 요구되고 있으며 다양한 의료영상분할 방법이 연구되고 있다. 의료영상에는 특성이 유사한 다양한 장기가 존재하기 때문에 분할영역의 정확한 판단이 필요하다. 그러나 의료영상은 장기의 일부가 작게 촬영되는 경우가 발생된다. 이 경우에는 분할영역을 판단하기 위한 정보가 부족하게 되며 그 결과 분할과정에서 분할영역이 제거된다. 본 논문에서는 볼륨 데이터와 선형 방정식을 이용하여 작은 영역에서의 분할결과를 개선하였다. 제안한 방법의 성능을 확인하기 위하여 흉부 CT 영상의 폐 분할을 수행하였다. 실험 결과, 의료영상의 분할 정확도는 0.978에서 0.981로 표준편차는 0.281에서 0.187로 개선되는 것을 확인하였다.

• 한국멀티미디어학회논문지
• /
• 제14권11호
• /
• pp.1383-1391
• /
• 2011

향상된 기능을 가진 최신 의료장비들의 등장으로 하드웨어 성능에 부합하는 효과적인 영상처리 및 분석의 중요성이 부각되고 있으며, 2차원 의료 영상처리 및 3차원 영상 재구성에 관한 많은 연구들이 진행되고 있다. 본 논문은 흉부 CT 영상을 사용하여 신체 장기를 단계별로 분할 하였으며, 분할된 결과 영상을 3차원으로 재구성 하였다. 다양한 영상분할 방법중 영역 확장법 및 효과적인 분할을 위해 선명화와 감마 조절등과 같은 영상 향상 기법을 적용하였으며, 기관지를 포함한 폐, 기관지, 폐 등의 순서로 영상을 분할하였다. 분할된 신체 장기 영상을 VTK를 사용하여 3차원 영상으로 재구성 하였으며, 병변 진단을 위한 2차원 및 3차원 의료 영상 처리와 분석에 활용될 것으로 판단된다.

• 한국멀티미디어학회논문지
• /
• 제25권5호
• /
• pp.757-768
• /
• 2022

In this paper, we proposed a system that visualizes a hologram device in 3D by utilizing the CT image segmentation function based on artificial intelligence deep learning. The input axial CT medical image is converted into Sagittal and Coronal, and the input image and the converted image are divided into 3D volumes using ResUNet, a deep learning model. In addition, the volume is created by segmenting the tumor region in the segmented liver image. Each result is integrated into one 3D volume, displayed in a medical image viewer, and converted into a video. When the converted video is transmitted to the hologram device and output from the device, a 3D image with a sense of space can be checked. As for the performance of the deep learning model, in Axial, the basic input image, DSC showed 95.0% performance in liver region segmentation and 67.5% in liver tumor region segmentation. If the system is applied to a real-world care environment, additional physical contact is not required, making it safer for patients to explain changes before and after surgery more easily. In addition, it will provide medical staff with information on liver and liver tumors necessary for treatment or surgery in a three-dimensional manner, and help patients manage them after surgery by comparing and observing the liver before and after liver resection.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권3호
• /
• pp.1167-1187
• /
• 2020

With the development of medical imaging technology, image registration has been widely used in the field of disease diagnosis. The registration between different modal images of brain magnetic resonance (MR) is particularly important for the diagnosis of brain diseases. However, previous registration methods don't take advantage of the prior knowledge of bilateral brain symmetry. Moreover, the difference in gray scale information of different modal images increases the difficulty of registration. In this paper, a multimodal medical image registration method based on image segmentation and symmetric self-similarity is proposed. This method uses modal independent self-similar information and modal consistency information to register images. More particularly, we propose two novel symmetric self-similarity constraint operators to constrain the segmented medical images and convert each modal medical image into a unified modal for multimodal image registration. The experimental results show that the proposed method can effectively reduce the error rate of brain MR multimodal medical image registration with rotation and translation transformations (average 0.43mm and 0.60mm) respectively, whose accuracy is better compared to state-of-the-art image registration methods.

• 산업공학
• /
• 제21권4호
• /
• pp.394-402
• /
• 2008

The objective of this paper is to design segmentation algorithm for applying the breast ultrasound image to CAD(Computer Aided Diagnosis). This study is conducted after understanding limits, used algorithm and demands of CAD system by interviewing with a medical doctor and analyzing related works based on a general CAD framework that is consisted of five step-establishment of plan, analysis of needs, design, implementation and test & maintenance. Detection function of CAD is accomplished by Canny algorithm and arithmetic operations for segmentation. In addition to, long computing time is solved by extracting ROI (Region Of Interests) and applying segmentation technical methods based morphology algorithm. Overall course of study is conducted by verification of medical doctor. And validity and verification are satisfied by medical doctor's confirmation. Moreover, manual segmentation of related works, restrictions on the number of tumor and dependency of image resolution etc. was solved. This study is utilized as a support system aided doctors' subjective diagnosis even though a lot of future studies is needed for entire application of CAD system.

• 대한의용생체공학회:의공학회지
• /
• 제26권6호
• /
• pp.407-412
• /
• 2005

Convenience is one of the most important factors in medical image segmentation. Convenience is defined by compiling opinions from radiologists, and can be described as controllable maximum automation on the condition of producing only accurate results. The components of convenience are inclusive automation and inclusive modification. Inclusive modification consists of verify-and-confirm, undo-redo, exchange of segmentation methods, and intelligent modification tools. Inclusive automation is composed of automatic selection of a method, automatic selection of a confident segment, and automated chores. The convenient segmentation tool has been developed to segment X-ray images for orthopedic surgery, and has received an excellent evaluation from radiologists.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제14권4호
• /
• pp.201-210
• /
• 2010

We propose a new robust and accurate method for the numerical solution of medical image segmentation. The modified Allen-Cahn equation is used to model the boundaries of the image regions. Its numerical algorithm is based on operator splitting techniques. In the first step of the splitting scheme, we implicitly solve the heat equation with the variable diffusive coefficient and a source term. Then, in the second step, using a closed-form solution for the nonlinear equation, we get an analytic solution. We overcome the time step constraint associated with most numerical implementations of geometric active contours. We demonstrate performance of the proposed image segmentation algorithm on several artificial as well as real image examples.