Search

Diabetic kidney disease (DKD) is a critical microvascular complication of diabetes. With the continuous increase in the prevalence of diabetes since 2000, the prevalence of DKD has also been increasing in past years. The prevalence of DKD among individuals with type 2 diabetes in Taiwan increased from 13.32% in 2000 to 17.92% in 2014. The cumulative incidence of DKD among individuals with type 1 diabetes in Taiwan was higher than 30% during 1999–2012. DKD is the leading cause of end‐stage renal disease (ESRD), with a prevalence of approximately 45% in a population on chronic dialysis in Taiwan. Among individuals with type 2 diabetes, the prevalence of ESRD in the receipt of dialysis also increased from 1.32% in 2005 to 1.47% in 2014. Risk factors for DKD development are age, race, family history, hyperglycemia, hypertension, dyslipidemia, dietary patterns, and lifestyles. Prognostic factors that aggravate DKD progression include age, family history, sex, glycemic control, blood pressure (BP), microvascular complications, and atherosclerosis. This review summarizes updated information on the onset and progression of DKD, particularly in the Taiwanese population. Translating these epidemiological features is essential to optimizing the kidney care and improving the prognosis of DKD in Asian populations., This review summarizes updated information on the onset and progression of DKD, particularly in Taiwan. Translating these epidemiological features is essential to optimizing kidney care and improving the prognosis of DKD in Asian populations.

Background Food-mediated transmission of pathological viruses is an enduring issue in food safety and public health. The globalization of manufacturing and supply chains has worsened the situation. A containable food-related disease, if not sensed early, could quickly escalate into a global calamity. To control the spread of these infectious diseases, especially when their route of transmission is unknown, timely detection of the causal virus for determining its origin and accurate diagnosis is as crucial as, if not more than medical interventions. The challenges for the developers of such detection methods are sensitivity, specificity, and beyond. Scope and approach Raman spectroscopy provides characteristic information of molecular vibrations for analytes ranging from small molecules to biological compounds and cells. The sensitivity of Raman spectroscopy has improved over time. Confocal Raman technique, surface-enhanced Raman scattering, and coherent anti-Stokes Raman scattering are now available for use in food inspection. We conduct a systematic and comprehensive survey of recent studies on Raman spectroscopic techniques aimed at detecting viruses. Reports relating to the implementation of food safety, either intended or incidental, are selected and reviewed. The basic concepts of the involved Raman techniques are described and compared. And their applications to specific viruses and diseases are summarized. Key findings and conclusion The detection of viruses by Raman scattering takes either the nanotag or label-free approach. The competitive advantages of Raman spectroscopy allow rapid acquisition of spectral data, with minimal sample preparation and the applicability to point-of-care testing. Some of the reviewed methods are highly sophisticated. However, neither approach is free from limits. The emerging challenges and perspectives of the further development of Raman spectroscopic techniques in food-adherent pathogen testing are discussed and concluded.

Multiferroic toxicity-free perovskite BiFeO3-based materials reveal potentials in photoelectrical conversion, energy harvesting and photodetector. This work highlights polarization-enhanced photovoltaic (PV) effects and mechanisms in B-site Ni-doped (Bi0.93Gd0.07)FeO3 ceramics with indium tin oxide (top electrode) and Au (bottom electrode) thin films. The enhanced PV effects are attributed to the reduced bandgap, polarization-modulated Schottky barrier, increased O 2p-Fe 3d orbital hybridization, and nucleation of domain walls to improve transportation of charge carriers. Improved PV parameters of responsivity (R) ∼1.71 × 10−2 A/W, detectivity (D*) ∼9.56 × 1011 Hz1/2/W (Jones), open-circuit voltage (Voc) ∼0.65 V, and short-circuit current density (Jsc) ∼80 μA/cm2 were obtained from the ITO/ (Bi0.93Gd0.07)(Fe0.97Ni0.03)O3/ Au heterostructure after 2 kV/cm poling under 405 nm irradiation at 103 W/m2 intensity. This work demonstrates that Ni-doped (Bi0.93Gd0.07)FeO3 ceramics can be potential candidates for self-powered UV photodetector.

Monolayer doping is a possible method for achieving complex-geometry structures with different semiconductors. Understanding the dopant diffusion behavior of monolayer doping, especially under different heating sources, is essential for further improvement. We examine and compare the doping profile and dopant activation with two different heating sources (rapid thermal annealing and microwave annealing), especially focused on SiO2/Si interface. These heating sources are used for junction diode fabrication, to realize current switching behavior. Direct observations of monolayer doping profiles, especially inside the capping oxide, are discussed to provide quantitative information for dopant concentration. This can provide significant information for better tuning of surface chemistries and process protocols applied in monolayer doping methodologies.

A framework is presented to identify an optimal accelerated test region and accelerated test conditions for the accelerated test of logic circuits for time-dependent dielectric breakdown (TDDB). Both gate-oxide breakdown and middle-of-line (MOL) TDDB are investigated. Separate test regions are identified for each wearout mechanism. Two digital circuits, an 8-bit fast Fourier transform (FFT) circuit and a Leon3 microprocessor are used to demonstrate the capability of the framework. The lifetimes of standard cells are combined to compute the circuit lifetime, by combining the Weibull distributions that characterize the lifetime distribution of each of the standard cells. The errors in estimating wearout parameters consist of two parts: the error in estimating the wearout parameters at accelerated test conditions and the forecasting accuracy at use conditions. By estimating the errors in wearout parameters at accelerated test conditions, the optimal accelerated test region is found by determining the test conditions producing a minimal error. Test conditions are selected by minimizing the error at use conditions. Given a forecasting error target, the required sample size at each test condition is found. This work also considers the impact of variation in circuit size, type, and process parameters on the selection of optimal test conditions.

Some studies showed patients with chronic urticaria have a higher rate of peptic ulcer disease (PUD). Whether PUD is a risk factor for chronic urticaria is unclear. The objective of this study was to evaluate the incidence of and risk factors for chronic urticaria in patients with PUD using the Taiwan National Health Insurance Research Database. We conducted a retrospective nationwide cohort study of the period 2000–2012 and involving 11,901 patients with PUD who underwent Helicobacter pylori (HP) therapy (PUD + HP group) and an equal number of matched patients without HP infection (PUD − HP group). Furthermore, we enrolled 23,802 patients without PUD for comparison (non-PUD group). The Cox proportional hazards regression model was used to analyze chronic urticaria risk after adjusting for potential confounding factors. The mean ages of the three groups were around 50 years. Approximately 42.6% were female. Chronic urticaria incidences in the PUD + HP and PUD − HP groups were both significantly higher than that in the non-PUD group. The hazard ratios of chronic urticaria in the PUD + HP group and the PUD − HP group were 1.34 (95% confidence interval 1.09–1.64) and 1.45 (95% confidence interval 1.19–1.79), respectively. The risk difference became significant 2 years after patients with PUD had the HP infection tests and persisted till the end of follow-up. The risk increase was significant in patients with PUD who were female or aged 40–64 years. There was no difference in the risk comparison between PUD + HP and PUD − HP groups. Peptic ulcer disease, independent of HP infection, is associated with an increased chronic urticaria risk. Patients with PUD who were female or aged 40–64 years are more likely to have chronic urticaria.

The development of infrared photodetectors is mainly limited by the choice of available materials and the intricate crystal growth process. Moreover, thermally activated carriers in traditional III-V and II-VI semiconductors enforce low operating temperatures in the infrared photodetectors. Here we demonstrate infrared photodetection enabled by interlayer excitons (ILEs) generated between tungsten and hafnium disulfide, WS2/HfS2. The photodetector operates at room temperature and shows an even higher performance at higher temperatures owing to the large exciton binding energy and phonon-assisted optical transition. The unique band alignment in the WS2/HfS2 heterostructure allows interlayer bandgap tuning from the mid- to long-wave infrared spectrum. We postulate that the sizeable charge delocalization and ILE accumulation at the interface result in a greatly enhanced oscillator strength of the ILEs and a high responsivity of the photodetector. The sensitivity of ILEs to the thickness of two-dimensional materials and the external field provides an excellent platform to realize robust tunable room temperature infrared photodetectors. Agency for Science, Technology and Research (A*STAR) Ministry of Education (MOE) National Research Foundation (NRF) Accepted version The work is supported by the Agency for Science, Technology and Research (A*STAR) under the 2D Materials Pharos Program (grant no. 152 700014 and grant no. 152 700017). Q.J.W. acknowledges funding from the National Research Foundation Competitive Research Program (NRF-CRP18-2017-02 and NRF–CRP19–2017–01). K.S.T. acknowledges support from the Center for Nanostructured Graphene (CNG) under the Danish National Research Foundation (project DNRF103) and from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant no. 773122, LIMA). G.L. acknowledges the supported under the grant MOE2017-T2-1-114. H.L. acknowledges the support by the Ministry of Science and Technology (MOST) in Taiwan under grant no. MOST 109-2112-M-001-014-MY3. J.T. and S. Lukman thank C. W. Lee, A. Ngo and M. Zhao for their valuable inputs and K Hippalgaonkar for sharing tools in the device fabrication.

Random forest has been widely used to classifying objects recently because of its efficiency and accuracy. On the other hand, nonvolatile memory has been regarded as a promising candidate to be a part of a hybrid memory architecture. For achieving the higher accuracy, random forest tends to construct lots of decision trees, and then conducts some post-pruning methods to fell low contribution trees for increasing the model accuracy and space utilization. However, the cost of writing operations is always very high on non-volatile memory. Therefore, writing the to-be-pruned trees into non-volatile memory will significantly waste both energy and time. This work proposed a framework to ease such hurt of training a random forest model. The main spirit of this work is to evaluate the importance of trees before constructing it, and then adopts different writing modes to write the trees to the non-volatile memory space. The experimental results show the proposed framework can significantly mitigate the waste of energy with high accuracy.

(1) Purpose: Undesirable health care outcomes could conceivably increase as a result of the entry of new, less experienced health care personnel into patient care during the month of July (the July effect) or as a result of the less balanced allocation of health care resources on weekends (the weekend effect). Whether these two effects were present in Taiwan’s National Health Insurance (NHI) system was investigated. (2) Methods: The current study data were acquired from the NHI Research Database. The research sample comprised ≥18-year-old patients diagnosed as having a stroke for the first time from 1 January 2006 to 30 September 2012. The mortality rate within 30 days after hospitalization and readmission rate within 14 days after hospital discharge were used as health care quality indicators, whereas health care utilization indicators were the total length and cost of initial hospitalization. (3) Results: The results revealed no sample-wide July effect with regard to the four indicators among patients with stroke. However, an unexpected July effect was present among in-patients in regional and public hospitals, in which the total lengths and costs of initial hospitalization for non-July admissions were higher than those for July admissions. Furthermore, the total hospitalization length for weekend admissions was 1.06–1.07 times higher than that for non-weekend admissions; the total hospitalization length for weekend admissions was also higher than that for weekday admissions during non-July months. Thus, weekend admission did not affect the health care quality of patients with stroke but extended their total hospitalization length. (4) Conclusions: Consistent with the NHI’s general effectiveness in ensuring fair, universally accessible, and high-quality health care services in Taiwan, the health care quality of patients examined in this study did not vary significantly overall between July and non-July months. However, a longer hospitalization length was observed for weekend admissions, possibly due to limitations in personnel and resource allocations during weekends. These results highlight the health care efficiency of hospitals during weekends as an area for further improvement.

Horizontally stacked pure-Ge-nanosheet gate-all-around field-effect transistors (GAA FETs) were developed in this study. Large lattice mismatch Ge/Si multilayers were intentionally grown as the starting material rather than Ge/GeSi multilayers to acquire the benefits of the considerable difference in material properties of Ge and Si for realising selective etching. Flat Ge/Si multilayers were grown at a low temperature to preclude island growth. The shape of Ge nanosheets was almost retained after etching owing to the excellent selectivity. Additionally, dislocations were observed in suspended Ge nanosheets because of the absence of a Ge/Si interface and the disappearance of the dislocation-line tension force owing to the elongation of misfit dislocation at the interface. Forming gas annealing of the suspended Ge nanosheets resulted in a significant increase in the glide force compared to the dislocation-line tension force; the dislocations were easily removed because of this condition and the small size of the nanosheets. Based on this structure, a new mechanism of dislocation removal from suspended Ge nanosheet structures by annealing was described, which resulted in the structures exhibiting excellent gate control and electrical properties.

BackgroundFew investigations have evaluated the influences on peripheral arterial disease (PAD) risk of statin treatment in hemodialysis (HD) subjects with hyperlipidemia (HL).MethodsFrom the National Health Insurance Research Dataset, we identified 3658 HD patients with statin therapy for HL as the statin cohort, and then selected, by 1:1 propensity score matching, 3658 HD patients with HL but without statin use as the nonstatin cohort in 2000–07. The cohorts were followed through until the end of 2011. We used Cox proportional hazards regression analysis to assess the hazard ratio (HR) of PAD development.ResultsThe average follow-up period was 4.18 years; the incident PAD risk was 1.35-fold greater in statin users than in nonusers (16.87 versus 12.46/1000 person-years), with an adjusted HR (aHR) of 1.34 for PAD [95% confidence interval (CI) 1.12–1.62]. The PAD risk increases were significant for patients receiving fluvastatin (aHR 1.88; 95% CI 1.12–3.14) and atorvastatin (aHR 1.60; 95% CI 1.24–2.08). The risk increased with higher annual average statin dosage (P for trend ConclusionsHD patients with HL on statin medication were at increased PAD risk, which increased with cumulative statin dosage. Thorough considerations are needed before prescribing statins to HD patients.

E-retailers and home delivery companies have struggled with logistics congestion. Hence, this study aims to develop a pricing model for alleviation of congestion in e-commerce logistics. The analytical results reveal that a congestion surcharge mechanism influences the behavior of a home delivery company and e-commerce consumers in terms of early delivery and delayed delivery, cargo volume, and consumers’ favorite delivery time. Moreover, e-commerce logistics fees in practice are established by negotiation between e-retailers and home delivery companies. The proposed model can be treated as a reference for e-retailers and home delivery companies to determine appropriate logistics fees. This pioneering study develops a congestion pricing model to deal with congestion problems in e-commerce logistics, which can effectively reflect the impact of various congestion fees on the behavior of e-commerce stakeholders.

BACKGROUND Heterotopic ossification (HO) is a well-documented complication of cervical disc arthroplasty (CDA), although it rarely causes adverse clinical effects. Despite high-grade HO possibly limiting segmental mobility, it is reportedly seldom associated with symptoms. OBSERVATIONS The authors report a case of a 46-year-old male patient who underwent hybrid CDA and anterior cervical discectomy and fusion for 3-level cervical disc herniation that caused myeloradiculopathy. The surgery was successful; the patient experienced nearly complete recovery postoperatively. The follow-up images, including computed tomography and magnetic resonance imaging scans, showed satisfactory decompression at the indexed levels without residual osteophytes or ossification of the posterior longitudinal ligament. However, 10 years later, the patient presented with symptomatic compressive myelopathy caused by severe HO that prompted a secondary surgery. LESSONS Although it is generally reported in the literature that HO is clinically innocuous, in this patient, it gradually and progressively developed and caused myelopathy, requiring a secondary surgery. Symptomatic HO can be expected over time, and patients with a high risk of HO deserve long-term follow-up after CDA. Further investigations are warranted to corroborate these risk factors, including multilevel calcified disc herniation, severe spondylosis, and suboptimal placement of the device during primary CDA surgery.

The possibilities of wireless sensor network (WSN) have been shown in many studies in recent years. A critical research topic among them is certainly how to reduce the power consumption of a WSN to prolong its lifetime. In addition to extending the lifetime of static wireless sensors of a WSN, some recent studies have attempted to add mobile wireless sensors to increase the coverage rate of a WSN. For this reason, this study is aimed to find the optimal deployment of a hybrid mobile and static WSN. In this study, an effective search algorithm is presented to maximize the coverage rate of all the targets while minimizing the energy consumption of the static and mobile sensors. The proposed algorithm is an improved version of search economics for single-objective real-parameter optimization problem (SE-SOP), called self-cutting search economics (SCSE). A new division method is also presented for the SE-based algorithm to divide the solution space into a certain number of subspaces and assign suitable computation resources to these subspaces according to their potentialities. The number of subspaces will be decreased so that particular subspaces are apt to acquire more investments (i.e., searches) due to the combining region method, which is helpful to speedup the convergence process. Simulation results show that the proposed algorithm can find out a better plan of network management than all the other search algorithms compared in this study for the WSN containing static and mobile sensors with adjustable sensing range functionality.

Plasmonic waveguides can offer a promising solution beyond the optical diffraction limit. However, the cost of shrinking mode sizes reflects in metallic ohmic losses that lead to a short propagation distance of light, hindering the practical applications of plasmonic waveguides. Herein, we tackled the practicality of a novel CMOS-compatible all-dielectric waveguide structure that exploits electromagnetic boundary conditions of both the continuous normal component of the electric displacement field and the tangential component of the electric field at a high-index-contrast interface, which allows the attainment of mode areas comparable with those of plasmonic waveguides and theoretical lossless. The proposed waveguide comprises two oppositely contacted nanoridges with semicircular tops embedded in a conventional slot waveguide. By stepping on the strong electric field in the low-index slot region of the slot waveguides, the nanoridges squeeze the mode areas further with a guiding mechanism identical to that of a surrounding slot waveguide. Through the design of the geometry parameters, the calculated mode area of the reported structure achieved an unprecedented order of 4.21 × 10−5 A0, where A0 is the diffraction-limited area. The mode area dependence on fabrication imperfections and spectral response showed the robustness and broadband operation. Moreover, on the basis of extremely tight mode confinements, the present waveguide even outperformed the hybrid plasmonic waveguides in lower crosstalk. The proposed idea makes the realization of practically feasible nanoscale photonic integrated circuits without any obstructions by the limited propagation distance of light for plasmonic waveguides, thereby expanding its applications in various nanophotonic and optoelectronics devices requiring strong light–matter interaction within nanoscale regions.

High-strength low-alloy steels (HSLAs) are widely used in the structural body components of many domestic motor vehicles owing to their better mechanical properties and greater resistance. The real production process of HSLA steelmaking can be regarded as a model that builds on the relationship between process parameters and product quality attributes. A surrogate modeling method is used, and the resulting production process model can be applied to predict the optimal manufacturing process parameters. We used different methods in this paper to build such a surrogate model, including linear regression, random forests, support vector regression, multilayer perception, and a simplified VGG model. We then applied three bio-inspired search algorithms, namely particle swarm optimization, the artificial bee colony algorithm, and the firefly algorithm, to search for the optimal controllable manufacturing process parameters. Through experiments on 9,000 test samples used for building the surrogate model, and 299 test samples for making the optimal process parameter selection, we found that the combination of a simplified VGG model and the firefly algorithm was the most successful at reaching a success rate of 100%—in other words, when the product quality attributes of all test samples satisfy the mechanical requirements of the end products.

Objective Fractures are a common reason for hospital admissions. However, regional and short-term studies show a varying incidence rate (IR) of fractures, and most of the surveys were conducted from only a few medical centers. Therefore, this study aims to investigate the epidemiological data of fracture hospitalizations of middle-aged and elderly persons in Taiwan between 2000 and 2015. Materials and methods Data from fractures occurring between 2000 and 2015 were obtained from the National Health Insurance Research Database in Taiwan for this study. The IR of fracture admission in patients, aged 40 years or older at the time of admission and first-time diagnosed with a fracture following admission, was calculated. Results We found that the IR of the fracture hospitalizations declined considerably from 95.70 per 10,000 person-years in 2000, to 68.48 per 10,000 person-years in 2015. The three most common fracture hospitalizations accounting for more than 50% of all fractures were fractures of the femur/hip, radius or ulna, and vertebral column. The IR of fracture hospitalization increased with age and was found to be higher in women than in men. The most common cause of fracture hospitalization for men and patients under 65 years of age was traffic accidents. In contrast, falls were the most frequent causes of fracture hospitalization for women and patients over 65 years of age. Conclusion The present study furnishes an updated picture of the incidence of fracture hospitalization over a 16-year period among middle-aged and elderly persons in Taiwan.

Background: The field of physician health is gaining increasing attention; however, most research and interventions have concentrated on factors such as job stress, mental health, and substance abuse. The risks of major cancers in physicians remain unclear. We used a propensity score-matched analysis to investigate the risk of cancer in physicians relative to the general population who had no healthcare-related professional background. Methods: Data were obtained from the National Health Insurance system in Taiwan. The physician cohort contained 29,713 physicians, and each physician was propensity score-matched with a person from the general population. Results: The physicians demonstrated a 0.90-fold lower risk of all-cancers (95% confidence interval [CI] = 0.83 – 0.96) when compared with the general population. Female physicians had a higher risk of cancer than male physicians (adjusted hazard ratio [HR] = 1.59; 95% CI = 1.28 – 1.96). Physicians had higher risks of prostate (HR = 1.26; 95% CI = 1.00 – 1.59) and thyroid cancers (HR = 3.16; 95% CI = 1.69 – 5.90) when compared with the general population. Conclusion: Physicians have lower rates of overall cancer risk than the general population. Female physicians have higher cancer risks than male physicians. Male physicians have higher risks of thyroid and prostate cancer relative to the general population.

We measured the overshoot effect in a linac and reduced it using block correction, reverse-sequence correction, and index correction. A StarTrack detector was used on a Varian iX. Five segments, 1 × 10 cm2 in area, were designed; the centers were at −4, −2, 0, 2, and 4 cm on the x axis for measuring the overshoot effect on a 10 × 10 cm2 collimator setting. Block correction was applied to two segments. The first was on the new first segment at −6 cm, and the other was on the new last segment at 6 cm. Both two new segments were obtained from the 10 × 10 cm2 collimator setting. The order of segments was reversed in reverse-sequence correction. Reverse-sequence correction averages the dose at every segment after two irradiations. When we used MLC Shaper, index correction reduced the first segment’s index (cumulative radiation occupation) by 60% and increased the last segment’s radiation occupation by 60% in a new MLC.log file. As for relative dose, the first segment had an overdose of 52.4% and the last segment had an underdose of 48.6%, when irradiated at 1 MU at 600 MU/min. The relative doses at the first segment, irradiated at 1 MU, after block correction, reverse-sequence correction, and index correction were applied decreased from 152.5% to 95.1%, 104.8%, and 100.1%, respectively. The relative doses at the last segment, irradiated at 1 MU, after block correction, reverse-sequence correction, and index correction were applied increased from 48.6% to 97.3%, 91.1%, and 95.9%, respectively. The overshoot effect depended on the speed of irradiation. High irradiation speeds resulted in notable overdosing and underdosing at the first and last segments, respectively. The three corrections mitigated the overshoot effect on dose. To save time and effort, the MLC.log file should be edited with a program in the future.

There was a reduction in the hospitalization rate for major osteoporotic fractures. As per our analysis, hospitalization for site-specific fractures showed a declining trend for hip and vertebral fractures for both the sexes. However, an increasing trend was noted in women regarding hospitalization for forearm fracture. Major osteoporotic fractures (MOFs) constitute a large proportion of the total expenditures for public healthcare. Knowing the secular trends of MOF will allow for more efficient use of healthcare resources, but such data are insufficient for the current population of Taiwan. Therefore, we investigated the epidemiological data of MOF hospitalization from adults 50 years of age or older in Taiwan during the period 2000–2015. The data analyzed were acquired from the Taiwan National Health Insurance Research Database (NHIRD) entries between 2000 and 2015. All study subjects were 50 years of age or older at the time of admission and diagnosed as having MOF. A general decline was observed in the incidence rate (IR) of MOF hospitalization for the whole population, from 74.52 per 10,000 person-years (PYs) in 2000 to 55.19 in 2015. Females aged ≥65 years had the highest rates of hospitalization for MOF among the subgroups analyzed. Apart from the wrist fracture hospitalization rates in both sexes, which remained steady, all other site-specific fracture hospitalization rates exponentially increased with age. Among men, the IRs of all MOF hospitalization were steady, except for a slight decrease in hip and vertebral fracture hospitalizations. In women, hip and vertebral fracture hospitalization rates gradually decreased, humerus and wrist fracture hospitalization remained steady, and forearm fracture hospitalization increased. Hospitalization rates of MOF decreased. The trend of site-specific fracture hospitalization analysis showed that hip and vertebral fractures decreased for both sexes. However, an increasing trend in forearm fracture hospitalization was noticed among females.

TrackNet, a deep learning network, was proposed to track high-speed and tiny objects such as tennis balls and shuttlecocks from videos. To conquer low image quality issues such as blur, afterimage, and short-term occlusion, some number of consecutive images are input together to detect an flying object. In this work, TrackNetV2 is proposed to improve the performance of TrackNet from various aspects, especially processing speed, prediction accuracy, and GPU memory usage. First of all, the processing speed is improved from 2.6 FPS to 31.8 FPS. The performance boost is achieved by reducing the input image size and re-engineering the network from a Multiple-In Single-Out (MISO) design to a Multiple-In Multiple-Out (MIMO) design. Then, to improve the prediction accuracy, a comprehensive dataset from diverse badminton match videos is collected and labeled for training and testing. The dataset consists of 55563 frames from 18 badminton match videos. In addition, the network mechanisms are composed of not only VGG16 and upsampling layers but also U-net. Last, to reduce GPU memory usage, the data structure of the heatmap layer is remodeled from a pixel-wise one-hot encoding 3D array to a real-valued 2D array. To reflect the change of the heatmap representation, the loss function is redesigned from a RMSE-based function to a weighted cross-entropy based function. An overall validation shows that the accuracy, precision and recall of TrackNetV2 respectively reach 96.3%, 97.0% and 98.7% in the training phase and 85.2%, 97.2% and 85.4% in a test on a brand new match. The processing speed of the 3-in and 3-out version TrackNetV2 can reach 31.84 FPS. The dataset and source code of this work are available at https://nol.cs.nctu.edu.tw:234/open-source/TrackNetv2/.

The complexity of building, deploying, and managing cross-organizational enterprise computing services with self-service, security, and quality assurances has been increasing exponentially in the era of hybrid multiclouds. AI-accelerated material discovery capabilities, for example, are desirable for enterprise application users to consume through business API services with assurance of satisfactory nonfunctional properties, e.g., enterprise-compliant self-service management of sharable sensitive data and machine learning capabilities at Internet scale. This paper presents a composable microservices based approach to creating and continuously improving enterprise computing services. Moreover, it elaborates on several key architecture design decisions for Navarch, a composable enterprise microservices fabric that facilitates consuming, managing, and composing enterprise API services. Under service management model of individual administration, every Navarch microservice is a managed composable API service that can be provided by an internal organization, an enterprise partner, or a public service provider. This paper also illustrates a Navarch-enabled systematic and efficient approach to transforming an AI-accelerated material discovery tool into secure, scalable, and composable enterprise microservices. Performance of the microservices can be continuously improved by exploiting advanced heterogeneous microservice hosting infrastructures. Factual comparative performance analyses are provided before the paper concludes with future work.

A new system-level RF desense characterization technique is investigated to assess platform noise in mobile computer systems. The technique uses a highly tuned antenna probe that provides 8-dB higher noise detection sensitivity in Wi-Fi band of interest than conventional near-field probe technique. This characterization technique is validated with mobile PC system and demonstrates accurate spatial RFI noise distribution. In addition, the low-RFI regions are identified and show improved Wi-Fi throughput performance. The presented desense characterization technique can be used to improve radio system performance and mitigate potential RFI risk in mobile computer system applications.

Background Conventional transbronchial needle aspiration (TBNA) is advantageous for the one-step diagnosis and staging of lung adenocarcinoma under topical anesthesia and conscious sedation. We examined its efficacy for identifying EGFR mutations. Methods Forty-seven patients with proven or suspected lung adenocarcinoma indicated for hilar-mediastinal lymph node (LN) staging between June 2011 and December 2017 were enrolled. The cellblock was prepared using the plasma-thrombin method. TaqMan PCR was used to detect mutations. Considering cost effectiveness, only the sample with the highest tumor cell fraction in the same patient was chosen for analysis. Results TBNA provided positive results of malignancy in 27 patients. Seventeen patients (63.0%) had cellblocks eligible for mutation testing. Bronchial biopsy (n = 6), neck LN fine needle aspiration (n = 1), and brushing (n = 1), provided higher tumor cell fractions for analysis in eight patients. TBNA was the exclusive method used in nine patients (19.1%). For patients with an inadequate TBNA cellblock, bronchial biopsy (n = 5), neck LN fine needle aspiration (n = 3), computed tomography-guided transthoracic needle biopsy (n = 1), and brushing (n = 1) were used for analysis. Modification to specimen processing to prevent exhaustion by cytology after June 2016 improved the adequacy of cellblock samples (9/10, 90% vs. 8/17, 47.1%; P = 0.042). Conclusions These findings suggest the promising role of conventional TBNA and highlight the challenges of doing more with less in an era of precision medicine.

Objective: Recent studies have reported that reduced excretion of urinary uromodulin is associated with renal tubular function and risks of progressive kidney disease. Gouty nephropathy is usually seen in patients with gout. Patients with chronic gouty nephropathy are characterized by the deposition of monosodium urate crystals primarily involving the collecting ducts in the medulla. We postulated that this correlation may be specific to gout and may serve as a useful biomarker for chronic kidney disease (CKD). Materials and Methods: A total of 114 Taiwanese patients diagnosed with gout (n = 72), CKD (n = 26), or healthy volunteers (n = 16) were prospectively enrolled for this study from the Rheumatology and Nephrology Outpatient Clinics of our institution. We obtained urine and blood samples on patient visits to the outpatient clinics. Demographic data were obtained from medical records. Results: In patients with gout, the spot urinary uromodulin/creatinine ratio (uUMCR; mg/g) in patients with CKD was significantly lower than that in those without CKD (CKD group: 2.2; non-CKD group: 5.6, p = 0.005). Multivariate analysis revealed that patients with CKD and gout had a lower uUMCR than those with gout alone (p = 0.028). A significant association was not observed in our non-gout cohort. Conclusion: The association of decreased uUMCR with CKD status was identified only in patients with gout in the present study. We believe that uUMCR might serve as an indicator of differential CKD in patients with gout.

Background Patients with end-stage renal disease (ESRD) who received parathyroidectomy (PTX) had persistently reduced levels of parathyroid hormone. The study investigated the risk of acute coronary syndrome (ACS) in patients with ESRD who underwent PTX using a nationwide health insurance claims database. Methods Of all ESRD patients, we selected 1047 individuals who had undergone PTX between 2000 and 2008 as the PTX group and 4188 patients who did not undergo PTX (non-PTX group) matched by propensity score. Multivariable Cox proportional hazards regression analysis was conducted for assessing the excess ACS risk for the PTX group compared to the non-PTX group. Results The mean follow-up periods were 4.63 and 4.04 years for the PTX and non- PTX groups, respectively. A significant reduction in the risk of ACS (adjusted hazard ratio = 0.74, 95% confidence interval = 0.57-0.96) was observed for the ESRD patients after PTX. Conclusions PTX is associated with reduced risk of ACS in patients with ESRD.

Facing the global trend of Industry 4.0, many countries around the world proposed its industry policy of smart factory, in which, cyber-physical production system (CPPS) plays the critical role. A CPPS has the characteristics of distributed control, real time communication, computation and modularized. In Taiwan, most of the manufacturing industries (e.g., metal manufacturing, machinery and equipment manufacturing etc.) still heavily rely on the excel-based production planning and scheduling approach although ERP systems are widely used today. Furthermore, the advanced planning and scheduling (APS) systems have been developed over 15 years in Taiwan, however, very few of small and medium-sized manufacturing enterprises (SMEs) have implemented APS system due to the high implementation cost for satisfying the specific production characteristics and planning constraints. Therefore, this research will consider the characteristics of manufacturing process, automation and digitalization of metal manufacturing, machinery and equipment manufacturing industry and develop a cloud-based advanced planning and scheduling (CAPS) system, the planning engine is based on an intelligent dynamic planning & scheduling (IDPS) system, which can dynamically and effectively generate production and operations schedule, and provide the timely visualized analysis data for production planners in SMEs. Applications of CAPS in a metal manufacturing company has been demonstrated in this paper.

The International Commission on Radiological Protection (ICRP) recommended that the equivalent dose limit for the lens of the eye be decreased from 150 mSv/y ( ICRP, 2007 ) to 20 mSv/y averaged over five years ( ICRP, 2011 ). How to accurately measure the eye-lens dose has, therefore, been an issue of interest recently. Interventional radiologists are at a higher risk of radiation-induced eye injury, such as cataracts, than all other occupational radiation workers. The main objective of this study is to investigate the relationship between the doses to the eye lenses of interventional radiologists measured by different commercial eye-lens dosimeters. This study measured a reference eye-lens dose, which involved placing thermoluminescent dosimeter (TLD) chips at the surface of the eye of the Rando Phantom, and the TLD chips were covered by a 3-mm-thick tissue-equivalent bolus. Commercial eye-lens dosimeters, such as a headband dosimeter and standard personnel dose badges, were placed at the positions recommended by the manufacturers. The results show that the personnel dose badge is not an appropriate dosimeter for evaluating eye-lens dose. Dose deviations for different dosimeters are discussed and presented in this study.

Aims Gestational diabetes mellitus (GDM) has always been a concerning issue for pregnant women. In recent studies, GDM was found to be related to epigenetic modifications, which would alter gene expressions, thus affecting the patients’ and their offspring’s health, leading to a higher probability of developing metabolic syndromes and diabetes later in life. Methods In this study, we collected both maternal and cord blood samples from 16 pregnant women and their newborns, including eight exposed to GDM. GDM was diagnosed via a 75 g oral glucose tolerance test (OGTT) at 24–28 weeks of pregnancy. DNA methylation was measured at 841,573 CpG sites via the Infinium HumanMethylationEPIC BeadChip. An Ingenuity Pathway Analysis was conducted afterwards to identify genes and pathways epigenetically affected by GDM. Results We identified the top 200 loci and their corresponding genes in the maternal blood group (n = 151) and cord blood group (n = 167), both of which were methylated differently in the GDM and unexposed group. Metabolic disease-related pathways and molecules, such as interleukin-6 and interleukin-10 were identified in both groups. These results suggested that GDM has epigenetic effects on both mother and their offspring, which might result in future metabolic syndromes or diabetes. Conclusions The high-throughput platform enabled us to analyze methylation sites throughout the genome and identify the most promising genes and pathways associated with GDM.

Lactacystin has been used to establish rodent models of Parkinson disease (PD), with cerebral α-synuclein inclusions. This study evaluated the uptake of [18F]9-fluoropropyl-(+)-dihydrotetrabenazine ([18F]FP-(+)-DTBZ), a vesicular monoamine transporter type 2 (VMAT2)-targeting radiotracer, through positron emission tomography (PET) in lactacystin-treated rat brains. Adult male Sprague–Dawley rats were randomly treated with a single intracranial dose of lactacystin (2 or 5 μg) or saline (served as the sham control) into the left medial forebrain bundle. A 30-min static [18F]FP-(+)-DTBZ brain PET scan was performed following an intravenous [18F]FP-(+)-DTBZ dose (approximately 22 MBq) in each animal at 2 and 3 weeks after lactacystin treatment. Upon completing the last PET scans, the animals were killed, and their brains were dissected for ex vivo autoradiography (ARG) and immunohistochemical (IHC) staining of tyrosine hydroxylase (TH) as well as VMAT2. Both the 2- and 5-μg lactacystin-treated groups exhibited significantly decreased specific [18F]FP-(+)-DTBZ uptake in the ipsilateral striata (I-ST) at 2 weeks (1.51 and 1.16, respectively) and 3 weeks (1.36 and 1.00, respectively) after lactacystin treatment, compared with the uptake in the corresponding contralateral striata (C-ST) (3.48 and 3.08 for the 2- and 5-μg lactacystin-treated groups, respectively, at 2 weeks; 3.36 and 3.11 for the 2- and 5-μg lactacystin-treated groups, respectively, at 3 weeks) and the sham controls (3.34–3.53). Lactacystin-induced decline in I-ST [18F]FP-(+)-DTBZ uptake was also demonstrated through ex vivo ARG, and the corresponding dopaminergic neuron damage was confirmed by the results of TH- and VMAT2-IHC studies. In this PD model, lactacystin-induced dopaminergic terminal damage in the ipsilateral striatum could be clearly visualized through in vivo [18F]FP-(+)-DTBZ PET imaging. This may serve as a useful approach for evaluating the effectiveness of new treatments for PD.

Computer vision based object tracking has been used to annotate and augment sports video. For sports learning and training, video replay is often used in post-match review and training review for tactical analysis and movement analysis. For automatically and systematically competition data collection and tactical analysis, a project called CoachAI has been supported by the Ministry of Science and Technology, Taiwan. The proposed project also includes research of data visualization, connected training auxiliary devices, and data warehouse. Deep learning techniques will be used to develop video-based real-time microscopic competition data collection based on broadcast competition video. Machine learning techniques will be used to develop a tactical analysis. To reveal data in more understandable forms and to help in pre-match training, AR/VR techniques will be used to visualize data, tactics, and so on. In addition, training auxiliary devices including smart badminton rackets and connected serving machines will be developed based on the IoT technology to further utilize competition data and tactical data and boost training efficiency. Especially, the connected serving machines will be developed to perform specified tactics and to interact with players in their training.

This paper presents a single antenna supporting all sub-6 GHz bands of 2G/3G/4G/5G, Bluetooth, and WiFi radio in a very thin and compact form-factor of 50 × 19.75 × 0.8 mm3 for mobile devices. The antenna consists of branched monopole antennas, T-shaped element, passive element, and one matching stub applied with tapering and meandering techniques, while the coupling among elements are optimized to minimize the size and boost performance. Impedance matching is better than -6 dB and the efficiency is between 60-90% within very wide bands up to 6 GHz. The proposed antenna design may enable further form-factor shrinking of 5G mobile devices.

A polymer lens attach process is developed on Optical Multi-Chip Modules (OMCM) with integrated optical waveguides and tested successfully in Deep Thermal Cycle (DTC) at -40°C/125°C for 1000 cycles. The lens attach is completed in a two step process: UV tack of the lens array using a high precision pick and place tool followed by the dispense of a viscous sidefill material along the lens array's periphery to mechanically secure it to the OMCM laminate. A screening protocol is developed based on manufacturing and reliability criteria for testing lens array UV tacking and sidefill adhesives. Manufacturing selection criteria include good adhesion and low infiltration in small gaps. Low infiltration under the lens array is necessary to prevent contamination of laminate optical features. For testing lens attach reliability, a list of critical process parameters are chosen including silicone and epoxy UV tacking and sidefill adhesives with a wide range of thermo-mechanical properties: Young's Modulus (E), Coefficient of Thermal expansion (CTE), glass transition temperature (Tg). Lens attach reliability performance is assessed by both visual inspection for mechanical integrity failures, and using lens to lens insertion loss optical test read-outs.

Background: Pleurodesis is often used to prevent the re-accumulation of a malignant pleural effusion (MPE). Intrapleural urokinase (IPUK) therapy facilitates lung re-expansion for patients with loculated MPE or a trapped lung that allows subsequent pleurodesis. MPE management has been traditionally regarded as a symptomatic treatment. We tried to evaluate their impact on patient survival. Methods: There were 314 consecutive patients with symptomatic MPE that underwent minocycline pleurodesis with ( n = 109) and without ( n = 205) the antecedent IPUK therapy between September 2005 and August 2015, who were recruited for the pleurodesis outcome and survival analysis. Results: The rate of successful pleurodesis was similar between the simple pleurodesis group and the IPUK therapy group followed by the pleurodesis group (69.0% versus 70.5%; p = 0.804). The patients who succeeded pleurodesis had a longer survival rate than those who failed in either the simple pleurodesis group (median, 414 versus 100 days; p < 0.001) or the IPUK therapy followed by pleurodesis group (259 versus 102 days; p < 0.001). The survival differences remained when the lung and breast cancer patients were studied separately. Conclusion: Successful pleurodesis translated into a better survival rate that promotes performing pleurodesis on lung re-expansion. The apparent shorter survival of the patients with loculated MPE or trapped lung, and those that did not respond to the IPUK therapy, lowered the probability of the survival benefit through the simple physical barrier by the fibrin formation to prevent the tumor spreading. The successfully induced inflammatory response by minocycline is supposed to prohibit the tumor invasion and metastasis. Further studies are warranted to clarify the mechanism and provide opportunities to develop novel therapeutic strategies.

Background Cardiovascular (CV) disease contributes to nearly half of the mortalities in patients with end-stage renal disease. Patients who received prehemodialysis arteriovenous access (pre-HD AVA) creation had divergent CV outcomes. Methods We conducted a population-based cohort study by recruiting incident patients receiving HD from 2001 to 2012 from the Taiwan National Health Insurance Research Database. Patients’ characteristics, comorbidities, and medicines were analyzed. The primary outcome of interest was major adverse cardiovascular events (MACEs), defined as hospitalization due to acute myocardial infarction, stroke, or congestive heart failure (CHF) occurring within the first year of HD. Secondary outcomes included MACE-related mortality and all-cause mortality in the same follow-up period. Results The patients in the pre-HD AVA group were younger, had a lower burden of underlying diseases, were more likely to use erythropoiesis-stimulating agents but less likely to use renin–angiotensin–aldosterone system blockers. The patients with pre-HD AVA creation had a marginally lower rate of MACEs but a significant 35% lower rate of CHF hospitalization than those without creation (adjusted hazard ratio (HR) 0.65, 95% confidence interval (CI) [0.48–0.88]). In addition, the pre-HD AVA group exhibited an insignificantly lower rate of MACE-related mortality but a significantly 52% lower rate of all-cause mortality than the non-pre-HD AVA group (adjusted HR 0.48, 95% CI [0.39–0.59]). Sensitivity analyses obtained consistent results. Conclusions Pre-HD AVA creation is associated with a lower rate of CHF hospitalization and overall death in the first year of dialysis.

Because data from a variety of wearout mechanisms is confounded in circuits, we apply machine learning techniques to detect the parameters of competing failure mechanisms in ring oscillators, which more closely mimic circuit behavior than test structures. This is the first known application using data analysis to distinguish competing wearout mechanisms in circuit-level data. To quickly and efficiently analyze failure data, we propose to use maximum likelihood estimation to separately determine the parameters of each underlying distribution by only observing the time-to-failure of samples. The quasi-Newton method is used to update and optimize the parameter extraction.

This paper applies a Diagonal BEKK model to investigate the risk spillovers of three major cryptocurrencies to ten leading traditional currencies and two gold prices (Spot Gold and Gold Futures). The daily data used are from 7 August 2015 to 15 June 2020. The dataset is analyzed in its entirety and is also subdivided into four distinct subsets in order to study and compare the patterns of spillover effects during economic turmoil, such as the 2018 cryptocurrency crash and the COVID-19 pandemic. The results reveal significant co-volatility spillover effects between cryptocurrency and traditional currency or gold markets, especially during the whole sample period and amid the uncertainty raised by COVID-19. The capabilities of cryptocurrency are time-varying and related to economic uncertainty or shocks. There are significant differences between normal and extreme markets with regard to the capabilities of cryptocurrency as a diversifier, a hedge or a safe haven. We find the significant co-volatility spillover effects are asymmetric in most cases especially during the COVID-19 pandemic period, which means the negative return shocks have larger impacts on co-volatility than positive return shocks of the same magnitude. Evidently, cryptocurrencies and traditional currencies or gold can be incorporated into financial portfolios for financial market participants who seek effective risk management and also for optimal dynamic hedging purposes against economic turmoil and downward movements.

Black phosphorus (BP) has been proposed as the channel material in the next generation ultrascaled CMOS devices. In order to gain insight into the current characteristics in 2-D layered materials, the current distribution of a few-layer BP Schottky barrier FET is investigated via state-of-the-art quantum device simulations. Approximately 40% of the total current was found to be concentrated in the top layer when the device was switched on, with the remaining current distributed among the other layers. In comparison, ~80% of the current concentrated below the surface in a Si device with the same structure. These features are related to the strength of the intra/interlayer interaction in few-layer BP and are unique to 2-D layered materials. Moreover, the current distribution and the device performance were different for the top- and side-contacted devices, with the side-contacted devices yielding lower resistance compared with the top-contacted devices.

Introduction Limited studies have evaluated risk of stroke associated with the use of NSAIDs in patients with end-stage kidney disease. We examined the adverse effects of selective and nonselective NSAID use on the risk of stroke in dialysis patients. Methods A case-crossover study was conducted using medical claims data from the National Health Insurance Research Database in Taiwan. We identified patients with ischemic and hemorrhagic stroke (defined as International Classification of Diseases, 9th revision, Clinical Modification codes 433, 434, and 436 for ischemic stroke and 430 and 431 for hemorrhagic stroke) from inpatient claims during the period from 2003 to 2012. Conditional logistic regression models with adjustment for potential confounders were used to determine the effects of NSAID use on stroke. Results A total of 1190 dialysis patients with stroke were identified from 2003 to 2012. The results indicate a 1.31-fold increased risk of stroke related to NSAID use during the 30 days prior to a stroke (AOR = 1.31; 95% CI: 1.03–1.66); likewise, an excessive risk of ischemic stroke was observed (AOR = 1.34; 95% CI: 1.02–1.77). When classifying NSAIDs into selective and nonselective groups, nonselective NSAID use was significantly associated with an increased risk of stroke (AOR = 1.27; 95% CI: 1.00–1.61). Discussion In summary, the results show supportive evidence that NSAID use increased the risk of stroke in dialysis patients, which suggests the importance of closely monitoring the transient effects of initial NSAID treatment to patients on dialysis.

Proteinopathy in the heart which often manifests excessive misfolded/aggregated proteins in cardiac myocytes can result in severe fibrosis and heart failure. Here we developed a mouse model, which transgenically express tetrameric DsRed, a red fluorescent protein (RFP), in an attempt to mimic the pathological mechanisms ofcardiac fibrosis. Whilst DsRed is expressed and forms aggregation in most mouse organs, certain pathological defects are specifically recapitulated in cardiac muscle cells including mitochondria damages, aggresome-like residual bodies, excessive ubiquitinated proteins, and the induction of autophagy. The proteinopathy and cellular injuries caused by DsRed aggregates may be due to impaired or overburdened ubiquitin-proteasome system and autophagy-lysosome systems. We further identified that DsRed can be ubiquitinated and associated with MuRF1, a muscle-specific E3 ligase. Concomitantly, an activation of NF-κB signaling and a strong TIMP1 induction were noted, suggesting that RFP-induced fibrosis was augmented by a skewed balance between TIMP1 and MMPs. Taken together, our study highlights the molecular consequences of uncontrolled protein aggregation leading to congestive heart failure, and provides novel insights into fibrosis formation that can be exploited for improved therapy.

Background In real-world practice settings, there is insufficient evidence on the efficacy of antiplatelet drugs, including clopidogrel, aspirin, and ticlopidine, in stroke prevention. Purpose To compare the efficacies between aspirin and clopidogrel and aspirin and ticlopidine in stroke prevention. Methods This population-based case-cohort study utilized the data obtained from a randomized sample of one million subjects in the Taiwan National Health Insurance Research Database. Patients who were hospitalized owing to the primary diagnosis of ischemic stroke from January 1, 2000 to December 31, 2010 and treated with aspirin, ticlopidine, or clopidogrel were included in the study. Propensity score matching with a 1:4 ratio was performed to compare aspirin with ticlopidine and clopidogrel. The criteria for inclusion were the use of one of the three antiplatelet drugs for more than 14 days within the first month after the stroke and then continued use of the antiplatelet drugs until the study endpoint of recurrent stroke. Results During the 3-year follow-up period, the recurrent stroke rates were 1.62% (42/2585), 1.48% (3/203), and 2.55% (8/314) in the aspirin, ticlopidine, and clopidogrel groups, respectively. Compared with the patients treated with aspirin, those treated with clopidogrel and ticlopidine showed competing risk-adjusted hazard ratios of recurrent stroke of 2.27 (1.02–5.07) and 0.62 (0.08–4.86), respectively. Conclusion Compared with the patients treated with aspirin, those treated with clopidogrel were at a higher risk of recurrent stroke. For stroke prevention, aspirin was superior to clopidogrel whereas ticlopidine was not inferior to aspirin.

Silica nanospheres with a well-controlled particle size were prepared via a nucleation-to-growth synthesis process. A facile method is proposed for improving the self-assembly behavior of silica colloidal particles in droplet coatings by the simple controlling of the drying temperature. It is shown that a periodically arranged, opal-structured, photonic crystal film with a large area of approximately 4.0 cm2 can be prepared, even when the particle size is up to 840 nm. When the band gap of the silica photonic crystals falls in the visible-light region, the crystals exhibit distinct structural colors. Moreover, the wavelength of the reflected light increases with an increasing particle size of silica. When the photonic band gap overlaps the wavelength of the laser source, the overall Raman spectrum intensity is significantly enhanced. Accordingly, the proposed nucleation-to-growth process and drop-coating method provides a cheap and simple approach for the manufacture of uniform sized silica and surface-enhanced Raman scattering substrates, respectively.

Released thyroid cancer patients treated with 131I may cause a radiation risk to their family members. However, no current regulations exist in Taiwan to limit the radiation dose among patients released post-treatment. This study assessed radiation exposure to caregivers of thyroid cancer patients treated with 131I to verify whether their doses exceed the 5-mSv limit per treatment session. The study population comprised 63 patients from four hospitals and their caregivers. Dose rates at a 1-m distance from patients were quantified before their release. Dosimetry data of caregivers were obtained using personalized thermoluminescent dosimetry badges. Upon the release of patients treated with 131I, Dose rates at a 1-m distance from the patients were 3.9–55 μSv/h. The radiation doses to caregivers ranged from 0.21 to 1.7 mSv after the patients released. When the 1-m dose rate of patients upon hospital release was 50 μSv/h, the highest possible radiation dose was assessed of 4.76 mSv for caregivers. In conclusion, as the dose rates at a 1-m distance from patients upon release were