Ming-Tse Kuo

Dr. Ming-Tse Kuo

Department of Ophthalmology, Kaohsiung Chang Gung Memorial Hospital
orcid

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
12 July 2021 in Diagnostics
Verified
Journal Article
Open access
PDF file
NHNing Hung
Ning Hung
KSKuan-Yu Shih
Kuan-Yu Shih
CLChihung Lin
Chihung Lin
Ming-Tse KuoMing-Tse Kuo
Ming-Tse Kuo
YHYih-Shiou Hwang
Yih-Shiou Hwang
CHChing-Hsi Hsiao
Ching-Hsi Hsiao
In this study, we aimed to develop a deep learning model for identifying bacterial keratitis (BK) and fungal keratitis (FK) by using slit-lamp images. We retrospectively collected slit-lamp images of patients with culture-proven microbial keratitis between 1 January 2010 and 31 December 2019 from two medical centers in Taiwan. We constructed a deep learning algorithm consisting of a segmentation model for cropping cornea images and a classification model that applies different convolutional neural networks (CNNs) to differentiate between FK and BK. The CNNs included DenseNet121, DenseNet161, DenseNet169, DenseNet201, EfficientNetB3, InceptionV3, ResNet101, and ResNet50. The model performance was evaluated and presented as the area under the curve (AUC) of the receiver operating characteristic curves. A gradient-weighted class activation mapping technique was used to plot the heat map of the model. By using 1330 images from 580 patients, the deep learning algorithm achieved the highest average accuracy of 80.0%. Using different CNNs, the diagnostic accuracy for BK ranged from 79.6% to 95.9%, and that for FK ranged from 26.3% to 65.8%. The CNN of DenseNet161 showed the best model performance, with an AUC of 0.85 for both BK and FK. The heat maps revealed that the model was able to identify the corneal infiltrations. The model showed a better diagnostic accuracy than the previously reported diagnostic performance of both general ophthalmologists and corneal specialists.

2 September 2020 in Scientific Reports
Verified
Journal Article
Open access
PDF file
Ming-Tse KuoMing-Tse Kuo
Ming-Tse Kuo
BHBenny Wei-Yun Hsu
Benny Wei-Yun Hsu
YYYu-Kai Yin
Yu-Kai Yin
PFPo-Chiung Fang
Po-Chiung Fang
HLHung-Yin Lai
Hung-Yin Lai
VTVincent S. Tseng
Vincent S. Tseng
Fungal keratitis (FK) is the most devastating and vision-threatening microbial keratitis, but clinical diagnosis a great challenge. This study aimed to develop and verify a deep learning (DL)-based corneal photograph model for diagnosing FK. Corneal photos of laboratory-confirmed microbial keratitis were consecutively collected from a single referral center. A DL framework with DenseNet architecture was used to automatically recognize FK from the photo. The diagnoses of FK via corneal photograph for comparing DL-based models were made in the Expert and NCS-Oph group through a majority decision of three non-corneal specialty ophthalmologist and three corneal specialists, respectively. The average percentage of sensitivity, specificity, positive predictive value, and negative predictive value was approximately 71, 68, 60, and 78. The sensitivity was higher than that of the NCS-Oph (52%, P < .01), whereas the specificity was lower than that of the NCS-Oph (83%, P < .01). The average accuracy of around 70% was comparable with that of the NCS-Oph. Therefore, the sensitive DL-based diagnostic model is a promising tool for improving first-line medical care at rural area in early identification of FK.

1 October 2009 in Investigative Ophthalmology & Visual Science
Verified
Journal Article
Open access
PDF file
HCHung-Chi Jesse Chen
Hung-Chi Jesse Chen
HCHsiang-Ling Chen
Hsiang-Ling Chen
Jui-Yang LaiJui-Yang Lai
Jui-Yang Lai
CCChih-Chun Chen
Chih-Chun Chen
YTYueh-Ju Tsai
Yueh-Ju Tsai
DMDavid Hui-Kang Ma
David Hui-Kang Ma
Purpose. To determine the expression of differentiation and progenitor cell markers in corneal tissues that previously underwent autologous cultivated oral mucosal epithelial transplantation (COMET). methods. Four eyes from three alkaline-injured patients and one thermally injured patient underwent COMET to promote re-epithelialization or corneal reconstruction. Between 10 and 22 months (mean, 14.2 ± 5.5 months [SD]) after COMET, the corneal tissues were obtained after penetrating keratoplasty (n = 1) or autologous limbal transplantation (n = 3). Immunoconfocal microscopy for keratin (K)3, -12, -4, -13, and -8; connexin (Cx)43; MUC5AC; laminin-5; pan-p63; ABCG2; and p75 was performed in those specimens as well as in the oral mucosa and cultivated oral mucosal epithelial cells (OMECs). results. All four specimens were unanimously positive for K3, -4, and -13 but negative for K8 and MUC5AC, suggesting that the keratinocytes were oral mucosa-derived. However, peripheral K12 staining was positive only in patient 2, suggesting a mixed oral and corneal epithelium in that case. Cx43 staining in the basal epithelium was negative in patients 1, 2, and 3, but was positive in patient 4. Small, compact keratinocytes in the basal epithelium preferentially expressed pan-p63, ABCG2, and p75. Although the staining of pan-p63 and ABCG2 tended to be more than one layer, signal for p75 was consistently localized only to the basal layer. conclusions. The study demonstrated the persistence of transplanted OMECs in human corneas. In addition, small, compact cells in the basal epithelium preferentially expressed the keratinocyte stem/progenitor cell markers, which may be indicative of the engraftment of the progenitor cells after transplantation.

17 April 2009 in Eye
Verified
Journal Article
Closed access
DMDavid Hui-Kang Ma
David Hui-Kang Ma
Ming-Tse KuoMing-Tse Kuo
Ming-Tse Kuo
YTY-J Tsai
Y-J Tsai
HCH-C J Chen
H-C J Chen
XCX-L Chen
X-L Chen
KLK-K Lin
K-K Lin

Verified
Journal Article
Open access
PDF file
Ming-Tse KuoMing-Tse Kuo
Ming-Tse Kuo
PFPo-Chiung Fang
Po-Chiung Fang
TCTsai-Ling Chao
Tsai-Ling Chao
ACAlexander Chen
Alexander Chen
YLYu-Hsuan Lai
Yu-Hsuan Lai
CTChia-Yi Tseng
Chia-Yi Tseng
Sjögren syndrome (SS) or dry eye disease (DED) is one of the most complicated ocular surface diseases. The goal of this study is to elucidate the relationship of the changes in clinical indices of tear film (TF) homeostasis with respect to tear components to allow for SS-DED monitoring and avoid stably controlled SS-DED patients from re-entering a vicious cycle. This prospective case-control study compared stable SS-DED patients with non-SS-DED control from several aspects, including clinical indices for TF homeostasis, 2 DED diagnostic biomarkers (MMP-9 and lactoferrin), and the proteome of flush tears. Compared with non-SS-DED controls, stably controlled SS-DED subjects had less tear secretion and higher ocular surface inflammation, a higher concentration ratio of tear MMP-9/lactoferrin, a more diverse tear proteome, and lower spectral intensities of lipocalin-1, lacritin, and prolactin-inducible protein among the abundant tear proteins. For stable SS-DED patients, the concentration ratio of tear MMP-9/lactoferrin and the corrected lipocalin-1 signal was positively correlated with ocular inflammation and TF stability, respectively. MMP-9 released from stressed ocular surface epithelium and lipocalin-1 secreted from the energetic lacrimal gland are two tear biomarkers responding well to TF homeostasis. The tear proteomics approach through flush tears is a promising method for monitoring SS-DED patients with a standardized sampling procedure and lactoferrin-corrected analysis.

12 April 2018 in BMC Ophthalmology
Verified
Journal Article
Open access
PDF file
CWChia-Yu Wang
Chia-Yu Wang
RHRen-Wen Ho
Ren-Wen Ho
PFPo-Chiung Fang
Po-Chiung Fang
HYHun-Ju Yu
Hun-Ju Yu
CCChun-Chih Chien
Chun-Chih Chien
Ming-Tse KuoMing-Tse Kuo
Ming-Tse Kuo
To investigate function and morphology of the meibomian gland (MG) in patients with thyroid eye disease (TED). In this prospective case series study, patients with unilateral or bilateral TED were consecutively enrolled. The diagnosis of TED was based on the typical orbital findings and/or radiographic evidence. The disease activity of TED was classified according to the clinical activity score (CAS). Degrees of lagophthalmos and exophthalmos, blinking rates, and results of the Schirmer test 1 were also recorded. All patients completed the SPEED questionnaire and underwent MG assessment, including lipid layer thickness (LLT), MG dropout (MGd), and MG expression. In total 31 eyes from 17 patients with unilateral or bilateral TED were included. Patients were divided into inactive TED (CAS 0−1; 20 eyes from 11 patients) and active TED (CAS 2−3, 11 eyes from 6 patients) groups. MGd was significantly more severe in the active TED than the inactive TED group [Median (Inter-quartile region): 3.0 (2.0−3.0) vs. 2.0 (1.0−2.0) degree, P = 0.04]. However, patients with active TED had thicker LLT than those with inactive TED (90.0 [80.0−100.0] vs. 65.0 [47.8−82.5] nm, P = 0.02), and LLT was positively correlated with lagophthalmos (r = 0.37, P = 0.04). Patients with active TED had more severe MGd, but thicker LLT. Active TED may cause periglandular inflammation of MGs, leading to MGd, but compensatory secretion from residual MGs and lagophthalmos-induced forceful blinking might temporarily release more lipids over the tear film.