• Journal of the Korea Computer Graphics Society
• /
• v.28 no.2
• /
• pp.11-19
• /
• 2022

Recently, various researches on medical image generation have been suggested, and it becomes crucial to accurately evaluate the quality and diversity of the generated medical images. For this purpose, the expert's visual turing test, feature distribution visualization, and quantitative evaluation through IS and FID are evaluated. However, there are few methods for quantitatively evaluating medical images in terms of fidelity and diversity. In this paper, images are generated by learning a chest CT dataset of non-small cell lung cancer patients through DCGAN and PGGAN generative models, and the performance of the two generative models are evaluated in terms of fidelity and diversity. The performance is quantitatively evaluated through IS and FID, which are one-dimensional score-based evaluation methods, and Precision and Recall, Improved Precision and Recall, which are two-dimensional score-based evaluation methods, and the characteristics and limitations of each evaluation method are also analyzed in medical imaging.

• Journal of the Korean Society of Radiology
• /
• v.17 no.4
• /
• pp.565-572
• /
• 2023

This study was conducted to analyze the status radiological imaging examinations assessment in Korea medical institutions conducted in the public sector and suggest a direction for improvement. Among the assessment of medical institutions, the main assessment related to radiographic imaging examinations are the certification evaluation of medical institutions and the adequacy assessment of radiographic imaging examinations. The certification evaluation of medical institutions evaluates the image inspection operation process, provision of accurate results, and compliance with safety management procedures. In the assessment of adequacy of radiographic imaging examinations, structural indicators related to manpower and equipment, patient evaluation implementation rate, and exposure reduction programs were included. However, for safer and higher-quality radiological imaging examinations, it is necessary to increase the participation rate of medical institutions in certification evaluations. In addition, it is necessary to improve the manpower indicator, and incentive payments can be considered to induce quality improvement of medical institutions in the future. Integrated management of radiation exposure at the national level should also be carried out simultaneously.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.10 no.6
• /
• pp.713-720
• /
• 2015

As the hand-held mobile devices are widely used, they are recently coming into convergence with medical diagnostic systems. Furthermore, the wireless mobile Internet and the various kinds of communication devices are rapidly coming into wide use converging with medical technology. The mobile communication environments can make people get more health care services beyond space and time. In this paper, we implement and evaluate the mobile client and the medical diagnostic information server for transmitting, searching and updating the medical diagnostic information. The DICOM CT image and the compressed JPEG 2000 CT image are statistically evaluated by t-test performance whether those images are clinically appropriate. In the case of the DICOM CT image, we realize that the average value is relatively more appropriate to the clinical diagnosis than the JPEG 2000 CT image.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.903-906
• /
• 2021

인공지능 기술 발전으로, 의료영상 분야에서도 딥러닝 기반 질병 진단 연구가 활발히 진행되고 있다. 딥러닝 모델 개발 시, 학습 데이터 품질은 모델의 성능과 신뢰성에 매우 큰 영향을 미친다. 그러나 의료 분야의 경우 도메인 지식에 대한 진입 장벽이 높아 개발자가 학습에 사용되는 의료영상 데이터의 품질을 평가하기 어렵다. 이로 인해, 많은 의료영상 분야에서는 각 분야의 특성(질병의 종류, 관찰 아나토미 등)에 따른 영상 품질 평가 방법을 제시해왔다. 그러나 기존의 방법은 특정 질병에 초점이 맞춰져, 일반화된 품질 평가 기준을 제시하고 있지 않다. 따라서 본 논문에서는 대부분의 흉부 질환을 진단하기 위한 흉부 X선 영상의 품질을 평가할 수 있는 기준을 제안한다. 우선, 흉부 X선 영상을 대상으로 관찰된 영역인 심장, 횡격막, 견갑골, 폐 등을 분할하여, 3D 히스토그램을 기반으로 각 영역별 통계적인 정밀 품질 평가 기준을 제안한다. 본 연구에서는 JSRT, Chest 14의 오픈 데이터셋을 활용하여 적용 실험을 수행하였으며, 민감도는 97.6%, 특이도는 92.8%의 우수한 성능을 확인하였다.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2007.06a
• /
• pp.369-369
• /
• 2007

The physical and visual evaluation of the identical X-rays are analyzed for medical image clarity with CD-RAD Phantom on this study. The method of measurement is to research CD-RAD by X-rays and to acquire images through image processing equipment, the image analyses are carried out by physical evaluation with statistical method through CD-RAD analyser program, and the visual evaluation of the identical X-rays is carried out by blind test for 20 observers. The result of it is that IQF value of the physical evaluation of Contrast-detail curve is 25 and IQF value of the visual evaluation is 30, so it is revealed that the physical evaluation is superior to the visual one. The special qualities of medical images have much importance of the transmission capacity of information to the image analyser, so it is concluded that 0비ective methods of the physical and visual analyses should be carried out side by side.

• Journal of the Institute of Convergence Signal Processing
• /
• v.18 no.2
• /
• pp.76-83
• /
• 2017

In this study, the original medical image, DICOM file, was converted into TIFF, BITMAP, GIF, JPEG image file, and then the conversion loss ratio according to the image compression and conversion process was quantitatively evaluated using Origin pro and ICY image analysis program. As the evaluation method, 50% MTF, structural similarity index, MSE, RMSE, maximum signal - to - noise ratio and so on were evaluated. The TIFF image file showed the same result as DICOM image in all experimental groups, Image file format. In this study, we propose a new method for evaluating the quality of digital images by applying original evaluation program such as Origin pro or ICY medical image analysis program. Is expected to be used as research data in the field of medical image processing, and TIFF image file showing the same result as DICOM image in the basic research field using digital medical image and evaluation program that does not support DICOM file Therefore, it is believed that it will help to secure reliability in digital medical image processing research using image file.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.904-905
• /
• 2023

본 논문에서는 의료 영상 생성을 위한 Med-StyleGAN2를 제안한다. 생성적 적대 신경망은 이미지 생성에는 효과적이지만, 의료 영상 생성에는 한계점을 가지고 있다. 따라서 본 연구에서는 의료 영상 생성에 특화된 StyleGAN 기반 학습 모델을 제안한다. 이는 다양한 의료 영상 어플리케이션에 활용할 수 있으며, 생성된 의료 영상에 대한 정량적, 정성적 평가를 수행함으로써 의료 영상 생성 분야의 발전 가능성에 대해 연구한다.

• Korean Journal of Digital Imaging in Medicine
• /
• v.2 no.1
• /
• pp.150-157
• /
• 1996

• Journal of the Institute of Convergence Signal Processing
• /
• v.20 no.3
• /
• pp.132-137
• /
• 2019

The amount of data generated from medical images is increasingly exceeding the limits of professional visual analysis, and the need for automated medical image analysis is increasing. For this reason, this study evaluated the classification and accuracy according to the presence or absence of tumor using Inception V3 deep learning model, using MRI medical images showing normal and tumor findings. As a result, the accuracy of the deep learning model was 90% for the training data set and 86% for the validation data set. The loss rate was 0.56 for the training data set and 1.28 for the validation data set. In future studies, it is necessary to secure the data of publicly available medical images to improve the performance of the deep learning model and to ensure the reliability of the evaluation, and to implement modeling by improving the accuracy of labeling through labeling classification.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.05a
• /
• pp.595-598
• /
• 2003

본 논문에서는 방사선투영영상을 기존 압축방법인 JPEG 압축과 새로운 표준으로 채택중인 JPEG2000을 적용하여 압축율 및 영상의 품질을 비교 실험하였다. 기존의 의료영상압축 표준의 하나인 JPEG 압축은 압축비율이 높아짐에 따라 블륵킹 현상의 발생으로 원 영상이 회손되는 압축의 한계를 인식하고 있다. 따라서 원 영상의 보호와 압축율 증가의 두 가지 면을 만족시키기 위해 Wavelet 을 사용하는 JPEG2000을 실험 평가하여 의료영상압축에 적용하고자 한다. 실험대상으로 환자 10명 정상인 10명의 투영영상을 사용하였으며, 영상의 품질, 손상도 등을 평가하기 위해 PSNR( Peak Signal to Noise Ratio )과 판독의에 의한 ROC( Receiver Operating Characteristic )분석을 실행하였다. 실험결과, 영상의 품질, 손상도를 평가하기 위한 PSNR 은 15:1 압축에서 $46.05{\pm}1.1dB$의 값을 얻었으며, JPEG의 같은 압축비율에 비해 $1.78{\pm}0.1dB$의 값이 높음을 알 수 있었다. 종합적으로 3명의 판독의에 의해 ROC 분석을 실행한 결과 15:1의 압축비율에서 압축비율과 품질을 종합하였을 때 진단에 적합한 최적 압축비율임을 보였다.