Browsing by Issue Date, starting with "2020-05-14"

Now showing 1 - 20 of 5029

Deep learning assisted detection of glaucomatous optic neuropathy and potential designs for a generalizable model
(2020-05-14)
Ko, Yu-Chieh
;
Wey, Shih-Yu
;
Chen, Wei-Ta
;
Chang, Yu-Fan
;
Chen, Mei-Ju
Purpose To evaluate ways to improve the generalizability of a deep learning algorithm for identifying glaucomatous optic neuropathy (GON) using a limited number of fundus photographs, as well as the key features being used for classification. Methods A total of 944 fundus images from Taipei Veterans General Hospital (TVGH) were retrospectively collected. Clinical and demographic characteristics, including structural and functional measurements of the images with GON, were recorded. Transfer learning based on VGGNet was used to construct a convolutional neural network (CNN) to identify GON. To avoid missing cases with advanced GON, an ensemble model was adopted in which a support vector machine classifier would make final classification based on cup-to-disc ratio if the CNN classifier had low-confidence score. The CNN classifier was first established using TVGH dataset, and then fine-tuned by combining the training images of TVGH and DrishtiGS datasets. Class activation map (CAM) was used to identify key features used for CNN classification. Performance of each classifier was determined through area under receiver operating characteristic curve (AUC) and compared with the ensemble model by diagnostic accuracy. Results In 187 TVGH test images, the accuracy, sensitivity, and specificity of the CNN classifier were 95.0%, 95.7%, and 94.2%, respectively, and the AUC was 0.992 compared to the 92.8% accuracy rate of the ensemble model. For the Drishti-GS test images, the accuracy of the CNN, the fine-tuned CNN and ensemble model was 33.3%, 80.3%, and 80.3%, respectively. The CNN classifier did not misclassify images with moderate to severe diseases. Class-discriminative regions revealed by CAM co-localized with known characteristics of GON. Conclusions The ensemble model or a fine-tuned CNN classifier may be potential designs to build a generalizable deep learning model for glaucoma detection when large image databases are not available.
【完美求職系列講座】傻瓜理財法～畢業存出一桶金
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
【反毒宣導月活動】「藥物濫用防制認知檢測」問卷填答抽獎活動
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
Resource provisioning towards OPEX optimization in horizontal edge federation
(2020-05-15)
Baghban, Hojjat
;
Huang, Ching-Yao
;
Hsu, Ching-Hsien
;
電子工程學系及電子研究所
;
電機工程學系
Energy consumption is a key performance metric in multi-access edge computing (MEC) system. Therefore, minimizing consumed energy cost is critically essential. The 5G networks deal with edge computing resources so that the operator would face with power supply limitation. Therefore, it may not be able to provide sufficient resources to ever-increasing user's requests. One way to compensate such limitation is to form horizontal edge federation (HEF) so that all participant can share the resource capacities as well as request workloads. Energy efficient and ultra-low latency HEF involves the setting of critical factor in each participant: offloading ratios. The decided offloading ratios must provide satisfactory service level to meet latency and physical resource types capacity constraints demanded by requests. Our proposed problem is an energy efficient operational cost (eOPEX) optimization problem. In this paper, we formulate it as a mixed integer linear program and demonstrate that the problem is NP-hard and proposed a federated multidimensional fractional knapsack based algorithm (FMFK) as our approach. The result shows that the horizontal edge federation based on the FMFK performs better and saves the more e-OPEX as well as serving more input requests compare with the non-federation approach. The experimental results show that our approach save about 40% of e-OPEX specially for high latency sensitive application requests in hotspot zone. It shows around 50% e-OPEX saving in the case of high computation unit cost compare with non-federation approach.
陳右穎教授團隊利用「石墨烯」打造奈米生物晶片，讓診斷治療合而為一
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
如何縮短診斷與治療的時間差，一直是亟待克服的難題。陽明醫工系陳右穎教授團隊利用奈米材料「石墨烯」，重新打造神經探針晶片深
【懷生社講座】歡迎餵食～兩隻偽浪浪的徒步環台記
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
Timing jitter reduction of passively Q-switched solid-state lasers by coupling resonance between pumping and firing rates
(2020-05-15)
Chen, Y. F.
;
Chien, P. Y.
;
Lee, C. C.
;
Huang, K. F.
;
Liang, H. C.
The coupling resonance between pumping and firing rates is originally proposed to achieve the timing jitter reduction of a Nd : YVO4 laser passively Q-switched with a saturable absorber. When the pumping rate is higher than the spontaneous emission rate, it is experimentally confirmed that the pulse firing rate can be fractionally locked with the pumping rate by controlling the pump power. The locking characteristics of the firing rate display a variety of complex plateaus that can be excellently manifested with the sine-circle map. From numerical analyses, the coupling strength can be verified to be effectively enhanced by reducing the duty cycle of the pumping rate. (C) 2020 Optical Society of America
NiCo2O4/graphene quantum dots (GQDs) for use in e fficient electrochemical energy devices: An electrochemical and X-ray absorption spectroscopic investigation
(2020-05-15)
Chang, Han-Wei
;
Fu, Ji-Xuan
;
Huang, Yu-Cheng
;
Lu, Ying-Rui
;
Kuo, Chun-Hong
NiCo2O4/graphene quantum dots (GQDs) were synthesized by the incorporation of GQDs into NiCo2O4 as electrode materials for use in electrochemical energy devices. The electronic structure and the electrochemical performance of NiCo2O4/GQDs were investigated by X-ray absorption spectroscopy (XAS) and electrochemical measurements. The rich redox active sites were revealed on the NiCo2O4/GQDs, which were formed by the strong CeOeNi/Co interaction between NiCo2O4 and GQDs. The synergetic double-layer capacitive and pseudocapacitive properties were suggested to provide favorable electrochemical characteristics. In/ex situ XAS elucidates local electronic and chemical environments and provides insights into the energy storage/conversion mechanisms. Analytical results demonstrate that the NiCo2O4/GQDs have strong oxygen-carbon bridges and have more Ni active sites than does NiCo2O4, favoring metal-to-ligand charge transfer at the interface between NiCo2O4 and GQDs, critically influencing electrochemical performance. This work contributes importantly to the field by elucidating an energy storage/conversion mechanism for electrochemical energy devices.
A permutation-and-backtrack approach for reliability evaluation in multistate information networks
(2020-05-15)
Chen, Shin-Guang
;
Lin, Yi-Kuei
;
工業工程與管理學系
;
Department of Industrial Engineering and Management
Nowadays, people live heavily on information systems to handle their business and schedules. A reliable information system helps people smoothly and efficiently deal with their daily-life affairs. An information system usually consists of connected software and hardware components, and can be modeled by connected graphs. In network theory, such graphs are called multistate information networks. Therefore, the reliability of an information system can be evaluated by network theory, namely network reliability. The most popular approach to evaluate network reliability is the three-stage-approach (TSA), which involves (a) finding all minimal paths (MPs), (b) finding all d-system vectors (d-SVs), and (c) calculating the union probability of d-SVs for reliability. Approaches for creating all MPs has been reported. There is a need for a more efficient approach to generate all d-SVs for such calculation. In this article, an approach based on permutation-and-backtrack facility is proposed to generate all such d-SVs. The efficiency of this approach is generally accepted to be superior to the existing approaches. The proposed approach provides us an efficient, and simple way to speed up TSA as a more powerful tool for evaluating the reliability of multistate information networks. (C) 2020 Elsevier Inc. All rights reserved.
【陽明圖花樹報報Ｘ萬方知識查找】
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
2020「天下USR大學公民」調查，陽明名列公立中型大學第一
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
《天下雜誌》日前公布2020年「天下USR大學公民」調查結果，陽明在公立一般中型大學排行中名列第一。對於創校以來即以偏鄉醫療為使
Berezinskii-Kosterlitz-Thouless transition in an Al superconducting nanofilm grown on GaAs by molecular beam epitaxy
(2020-05-15)
Su, Guan-Ming
;
Wu, Bi-Yi
;
Fan, Yen-Ting
;
Kumar, Ankit
;
Chang, Chau-Shing
We have performed extensive transport experiments on a 4 nm thick aluminum (Al) superconducting film grown on a GaAs substrate by molecular beam epitaxy (MBE). Nonlinear current-voltage (I-V) measurements on such a MBE-grown superconducting nanofilm show that V similar to I-3, which is evidence for the Berezinskii-Kosterlitz-Thouless (BKT) transition, both in the low-voltage (T-BKT 1.97 K) and high-voltage regions (T-BKT 2.17 K). In order to further study the two regions where the I-V curves are BKT-like, our experimental data are fitted to the temperature-induced vortices/antivortices unbinding model as well as the dynamical scaling theory. It is found that the transition temperature obtained in the high-voltage region is the correct T-BKT as confirmed by fitting the data to the aforementioned models. Our experimental results unequivocally show that I-V measurements alone may not allow one to determine T-BKT for superconducting transition. Therefore, one should try to fit one's results to the temperature-induced vortices/antivortices unbinding model and the dynamical scaling theory to accurately determine T-BKT in a two-dimensional superconductor.
陽明聯手國體大打造防疫訓練系統，助選手備戰東澳
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
新型冠狀病毒造成全球大傳染，也導致運動訓練與比賽中斷，包括東京奧運都被迫延後一年舉行。陽明與國立體育大學聯手辦理「東京奧
《私人戰爭》（A Private War）
(國立陽明大學, 2020-05-15)
黃柏威
;
陽明電子報編輯小組
傳統與現代並陳，合校校徽設計初稿公布
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
秘書室與學生會5月6日晚間共同主辦「陽明交大校徽識別系統設計說明會」，會中公布Ａ、Ｂ兩組分別呈現傳統與現代兩種風格，以及由
藝文方盒子「第二屆藍花楹文學獎與攝影獎」揭曉
(國立陽明大學, 2020-05-15)
鍾庭宜
;
陽明電子報編輯小組
今年藍花楹文學獎各組首獎得獎人與作品為：小說組／公衛所博士班Joseph《梵林墩事件》、散文組／醫學五洪邦喩《鼠》、新詩組／醫學四許育寧《C1的愛情》；攝影獎首獎則是視文所碩三丁傳益《北方車站的聚會》。陽明九景文學獎首獎是醫學二張廷恩《老朋友》，攝影組首獎則是牙醫二林廷揚《向陽》。
Bio-inorganic nanocomposites of lysozyme amyloid fibrils and magnetic nanoparticles of different shape anisotropy
(2020-05-15)
Lackova, Veronika
;
Tomasovicova, Natalia
;
Olejniczak, Andrzej
;
Zakutanska, Katarina
;
Majorosova, Jozefina
The present study describes the interaction of an assorted shape of magnetic nanoparticles with lysozyme amyloid fibrils by atomic force microscopy and magnetization measurements. Mixture stability was verified by settling behaviour of NPs dispersions. A fibril solution doped with spherically shaped magnetic nanoparticles reveals the effect of particles' adsorption on the amyloid fibrils' surface. However, when spindle-like magnetic nanoparticles were doped, it led to zero interaction activity, which confesses subtle difference in fibrillary interaction with spherical particle that can be explained by diverse size of nanoparticles. Moreover, spherical MNPs showed excellent stability in sLAF, the spindle MNPs were unstable and exhibited limited dispersion.
Purely in-plane ferroelectricity in monolayer SnS at room temperature
(2020-05-15)
Higashitarumizu, Naoki
;
Kawamoto, Hayami
;
Lee, Chien-Ju
;
Lin, Bo-Han
;
Chu, Fu-Hsien
2D van der Waals ferroelectrics have emerged as an attractive building block with immense potential to provide multifunctionality in nanoelectronics. Although several accomplishments have been reported in ferroelectric switching for out-of-plane ferroelectrics down to the monolayer, a purely in-plane ferroelectric has not been experimentally validated at the monolayer thickness. Herein, an in-plane ferroelectricity is demonstrated for micrometer-size monolayer SnS at room temperature. SnS has been commonly regarded to exhibit the odd-even effect, where the centrosymmetry breaks only in the odd-number layers to exhibit ferroelectricity. Remarkably, however, a robust room temperature ferroelectricity exists in SnS below a critical thickness of 15 layers with both an odd and even number of layers, suggesting the possibility of controlling the stacking sequence of multilayer SnS beyond the limit of ferroelectricity in the monolayer. This work will pave the way for nanoscale ferroelectric applications based on SnS as a platform for in-plane ferroelectrics. Out-of-plane ferroelectricity is usually unstable in the two dimensional limit due to the presence of the depolarization field. Here, the authors successfully circumvent this issue by growing mu m-sized SnS monolayers that exhibit in-plane ferroelectricity that is stable at room temperature.
【歡慶45周年，5/31校慶活動】
(國立陽明大學, 2020-05-15)
陽明電子報編輯小組
Highly sensitive and selective sensing properties of modified green phosphorene monolayers towards SF6 decomposition gases
(2020-05-15)
Singh, Amit
;
Bae, Hyeonhu
;
Lee, Seunghan
;
Shabbiri, Khadija
;
Hussain, Tanveer
Gas sensing properties of pristine, defective and heteroatom substituted green phosphorene (GP) monolayers towards SF6 components such as H2S, SO2, SO2F2, SOF2 are investigated by using van der Waals corrected density function theory calculations. The adsorptions of H2S and SOF2 on pristine GP are found to be exothermic with weak binding energies (E-b) of- 0.065 and - 0.017 eV respectively, whereas the interactions of SO2 and SO2F2 are endothermic. We found that various vacancy defects and heteroatom substitutions enhance the E-b values of H2S, SO2, SO2F2 and SOF2 to - 0.84, - 1.04, - 0.40 and - 0.73 eV, respectively which are within desired range for efficient gas sensing applications. Presence of selected substituents (Al, B, C, S, Si) and vacancy defects result into measurable variations in electronic properties of GP upon the exposure of sulphur containing gases. We found that B- and S-doped GP systems show selectivity towards H2S and SO2, with E-b values of - 0.57eV for H2S and - 1.03eV for SO2. We believe that modified GP has potential to be used as excellent sensing material towards SF6 decomposition gases.