• ETRI Journal
• /
• v.46 no.3
• /
• pp.513-525
• /
• 2024

Deep neural networks trained on labeled medical data face major challenges owing to the economic costs of data acquisition through expensive medical imaging devices, expert labor for data annotation, and large datasets to achieve optimal model performance. The heterogeneity of diseases, such as Alzheimer's disease, further complicates deep learning because the test cases may substantially differ from the training data, possibly increasing the rate of false positives. We propose a reconstruction-based self-supervised anomaly detection model to overcome these challenges. It has a dual-subnetwork encoder that enhances feature encoding augmented by skip connections to the decoder for improving the gradient flow. The novel encoder captures local and global features to improve image reconstruction. In addition, we introduce an entropy-based image conversion method. Extensive evaluations show that the proposed model outperforms benchmark models in anomaly detection and classification using an encoder. The supervised and unsupervised models show improved performances when trained with data preprocessed using the proposed image conversion method.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.4
• /
• pp.211-216
• /
• 2017

This paper proposes image processing techniques that improve usability and performance in a diagnostic system of the contrast-enhanced ultrasonography. For a methodology for visualizing diagnostic parameter data in an ultrasonic medical image, an expression of transition time data with successive pixel values and a method of generating a lesion diagnostic parameter image with four categorized values are presented. We also introduce a MRF-based image enhancement technique to eliminate noises from generated parametric images. Such parametric image generation technique can overcome the difficulty of discriminating dynamic change in patterns in the ultrasonography. The technique clarifies the contour of the region in the original image and facilitates visual determination of the characteristics of the lesion through four colors. With regard to this MRF-based image enhancement, we define the energy function of consecutive pixel values and develop a technique to optimize it, and the usability of the proposed theory is examined through experiments with medical images.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.1
• /
• pp.59-64
• /
• 2023

With the recent development of IT technology, medical image processing technology is also widely used in the dental field, and the treatment effect is enhanced by using 3D data such as CT. In this paper, open source libraries such as ITK and VTK are introduced to develop dental medical image processing software, and how to use them to develop dental medical image processing software centering on 3D CBCT. In ITK, basic algorithms for medical image processing are implemented, so the image processing pipeline can be quickly implemented, and the desired algorithm can be easily implemented as a filter by the developer. The developed algorithm is linked with VTK to implement the visualization function. The developed SW can be used for dental diagnosis and treatment that overcomes the limitations of 2D images..

• Journal of Internet Computing and Services
• /
• v.24 no.4
• /
• pp.85-92
• /
• 2023

Deep learning applications are increasingly being leveraged for disease detection tasks in medical imaging modalities such as X-ray, Computed Tomography (CT), and Magnetic Resonance Imaging (MRI). Most data-centric deep learning challenges necessitate the use of supervised learning methodologies to attain high accuracy and to facilitate performance evaluation through comparison with the ground truth. Supervised learning mandates a substantial amount of image and label sets, however, procuring an adequate volume of medical imaging data for training is a formidable task. Various data augmentation strategies can mitigate the underfitting issue inherent in supervised learning-based models that are trained on limited medical image and label sets. This research investigates the enhancement of a deep learning-based rib fracture segmentation model and the efficacy of data augmentation techniques such as left-right flipping, rotation, and scaling. Augmented dataset with L/R flipping and rotations(30°, 60°) increased model performance, however, dataset with rotation(90°) and ⨯0.5 rescaling decreased model performance. This indicates the usage of appropriate data augmentation methods depending on datasets and tasks.

• Journal of the Korean Society for Precision Engineering
• /
• v.28 no.7
• /
• pp.834-850
• /
• 2011

In this paper, a novel tissue engineering scaffold design method based on triply periodic minimal surface (TPMS) is proposed. After generating the hexahedral elements for a 3D anatomical shape using the distance field algorithm, the unit cell libraries composed of triply periodic minimal surfaces are mapped into the subdivided hexahedral elements using the shape function widely used in the finite element method. In addition, a heterogeneous implicit solid representation method is introduced to design a 3D (Three-dimensional) bio-mimetic scaffold for tissue engineering from a sequence of computed tomography (CT) medical image data. CT image of a human spine bone is used as the case study for designing a 3D bio-mimetic scaffold model from CT image data.

• Journal of Biomedical Engineering Research
• /
• v.44 no.1
• /
• pp.11-18
• /
• 2023

Currently, studies to predict the risk of rectal cancer surgery select MRI image slices based on the clinical experience of surgeons. The purpose of this study is to semi-automatically select and classify 2D MRI image slides to predict the risk of rectal cancer surgery using biomarkers. The data used were retrospectively collected MRI imaging data of 50 patients who underwent laparoscopic surgery for rectal cancer at Gachon University Gil Medical Center. Expert-selected MRI image slices and non-selected slices were screened and radiomics was used to extract a total of 102 features. A total of 16 approaches were used, combining 4 classifiers and 4 feature selection methods. The combination of Random Forest and Ridge performed with a sensitivity of 0.83, a specificity of 0.88, an accuracy of 0.85, and an AUC of 0.89±0.09. Differences between expert-selected MRI image slices and non-selected slices were analyzed by extracting the top five significant features. Selected quantitative features help expedite decision making and improve efficiency in studies to predict risk of rectal cancer surgery.

• Journal of the Korea Computer Graphics Society
• /
• v.30 no.4
• /
• pp.19-29
• /
• 2024

In this paper, we examine the effectiveness of StyleGAN-generated images for data augmentation in training deep neural networks for medical image classification. We apply StyleGAN data augmentation to train VGG-16 networks for pneumonia diagnosis from chest X-ray images and focal liver lesion classification from abdominal CT images. Through quantitative and qualitative analyses, our experiments reveal that StyleGAN data augmentation expands the outer class boundaries in the feature space. Thanks to this expansion characteristics, the StyleGAN data augmentation can enhance classification performance when properly combined with real training images.

• Korea Journal of Hospital Management
• /
• v.18 no.4
• /
• pp.18-38
• /
• 2013

This study aims to analyze the effect of quality of health care on perceived value, patient satisfaction and revisit intention. Especially, it was focused on outdoor environment, indoor environment, admission procedure, hospital image, service quality of physicians nurses medical technicians medical staff that patients perceived. Inpatients and outpatients were selected from three hospital in D city Questionnaire survey was employed to collect data from the subjects. For inpatients, indoor environment, admission procedure, hospital image and service quality of physicians have an effect on perceived value. Admission procedure, hospital image and service quality of physicians nurses medical technicians has an effect on the patient satisfaction. Hospital image and service quality of physicians nurses medical technicians have an effect on revisit intention. Perceived value have an effect on the patient satisfaction. Perceived value have an effect on revisit intention. Patient satisfaction have an effect on revisit intention. For outpatients, Admission procedure, hospital image and service quality of physicians medical technicians have an effect on perceived value. Indoor environment, hospital image and service quality of physicians medical technicians medical staff has an effect on the patient satisfaction. Indoor environment, hospital image and service quality of physicians medical technicians have an effect on revisit intention. Perceived value have an effect on the patient satisfaction. Perceived value have an effect on revisit intention. Patient satisfaction have an effect on revisit intention. They should evaluate customer satisfaction on their services and analyze various factors that affect on it to improve middle hospitals.

• The Journal of the Korean dental association
• /
• v.43 no.1 s.428
• /
• pp.41-52
• /
• 2005

Information Technology has extended its scope to the medical field as well as dental field. Like medical field, network ststem for dental field requires acquisition, storage, and display of images. However, unlike the medical field, the system to integrate several information including medical images has not been developed according to industrial standard for management of digital image for medical use, so called DICOM conformance. which makes the digital environment in dental field more and more difficult and expensive for this standardization and comfortable communication in LAN and WAN. To solve this problem, the DICOM encoder and server has to be developed because the DICOM file can be easily retrieved with patient's information from the DICOM server in the system as DICOM file has the standard specification to integrate the patient's information. The information including image and other discrete data can be easily integrated in DICOM file and can be used without any difficulty for precise diagnosis and for contribution to the decision making for each treatment protocol. Therefore, the system composed of DICOM encoder and server in dental practive for DICOM file must be developed with prudent consideration of the several strategic factors: I) Enhanced diagnostic capability through the integrated information of image and clinical data. ii) Clinician-friendly interface to simulate the systemic treatment procedure in clinical practice iii) Implementation of multidisciplinary treatment protocol The development of DICOM encoder and server based on these strategic considerations will provide paperless and filmless hospital environments by the seamless integration and management of patient's history, several clinical data and clinical images through image processing for quantitative analysis. The system also allows clinicians to provide more predictable dental care for the patients.

• Proceedings of the Korean Society of Medical Physics Conference
• /
• 2003.09a
• /
• pp.68-68
• /
• 2003

SPM(Statistical Parametric Mapping)은 개인의 서로 다른 조건에서의 뇌영상이나 환자그룹간의 functional image data를 비교분석 하는데 강력한 기능을 가지고 있다.