http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Numerical Objective Assessment Using Structural Similarity for Diffuse Optical Reconstructed Images
Vicky Mudeng(비키 무댕 ),Se-woon Choe(최세운) 한국정보통신학회 2021 한국정보통신학회 종합학술대회 논문집 Vol.25 No.2
본 연구의 목표는 확산 광학 단층 촬영에 대한 기준 영상을 사용하여 동질성과 이질성을 분리하기 위한 재구성된 영상들간의 수치적 평가를 위해 구조적 유사성 지수에 기초한 알고리즘을 개발한다. 글로벌 지오메트리 및 관심 영역 평가는 유사성을 산출하기 위해 측정되었으며, 그 결과 구조적 유사성 지수의 평균이 모델 내부에 가시적 포함 여부를 판단할 수 있는 잠재적 성능을 나타낸다는 것을 알 수 있으며, 구조적 유사성 지수는 유방 구조 정보를 평가하기 위한 이미지 평가를 지원 가능한 것으로 확인 되었다. The work within this study develops an algorithm based on the structural similarity index to assess numerically between reconstructed images with a reference image to separate the homogeneity and heterogeneity for diffuse optical tomography. Global geometry and region of interest assessment have been measured to yield the similarity. The results indicate that the mean of structural similarity index shows potential performance to distinguish between visible and invisible inclusion inside the model. Therefore, the structural similarity index may promise to assist the image assessment for evaluating breast structural information.
Breast Masses Evaluation Using Deep Learning for Mammogram Imaging: A Pilot Study
Vicky Mudeng(비키 무뎅),Eonjin Lee(이언진),Se-woon Choe(최세운) 한국통신학회 2022 한국통신학회 학술대회논문집 Vol.2022 No.2
Following the improvements in the field of deep learning, the applications of convolutional neural networks emerge promising to apply in medical image analysis. This study demonstrates the classification of breast masses related to benign or malignant cancers from the mammographic image dataset of the digital database for screening mammography (DDSM). 50-layer residual network (ResNet50) architecture, adaptive gradient algorithm (Adagrad) optimizer, a learning rate scheduler, and a fine-tuning strategy are employed when training the convolutional neural network model. Additionally, to balance the dataset, image augmentations are employed. According to the accuracy and loss of the test dataset, the results indicate that the classifier in this study is reliable to classify breast cancer in the category of benign or malignant from the breast masses. However, more studies are necessary to expand the comprehensive understanding relative to the developed model in this research.
An Alternative Athlete Monitoring System Using Cost-Effective Inertial Sensing Instrumentation
Mudeng Vicky,Hakim Imam M.,Suprapto Sena S.,Choe Se-woon 대한전기학회 2022 Journal of Electrical Engineering & Technology Vol.17 No.6
An examination of the human gait is feasible using inertial sensing. The embedded accelerometer and gyroscope in an inertial measurement unit can evaluate physical activity-based sports and this unit is relatively aff ordable compared to global positioning systems or video recording quantifi cation. This study developed a cost-eff ective sports monitoring investigation method with an inertial sensor attached to the right leg of the athletes. In total, four parameters were simultaneously tracked to assess the entire sensor performance in real-time. The accelerometer measured the typical leg angle when walking and running, whereas the gyroscope processed the raw data to obtain the stride frequency from the time-domain data. Moreover, a comparison between the accelerometer and gyroscope was presented while simultaneously attaining the signal to convert the time-domain data to frequency results. Also, the number of strides and linear velocity was expressed as results in this study. To confi rm the results, a statistical hypothesis test was implemented for all obtained results. The results indicated that the inertial sensing instrumentation used in this study is promising and could be an aff ordable alternative option for a sports monitoring system.