Keh-Shih Chuang

Keh-Shih Chuang

Auto-generated profile, based on publicly available information

Keh-Shih Chuang

Department of Nuclear Science, National Tsing-Hua University, Taiwan

Research Focus & Metrics

Academic Career Length

25Years (2002-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2006 in Computerized Medical Imaging and Graphics
Verified
Journal Article
Closed access
KCKeh-Shih Chuang
Keh-Shih Chuang
HTHong-Long Tzeng
Hong-Long Tzeng
SCSharon Chen
Sharon Chen
JWJay Wu
Jay Wu
TCTzong-Jer Chen
Tzong-Jer Chen

1 October 2005 in Computerized Medical Imaging and Graphics
Verified
Journal Article
Closed access
KCKeh-Shih Chuang
Keh-Shih Chuang
MJMeei-Ling Jan
Meei-Ling Jan
JWJay Wu
Jay Wu
JLJeng-Chang Lu
Jeng-Chang Lu
SCSharon Chen
Sharon Chen
YFYing-Kai Fu
Ying-Kai Fu

1 June 2003 in Journal of Digital Imaging
Verified
Journal Article
Closed access
PDF file
TCTzong-Jer Chen
Tzong-Jer Chen
KCKeh-Shih Chuang
Keh-Shih Chuang
JWJay Wu
Jay Wu
SCSharon C. Chen
Sharon C. Chen
IHIng-Ming Hwang
Ing-Ming Hwang
MJMeei-Ling Jan
Meei-Ling Jan

9 July 2015 in Physica Medica
Verified
Journal Article
Closed access
PDF file
CSCheng-Ting Shih
Cheng-Ting Shih
YCYuan-Jen Chang
Yuan-Jen Chang
JHJui-Ting Hsu
Jui-Ting Hsu
KCKeh-Shih Chuang
Keh-Shih Chuang
SCShu-Jun Chang
Shu-Jun Chang
JWJay Wu
Jay Wu

12 November 2021 in Diagnostics
Verified
Journal Article
Open access
PDF file
YNYu-Ching Ni
Yu-Ching Ni
FTFan-Pin Tseng
Fan-Pin Tseng
MPMing-Chyi Pai
Ming-Chyi Pai
IHIng-Tsung Hsiao
Ing-Tsung Hsiao
KLKun-Ju Lin
Kun-Ju Lin
AIAlzheimer’s Disease Neuroimaging Initiative
Alzheimer’s Disease Neuroimaging Initiative
The correct differential diagnosis of dementia has an important impact on patient treatment and follow-up care strategies. Tc-99m-ECD SPECT imaging, which is low cost and accessible in general clinics, is used to identify the two common types of dementia, Alzheimer’s disease (AD) and Lewy body dementia (LBD). Two-stage transfer learning technology and reducing model complexity based on the ResNet-50 model were performed using the ImageNet data set and ADNI database. To improve training accuracy, the three-dimensional image was reorganized into three sets of two-dimensional images for data augmentation and ensemble learning, then the performance of various deep learning models for Tc-99m-ECD SPECT images to distinguish AD/normal cognition (NC), LBD/NC, and AD/LBD were investigated. In the AD/NC, LBD/NC, and AD/LBD tasks, the AUC values were around 0.94, 0.95, and 0.74, regardless of training models, with an accuracy of 90%, 87%, and 71%, and F1 scores of 89%, 86%, and 76% in the best cases. The use of transfer learning and a modified model resulted in better prediction results, increasing the accuracy by 32% for AD/NC. The proposed method is practical and could rapidly utilize a deep learning model to automatically extract image features based on a small number of SPECT brain perfusion images in general clinics to objectively distinguish AD and LBD.

1 April 2011 in American Journal of Roentgenology
Verified
Journal Article
Closed access
SDShang-Lung Dong
Shang-Lung Dong
TCTieh-Chi Chu
Tieh-Chi Chu
GLGong-Yau Lan
Gong-Yau Lan
YLYung-Chien Lin
Yung-Chien Lin
YYYu-Hsiu Yeh
Yu-Hsiu Yeh
KCKeh-Shih Chuang
Keh-Shih Chuang
OBJECTIVE. The currently used model breast for mammographic dosimetry assessment lacks the flexibility to change dimensions. The aim of this study was to develop an adjustable model breast for mammographic dosimetry assessment of Taiwanese women. MATERIALS AND METHODS. A retrospective review of 4226 craniocaudal (CC) views was conducted. The geometry of the model breast was defined as a semielliptical cylinder. Breast parameters, including compressed breast thickness, chest wall–to-nipple distance, compressed breast width, and percentage glandular content, were measured and analyzed. To validate the adjustable model breast, 44 mammograms were obtained. The expected values from the adjustable model breast were compared with the measured values. RESULTS. The average values of compressed breast thickness, chest wall–to-nipple distance, compressed breast width, and percentage glandular content of the women studied were 4.1 cm, 6.9 cm, 16.9 cm, and 54%, respectively. Variations of chest wall–to-nipple distance, compressed breast width, and percentage glandular content can be expressed as functions of compressed breast thickness, and the adjustable model breast developed was based on compressed breast thickness. The average area of the CC view obtained is a factor of 0.81 lower than that defined by the American College of Radiology protocol. For validation, the difference in average values between the expected and measured did not exceed 0.5 cm in breast dimensions and 6% in percentage glandular content. CONCLUSION. Compressed breast thickness is useful for quantifying dimensions and percentage glandular content of a model breast. The adjustable model breast developed in this study can offer greater flexibility in the determination of breast dimensions for mammographic dosimetry assessment of Taiwanese women.