Recognition of Glaucomatous Fundus Images Using Machine Learning Methods Based on Optic Nerve Head Topographic Features.

PURPOSE: The aim was to compare the effectiveness of clinical discriminant rules and machine learning classifiers in identifying glaucomatous fundus images based on optic disc topographic features. DESIGN: Retrospective case-control study. METHODS: The study used a total of 800 fundus images, half of which were glaucomatous cases and the other half non- glaucomatous cases obtained from an open database and clinical work. The images were randomly divided into training and testing sets with equal numbers of glaucomatous and non-glaucomatous images. An ophthalmologist framed the edge of the optic cup and disc, and the program calculated five features, including the vertical cup-to-disc ratio and the width of the optic rim in four quadrants in pixels, used to create machine learning classifiers. The discriminative ability of these classifiers was compared with clinical discriminant rules. RESULTS: The machine learning classifiers outperformed clinical discriminant rules, with the extreme gradient boosting method showing the best performance in identifying glaucomatous fundus images. Decision tree analysis revealed that the cup-to-disc ratio was the most important feature for identifying glaucoma fundus images. At the same time, the temporal width of the optic rim was the least important feature. CONCLUSIONS: Machine learning classifiers are an effective approach to detecting glaucomatous fundus images based on optic disc topographic features and integration with an automated program for framing and calculating the required parameters would make it a straightforward and effective approach.

Early Detection of Primary Open Angle, Angle Closure, and Normal Tension Glaucoma in an Asian Population Using Optical Coherence Tomography.

PRCIS: Spectral-domain optical coherence tomography (SD-OCT) facilitates early glaucoma detection in the Chinese population in Taiwan. The best parameters for primary open angle glaucoma (POAG), primary angle closure glaucoma (PACG), normal tension glaucoma (NTG), and suspected glaucoma (GS) detection are temporal inferior Bruch's membrane opening-minimum rim width (BMO-MRW), inner temporal macular ganglion cell layer (mGCL), temporal superior Circumpapillary retinal nerve fiber layer (cpRNFL), and mean global BMO-MRW, respectively. PURPOSE: We investigated the diagnostic capability of SD-OCT for different types of early glaucoma among the Chinese population in Taiwan. PARTICIPANTS AND METHODS: One eye each was assessed from 113 individuals with healthy eyes, 125 individuals with suspected glaucoma (GS), and 156 patients with early glaucoma (POAG, 87; PACG, 50; and NTG, 19). Circumpapillary (cp) RNFL thickness (global and sectoral), BMO-MRW, and macular parameters, including the macular RNFL (mRNFL), mGCL, and macular inner plexiform layer (mIPL), were assessed using SD-OCT. The areas under receiver operator characteristic curves (AUCs) were calculated to evaluate the diagnostic capacity of the parameters to differentiate between healthy and early glaucomatous eyes. RESULTS: The parameters most suitable for detecting early POAG, PACG, NTG, and GS were temporal inferior BMO-MRW (AUC, 0.847), inner temporal mGCL (AUC, 0.770), temporal superior cpRNFL (AUC, 0.861), and mean global BMO-MRW (AUC, 0.768), respectively. Among the macular parameters, the mGCL exhibited the highest diagnostic capacity. The diagnostic capacity of the mGCL was lower than that of cpRNFL and BMO-MRW for POAG and NTG but not PACG. After adjusting for confounding variables in multivariable analysis, the AUC was determined to be 0.935 for POAG and 0.787 for GS. CONCLUSION: SD-OCT facilitates the detection of early POAG, PACG, and NTG. Using a combination of cpRNFL, BMO-MRW, and macular parameters may enhance their diagnostic capacities. Further studies are necessary to validate these findings.

A U-Shaped Optical Fiber Temperature Sensor Coated with Electrospinning Polyvinyl Alcohol Nanofibers: Simulation and Experiment.

This study describes the fabrication of an electrospun, U-shaped optical fiber sensor for temperature measurements. The sensor is based on single mode fibers and was fabricated into a U-shaped optical fiber sensor through flame heating. This study applied electrospinning to coat PVA, a polymer, onto the sensor layer to reduce its sensitivity to humidity. The sensor is used to measure temperature variations ranging from 30 °C to 100 °C. The objectives of this study were to analyze the sensitivity variation of the sensor with different sensor layer thicknesses resulting from different electrospinning durations, as well as to simulate the wavelength signals generated at different electrospinning durations using COMSOL. The results revealed that the maximum wavelength sensitivity, transmission loss sensitivity, and linearity of the sensor were 25 dBm/°C, 70 pm/°C, and 0.956, respectively. Longer electrospinning durations resulted in thicker sensor layers and higher sensor sensitivity, that wavelength sensitivity of the sensor increased by 42%.

Glaucoma Detection Using Support Vector Machine Method Based on Spectralis OCT.

Spectralis optical coherence tomography (OCT) provided more detailed parameters in the peripapillary and macular areas among the OCT machines, but it is not easy to understand the enormous information (114 features) generated from Spectralis OCT in glaucoma assessment. Machine learning methodology has been well-applied in glaucoma detection in recent years and has the ability to process a large amount of information at once. Here we aimed to analyze the diagnostic capability of Spectralis OCT parameters on glaucoma detection using Support Vector Machine (SVM) classification method in our population. Our results showed that applying all OCT features with the SVM method had good capability in the detection of glaucomatous eyes (area under curve (AUC) = 0.82), as well as discriminating normal eyes from early, moderate, or severe glaucomatous eyes (AUC = 0.78, 0.89, and 0.93, respectively). Apart from using all OCT features, the minimum rim width (MRW) may be good feature groups to discriminate early glaucomatous from normal eyes (AUC = 0.78). The combination of peripapillary and macular parameters, including MRW_temporal inferior (TI), MRW_global (G), ganglion cell layer (GCL)_outer temporal (T2), GCL_inner inferior (I1), peripapillary nerve fiber layer thickness (ppNFLT)_temporal superior (TS), and GCL_inner temporal (T1), provided better results (AUC = 0.84). This study showed promise in glaucoma management in the Taiwanese population. However, further validation study is needed to test the performance of our proposed model in the real world.

Retinal artery occlusion following cosmetic injection of poly-L-lactic acid.

Ophthalmic vascular occlusion is an infrequent but devastating complication following cosmetic facial filler injection. We report a case of developing retinal artery occlusion after poly-L-lactic acid (PLLA) injection. A 49-year-old woman with multiple chronic diseases experienced sudden central visual loss and severe ocular pain in the right eye immediately after PLLA injection in the temporal region. Her best-corrected visual acuity in the right eye dropped from 20/20 to 20/200. Fundus photography showed marked optic disc edema and localized retinal whitening in the territory of the blocked vessels. Spectral-domain optical coherence tomography revealed localized hyperreflectivity of the inner retina and retinal edema. Fluorescein angiography showed delayed filling of the retinal arteries and absence of retinal perfusion in the affected areas. Despite prompt aggressive management of the condition with ocular massage, topical brimonidine eyedrops, and hyperbaric oxygen therapy, the patient suffered permanent visual loss due to optic atrophy. Among all the subcutaneous filler materials, PLLA has not been a common cause of vascular complications, especially when injected in the temporal region, as this area has not been considered dangerous in the previous literature. Practitioners should be aware of the risk of visual loss, and extra care should be given on those who originally have a higher risk for vascular complications.

Comparison of Different Machine Learning Classifiers for Glaucoma Diagnosis Based on Spectralis OCT.

Early detection is important in glaucoma management. By using optical coherence tomography (OCT), the subtle structural changes caused by glaucoma can be detected. Though OCT provided abundant parameters for comprehensive information, clinicians may be confused once the results conflict. Machine learning classifiers (MLCs) are good tools for considering numerous parameters and generating reliable diagnoses in glaucoma practice. Here we aim to compare different MLCs based on Spectralis OCT parameters, including circumpapillary retinal nerve fiber layer (cRNFL) thickness, Bruch's membrane opening-minimum rim width (BMO-MRW), Early Treatment Diabetes Retinopathy Study (ETDRS) macular thickness, and posterior pole asymmetry analysis (PPAA), in discriminating normal from glaucomatous eyes. Five MLCs were proposed, namely conditional inference trees (CIT), logistic model tree (LMT), C5.0 decision tree, random forest (RF), and extreme gradient boosting (XGBoost). Logistic regression (LGR) was used as a benchmark for comparison. RF was shown to be the best model. Ganglion cell layer measurements were the most important predictors in early glaucoma detection and cRNFL measurements were more important as the glaucoma severity increased. The global, temporal, inferior, superotemporal, and inferotemporal sites were relatively influential locations among all parameters. Clinicians should cautiously integrate the Spectralis OCT results into the entire clinical picture when diagnosing glaucoma.

Internal Residual Strain Measurements in Carbon Fiber-Reinforced Polymer Laminates Curing Process Using Embedded Tilted Fiber Bragg Grating Sensor.

Carbon fiber reinforced plastics (CFRP) have many mechanical properties that are superior to those of conventional structural materials and are becoming more and more widely used. Monitoring the curing process used to produce such composite material is important to ensure the quality of the process, especially for the characterization of residual strains after the material has been manufactured. In this study, we present a tilted fiber Bragg grating (TFBG) sensor used to monitor the curing of CFRP composite materials. The TFBG sensor was embedded into the layers of CFRP laminates to study the curing residual strain of the laminates. The experimental results showed that the curing residual stress was about -22.25 MPa, the axial residual strain was -281.351 µÎµ, and lateral residual strain of 89.91 µÎµ. The TFBG sensor was found to be sensitive to the curing residual strain of the CFRP, meaning that it has potential for use in applications involving composite curing processes. Moreover, it is indeed possible to improve the properties of composite materials via the optimization and monitoring of their curing parameters.

[Evaluating the impacts of C9 leakage on bay ecosystem services: A case study in x bay]. / C9æ³„æ¼äº‹æ• å¯¹æµ·æ¹¾ç”Ÿæ€ç³»ç»ŸæœåŠ¡çš„å½±å“è¯„ä¼°­­ä»¥x海湾为例.

The bay is the most susceptible area in the marine to human interference. It is of significance for maintaining ecological security of the bay to build an assessment framework of losses of bay ecosystem services caused by the C9 leakage event and evaluate it quantitatively. This study used market value, alternative cost, carbon tax and emergy analysis methods to construct a monetary value evaluation model for the lossses of key ecosystem services (food production, gas regulation, climate regulation, waste treatment, human health, nutrient cycling, species diversity maintenance, and recreation entertainment) caused by C9 leakage accident, and analyzed the losses of x-Bay ecosystem services. The results showed that total value of the losses of ecosystem services caused by C9 spill was 1.93×108 yuan, and the monetary value of loss per unit area was 1.19×108 yuan·km-2, which was more than 2800 times of the general marine oil spill events. Among all the components, the loss of food production services accounted for 77.1% of the total, being much higher than the impact of the general marine oil spills on human production and life. Our results could provide references to the assessment of ecosystem services loss caused by toxic substances like C9, and to the government decision-making and national territory spatial planning.

Glucose Sensor Using U-Shaped Optical Fiber Probe with Gold Nanoparticles and Glucose Oxidase.

In this study, we proposed a U-shaped optical fiber probe fabricated using a flame heating method. The probe was packaged in glass tube to reduce human factors during experimental testing of the probe as a glucose sensor. The U-shaped fiber probe was found to have high sensitivity in detecting the very small molecule. When the sensor was dipped in solutions with different refractive indexes, its wavelength or transmission loss changed. We used electrostatic self-assembly to bond gold nanoparticles and glucose oxidase (GOD) onto the sensor’s surface. The results over five cycles of the experiment showed that, as the glucose concentration increased, the refractive index of the sensor decreased and its spectrum wavelength shifted. The best wavelength sensitivity was 2.899 nm/%, and the linearity was 0.9771. The best transmission loss sensitivity was 5.101 dB/%, and the linearity was 0.9734. Therefore, the proposed U-shaped optical fiber probe with gold nanoparticles and GOD has good potential for use as a blood sugar sensor in the future.

Tilted Fiber Bragg Grating Sensor with Graphene Oxide Coating for Humidity Sensing.

In this study, we propose a tilted fiber Bragg grating (TFBG) humidity sensor fabricated using the phase mask method to produce a TFBG that was then etched with five different diameters of 20, 35, 50, 55 and 60 µm, after which piezoelectric inkjet technology was used to coat the grating with graphene oxide. According to the experimental results, the diameter of 20 µm yielded the best sensitivity. In addition, the experimental results showed that the wavelength sensitivity was -0.01 nm/%RH and the linearity was 0.996. Furthermore, the measurement results showed that when the relative humidity was increased, the refractive index of the sensor was decreased, meaning that the TFBG cladding mode spectrum wavelength was shifted. Therefore, the proposed graphene oxide film TFBG humidity sensor has good potential to be an effective relative humidity monitor.

Application of Robust, Packaged Long-Period Fiber Grating for Strain Measurement.

This paper proposes an optical fiber strain sensor based on packaged long-period fiber gratings (PLPFG) which is fabricated by the micro-electromechanical systems (MEMS) process and packaged with poly-dimethylsiloxane (PDMS) polymer materials. The optical fiber sensor packaged with PDMS improves robustness effectively. The proposed PLPFG sensors have periods of 610, 650, 660 µm and fiber diameter of 48, 60, 72 µm, respectively. The resonance dip of the PLPFG grows when a strain loaded onto the sensor. The results show that the largest strain sensitivity of the PLPFG strain sensor was -0.0652 dB/µÎµ from 0⁻1200 µÎµ and the linearity (R²) was 0.9812. Accordingly, the proposed PLPFG sensor has good potential for high-sensitivity strain sensing applications.

Sandwiched Long-Period Fiber Grating Fabricated by MEMS Process for CO2 Gas Detection.

This paper presents an optical fiber gas sensor based on sandwiched long-period fiber grating (SLPFG) that is fabricated via the microelectromechanical systems (MEMS) process and coated with amino silica adsorbent for carbon dioxide (CO2) gas sensing. The amine-modified nanoporous silica foams were coated onto the SLPFG for CO2 adsorption. To characterize the CO2 adsorption of the SLPFG sensor, a gas sensing test was conducted with a mixed gas consisting of 15% CO2 and 85% nitrogen at a flow rate of 0.2 L/min. The results showed that the spectra of the SLPFG were varied with the gas flow within 21 min. After that, the transmission spectra of the SLPFG held steady and exhibited no further change. This phenomenon was caused by the adsorption saturation of the amine-modified nanoporous silica foams which were coated onto the SLPFG. During the absorption process, the transmission was increasing by about 11.27 dB (from -23.11 to -11.84 dB), and the increasing rate of transmission was 0.4598 dB/min. Repeatable adsorption and desorption experiment results showed that the SLPFG CO2 gas sensor exhibited good repeatability and a short response time. The recovery rate for each cycle was about 85%, and the required recovery time was short. Therefore, elaborated SLPFG gas sensor could potentially be used as a gas sensor for monitoring CO2 adsorption in the context of various industrial, agricultural, and household applications.

Metabolic flux network analysis of fermentative hydrogen production: using Clostridium tyrobutyricum as an example.

This study applies metabolic flux network analysis (MFA) to evaluate the metabolic flux of fermentative hydrogen production (FHP) with the use of Clostridium tyrobutyricum fed with either glucose or lactate/acetate as substrates. The MFA results suggest that hydraulic retention time (HRT) presents significant impact on hydrogen production from glucose. At HRT between 4 and 18 h, increase of HRT increased hydrogen production but decreased lactate production, while at HRT below 4 h decrease of HRT increased hydrogen production but decreased lactate production. The flux for lactate, butyrate and acetate seemed to affect H2 production, due presumably to their impacts on the balance of NADH, ferredoxin and ATP. It is suggested that the MFA can be a useful tool to provide valuable information for optimization and design of the fermentative hydrogen production process.

A two-stage bioprocess for hydrogen and methane production from rice straw bioethanol residues.

This study evaluates a two-stage bioprocess for recovering hydrogen and methane while treating organic residues of fermentative bioethanol from rice straw. The obtained results indicate that controlling a proper volumetric loading rate, substrate-to-biomass ratio, or F/M ratio is important to maximizing biohydrogen production from rice straw bioethanol residues. Clostridium tyrobutyricum, the identified major hydrogen-producing bacteria enriched in the hydrogen bioreactor, is likely utilizing lactate and acetate for biohydrogen production. The occurrence of acetogenesis during biohydrogen fermentation may reduce the B/A ratio and lead to a lower hydrogen production. Organic residues remained in the effluent of hydrogen bioreactor can be effectively converted to methane with a rate of 2.8 mmol CH(4)/gVSS/h at VLR of 4.6 kg COD/m(3)/d. Finally, approximately 75% of COD in rice straw bioethanol residues can be removed and among that 1.3% and 66.1% of COD can be recovered in the forms of hydrogen and methane, respectively.

Fermentative biohydrogen production from lactate and acetate.

In this study, a continuous-flow stirred tank reactor (CSTR) fed with lactate and acetate was operated to enrich hydrogen-producing bacteria. By varying the influent substrate concentrations and hydraulic retention times (HRT), the volumetric loading rate (VLR) of 55.64 kg-COD/m(3)/day seemed to be optimum for this enriched culture for fermentative hydrogen production from lactate and acetate. The results of batch experiments confirmed that the enriched culture tended to fulfill the e(-) equiv requirement for cell growth at a lower VLR condition (21.77 kg-COD/m(3)/day), while it could largely distribute the e(-) equiv for hydrogen production at a higher VLR condition. However, a maximum lactate/acetate concentration allowed for enriching this culture existed, especially at a lower HRT condition in which wash-out can be an issue for this enriched culture. Finally, the results of cloning and sequencing indicated that Clostridium tyrobutyricum was considered the major hydrogen-producing bacteria in the CSTR fed with lactate and acetate.

Metabolic and energetic aspects of biohydrogen production of Clostridium tyrobutyricum: The effects of hydraulic retention time and peptone addition.

This study evaluates the microbial metabolism and energy demand in fermentative biohydrogen production using Clostridium tyrobutyricum FYa102 at different hydraulic retention times (HRT) over a period of 1-18 h. The hydrogen yield shows a positive correlation with the butyrate yield, the B/A ratio, and the Y(H2)/2(Y(HAc)+Y(HBu)) ratio, but a negative correlation with the lactate yield. A decrease in HRT, which is accompanied by an increased biomass growth, tends to decrease the B/A ratio, due presumably to a higher energy demand for microbial growth. The production of lactate at a low HRT, however, may involve an unfavorable change in e(-) equiv distribution to result in a reduced hydrogen production. Finally, the relatively high hydrogen yields observed in the bioreactor with the peptone addition may be ascribed to the utilization of peptone as an additional energy and/or amino-acid source, thus reducing the glucose demand for biomass growth during the hydrogen production process.