Deep Learning Estimation of 10-2 Visual Field Map Based on Macular Optical Coherence Tomography Angiography Measurements.

PURPOSE: To develop deep learning (DL) models estimating the central visual field (VF) from optical coherence tomography angiography (OCTA) vessel density (VD) measurements. DESIGN: Development and validation of a deep learning model. METHODS: A total of 1051 10-2 VF OCTA pairs from healthy, glaucoma suspects, and glaucoma eyes were included. DL models were trained on en face macula VD images from OCTA to estimate 10-2 mean deviation (MD), pattern standard deviation (PSD), 68 total deviation (TD) and pattern deviation (PD) values and compared with a linear regression (LR) model with the same input. Accuracy of the models was evaluated by calculating the average mean absolute error (MAE) and the R2 (squared Pearson correlation coefficients) of the estimated and actual VF values. RESULTS: DL models predicting 10-2 MD achieved R2 of 0.85 (95% confidence interval [CI], 74-0.92) for 10-2 MD and MAEs of 1.76 dB (95% CI, 1.39-2.17 dB) for MD. This was significantly better than mean linear estimates for 10-2 MD. The DL model outperformed the LR model for the estimation of pointwise TD values with an average MAE of 2.48 dB (95% CI, 1.99-3.02) and R2 of 0.69 (95% CI, 0.57-0.76) over all test points. The DL model outperformed the LR model for the estimation of all sectors. CONCLUSIONS: DL models enable the estimation of VF loss from OCTA images with high accuracy. Applying DL to the OCTA images may enhance clinical decision making. It also may improve individualized patient care and risk stratification of patients who are at risk for central VF damage.

Dehydrogenation and dehydration of formic acid over orthorhombic molybdenum carbide.

The dehydrogenation and dehydration of formic acid is investigated on the ß-Mo2C (100) catalyst surface using time independent density functional theory. The energetics of the two mechanisms are calculated, and the thermochemistry and kinetics are discussed using the transition state theory. Subsequently, microkinetic modelling of the system is conducted, considering the batch reactor model. The potential energy landscape of the reaction shows a thermodynamically favourable cleavage of H-COOH to form CO; however, the kinetics show that the dehydrogenation mechanism is faster and CO2 is continuously formed. The effect of HCOOH adsorption on the surface is also analysed, in a temperature-programmed desorption, with the conversion proceeding at under 350 K and desorption of CO2 is observed with a selectivity of about 100 %, in line with the experimental reports.

Study of Conversion of Bio-oil Model Compounds in Supercritical Water Using Density Functional Theory.

It is well known that supercritical water is a favourable medium for biomass conversion followed by its hydrodeoxygenation (HDO). Moreover, the actual kinetics and mechanism of reaction occurring in the supercritical water are not yet completely understood, either by experimental or computational approaches. Within the framework of DFT, the major challenge is non-availability of models to simulate supercritical phase. In this study, the authors manually define the descriptors of a solvation model to describe an implicit supercritical phase. In order to examine the suitability of supercritical water for thermal and hydrotreatment of bio-oil model compounds, nine different reactions involving conversion of furfural, tetrahydrofuran, xylose, phenol, guaiacol, ferulic acid, acetic acid, 2-hydroxybenzaldehyde and hydroxyacetone have been considered. Further these reactions are also studied in gas and liquid phase to compare results of different phases, including supercritical water. It was found that while HDO of aromatic compounds like phenol and 2-hydroxybenzaldehyde was favourable in the supercritical phase, smaller molecules like acetic acid and hydroxyacetone did not show much advantage in the supercritical phase over gas and liquid phase. It was also found that the thermochemical parameter - Gibbs free energy change (ΔG) was equally influenced by the solvation effect and the effect of temperature-pressure under supercritical conditions. In several instances, the two effects were found to offset each other in the supercritical phase.

DFT study on dibenzofuran conversion to cyclohexane and benzene in gas, water and methanol solvents.

Hydrodeoxygenation (HDO) of dibenzofuran is studied using density functional theory (DFT) with SMD implicit solvation model to give products like benzene and cyclohexane. Water and methanol are used as solvents, and gas-phase calculations are also performed for comparison. HDO is proposed via two major route- 1) saturation of phenyl rings followed by deoxygenation, 2) cleavage of C-O bond for oxy -removal followed by ring saturation. First, thermochemical parameters like Gibbs free energy change and enthalpy change are calculated at varying temperature in gas, water and methanol. Then, the solvation free energy is calculated for methanol and water to determine the ease of separation of products from the solvents. Finally, a Single Point Energy calculation is performed in the gas phase. It is found that water is the best solvent among three mediums and offers a significant advantage in some reactions such as the conversion of dibenzofuran to 1,2,3,4,4a,9b-hexahydrodibenzo[b,d]furan, 2-cyclohexyl phenol to bicyclo-hexan-2-ol and biphenyl to cyclohexyl benzene. Water is also found to be a better medium for separation of all the intermediates and products from the solvent over methanol. The effect of temperature was also studied, and it was found that the increase in temperature is unfavourable in all mediums for almost all reactions.

DFT investigation on thermochemical analyses of conversion of xylose to linear alkanes in aqueous phase.

Xylose is an integral part of hemicellulose fraction of lignocellulosic biomass. Its abundance in the lignocellulose makes it a desirable component for converting into various value-added compounds. In this study, conversion of xylose to four linear alkanes has been discussed by five different schemes including their thermochemistry under the framework of density functional theory. Main products are butane, pentane, octane and tridecane whereas the intermediate products include furfural, tetrahydrofuran, pentane-1,5-diol, etc. The simulations have been performed at B3LYP/6-31 + g(d,p) and M06-2X/6-31 + g(d,p) level of theories in aqueous phase using SMD solvation model. Thermochemical parameters (ΔG, ΔH and Keq) are obtained at a wide range of temperature, i.e. 298-698 K. Single point energy change (ΔE) of all the conversion steps has also been calculated at M05-2X/6-311++g(3df,2p) level of theory in the aqueous phase. It is observed that temperature plays a vital role in the formation of products. At high temperature, only scheme RS 1 (i.e. xylose to butane) can proceed to produce butane. The absolute difference between two functionals, B3LYP and M06-2X, was found to be small (<2 kcal/mol) for ring opening reactions making both the functionals suitable for a qualitative study. For saturation of cyclic compounds, a large difference (>10 kcal/mol) was observed between the two functionals making higher accuracy method more suitable for them. For all other reactions, use of M06-2X can be preferred.

Quantum chemical study on gas phase pyrolysis of p-isopropenylphenol.

In the pyrolysis of Sphagnum moss species, p-isopropenylphenol (p-IPP) is a major product which has been considered in this density functional theory based computational study for its conversion to various products such as benzene, phenol, 4-propenylphenol, indan-5-ol, 4-propylcyclohexanone, 4-cyclopropylphenol, etc. In order to achieve these products, eight different reaction schemes are performed using B3LYP/6-311 + g (d,p) level of theory. Further, thermodynamic properties such as reaction free energies and reaction enthalpies associated with these eight reaction schemes are developed in the temperature range of 298-898 K. The reaction schemes that include partial hydrogenation of the aromatic carbon followed by elimination of functional groups are found to demand low activation energy. The production of benzene from p-IPP with isopropenylbenzene as an intermediate product requiring only 19.83 kcal/mol of activation energy is the rate limiting reaction step. Indan-5-ol produced from p-IPP is validated with the literature results and found excellent agreement between two results. Furthermore, the temperature is found to have phenomenal effect in each reaction scheme.

Computational Study on Ring Saturation of 2-Hydroxybenzaldehyde Using Density Functional Theory.

Bio-oil produced from pyrolysis of lignocellulosic biomass consists of several hundreds of oxygenated compounds resulting in a very low quality with poor characteristics of low stability, low pH, low stability, low heating value, high viscosity, and so on. Therefore, to use bio-oil as fuel for vehicles, it needs to be upgraded using a promising channel. On the other hand, raw bio-oil can also be a good source of many specialty chemicals, e.g., 5-HMF, levulinic acid, cyclohexanone, phenol, etc. In this study, 2-hydroxybenzaldehyde, a bio-oil component that represents the phenolic fraction of bio-oil, is considered as a model compound and its ring saturation is carried out to produce cyclohexane and cyclohexanone along with various other intermediate products using density functional theory. The geometry optimization, vibrational frequency, and intrinsic reaction coordinate calculations are carried out at the B3LYP/6-311+g(d,p) level of theory. Furthermore, a single point energy calculation is performed at each structure at the M06-2X/6-311+g(3df,2p)//B3LYP/6-311+g(d,p) level of theory to accurately predict the energy requirements. According to bond dissociation energy calculations, the dehydrogenation of formyl group of 2-hydroxybenzaldehyde is the least energy demanding bond cleavage. The production of cyclohexane has a lower energy of activation than the production of cyclohexanone.