We present a new method of predicting the ages of galaxies using an artificial neural network (ANN) with an evaluation of the methods such that future work can be derived for other properties of galaxies with ANNs. We train our ANN to recognise the patterns between the equivalent widths of spectral indices and the mass-weighted ages of galaxies from the Galaxy and Mass Assembly (GAMA) survey estimated by the MAGPHYS model. We quantify the quality of our predictions by calculating the mean squared error as 0.02, mean absolute error as 0.11, R-squared score as 0.53 and presenting prediction uncertainties which show our network performs well. Finally, we describe how the predictions are physically motivated as they follow trends already present in the properties and EWs of the galaxies which shows the network is able to meaningfully derive age from the EWs of galactic spectra.

In this work, we present a new approach to produce spectroscopic constants and model first-principles synthetic spectra for all molecules of astrophysical interest. We have generalized our previous diatomic molecule simulation framework, employing transition-optimized shifted Hermite (TOSH) theory, thereby enabling the modeling of polyatomic rotational constants for molecules with three or more atoms. These capabilities are now provided by our new code Epimetheus. As a first validation of our approach, we confront our predictions and assess their accuracy against the well-studied triatomic molecule ozone 666 ( ^16 O _3 ), in addition to eight of its potential isotopomers: ozone 668 ( ^16 O ^16 O ^18 O), 686 ( ^16 O ^18 O ^16 O), 667 ( ^16 O ^16 O ^17 O), 676 ( ^16 O ^17 O ^16 O), 688 ( ^16 O ^18 O ^18 O), 868 ( ^18 O ^16 O ^18 O), 888 ( ^18 O _3 ), and 777 ( ^17 O _3 ). We then assess the accuracy of these rotational constants using the Epimetheus data in our code Pandora, and generate synthetic molecular spectra. The ozone spectra presented here are purely infrared and not Raman. Epimetheus builds upon the work from our previous code Prometheus, which used the TOSH theory to account for anharmonicity for the fundamental ν = 0 → ν = 1 band, going further to now account for triatomic molecules. This is combined with thermal profile modeling for the rotational transitions. We have found that this extended method performs well, typically approximating the spectroscopic constants with errors of less than 2%. Some issues do arise depending on the symmetry group of the ozone isotopomer. From these spectroscopic constants and using our own spectral modeling code, we show that we can provide the data to produce appreciable molecular spectra, which are good approximations until high-resolution studies can be done.

For this poster I will be presenting the approaches taken to produce synthetic spectra for molecules believed to be biosignatures. In astrochemistry, biosignatures describe a group of molecules which could be produced by life and therefore act as an indication of it. Naturally, by being able to identify these molecules it will be possible to screen exoplanets for the possibility of harbouring life. Recently, Seager, Bains and Petkowski (2016) produced a catalogue of possible biosignatures, which totalled above 14,000 molecules. There are groups such as the ExoMol group which model these molecules at extremely high precision. However, due to the large number of molecules, a high precision method would take an unreasonable amount of time and therefore a faster means of producing spectra is required. The work shown within is designed to be quick and produce data for the full rovibronic spectrum rather than selected bands. I will show that anharmonic corrections are needed to be able to simulate qualitatively correct ro-vibrational transitions. In my method, this is done by implementing TOSH, transition optimised shifted hermites, which was first detailed by Lin, Gilbert and Gill (2007). Finally I will be presenting my latest results using this analytical anharmonic approach.

This work presents the application of an Encoder-Decoder convolutional neural network (ED-CNN) model to automatically segment COVID-19 computerised tomography (CT) data. By doing so we are producing an alternative model to current literature, which is easy to follow and reproduce, making it more accessible for real-world applications as little training would be required to use this. Our simple approach achieves results comparable to those of previously published studies, which use more complex deep-learning networks. We demonstrate a high-quality automated segmentation prediction of thoracic CT scans that correctly delineates the infected regions of the lungs. This segmentation automation can be used as a tool to speed up the contouring process, either to check manual contouring in place of a peer checking, when not possible or to give a rapid indication of infection to be referred for further treatment, thus saving time and resources. In contrast, manual contouring is a time-consuming process in which a professional would contour each patient one by one to be later checked by another professional. The proposed model uses approximately 49 k parameters while others average over 1,000 times more parameters. As our approach relies on a very compact model, shorter training times are observed, which make it possible to easily retrain the model using other data and potentially afford “personalised medicine” workflows. The model achieves similarity scores of Specificity (Sp) = 0.996 ± 0.001, Accuracy (Acc) = 0.994 ± 0.002 and Mean absolute error (MAE) = 0.0075 ± 0.0005.

In this paper we provide a highly accurate value for the binding energy of benzene to proton-ordered crystalline water ice (XIh), as a model for interstellar ices. We compare our computed value to the latest experimental data available from temperature programmed desorption (TPD) experiments and find that our binding energy value agrees well with data obtained from binding to either crystalline or amorphous ice. Importantly, our new value is lower than that used in most astrochemical networks by about nearly half its value. We explore the impact of this revised binding energy value for both an AGB outflow and a protoplanetary disk. We find that the lower value of the binding energy predicted here compared with values used in the literature (4050 K versus 7587 K) leads to less depletion of gas-phase benzene in an AGB outflow, and leads to a shift outwards in the benzene snowline in the midplane of a protoplanetary disk. Using this new value, the AGB model predicts lower abundances of benzene in the solid phase throughout the outflow. The disk model also predicts a larger reservoir of gas-phase benzene in the inner disk, which is consistent with the recent detections of benzene for the first time in protoplanetary disks with JWST.

We present an application of autoencoders to the problem of noise reduction in single-shot astronomical images and explore its suitability for upcoming large-scale surveys. Autoencoders are a machine learning model that summarises an input to identify its key features, then from this knowledge predicts a representation of a different input. The broad aim of our autoencoder model is to retain morphological information (e.g., non-parametric morphological information) from the survey data whilst simultaneously reducing the noise contained in the image. We implement an autoencoder with convolutional and maxpooling layers. We test our implementation on images from the Panoramic Survey Telescope and Rapid Response System (Pan-STARRS) that contain varying levels of noise and report how successful our autoencoder is by considering Mean Squared Error (MSE), Structural Similarity Index (SSIM), the second-order moment of the brightest 20 percent of the galaxy's flux M20, and the Gini coefficient, whilst noting how the results vary between the original images, stacked images, and noise reduced images. We show that we are able to reduce noice, over many different targets of observations, whilst retaining the galaxy's morphology, with metric evaluation on a target by target analysis. We establish that this process manages to achieve a positive result in a matter of minutes, and by only using one single shot image compared to multiple survey images found in other noise reduction techniques.

Objective: Gamma passing rate (GPR) predictions using machine learning methods have been explored for treatment verification of radiotherapy plans. However, these methods presented datasets with unbalanced number of plans having different treatment conditions (heterogeneous datasets), such as anatomical sites or dose per fractions, leading to lower model interpretability and prediction performance. Methods: We investigated the impact of the dataset composition on GPR binary classification (pass/fail) using random forest (RF), XG-boost, and neural network (NN) models. 945 plans were used to create one reference dataset (randomly assembled) and 24 customized datasets that considered four heterogeneity factors independently (anatomical region, number of arcs, dose per fraction, and treatment unit). 309 predictor features were extracted and calculated from plan parameters, modulation complexity metrics, and radiomic analysis (leave-trajectory maps, 3D dose distributions, and portal dosimetry images). The models’ performances were measured using the area under the curve from the receiver operating characteristic (ROC-AUC). Results: Radiomics features for reference models increased ROC-AUC values up to 13%, 15%, and 5% for RF, XG-Boost, and NN, respectively. The datasets with higher heterogeneous conditions presented the lower ROC-AUC values (RF: 0.72 ± 0.11, XG-Boost: 0.67 ± 0.1, NN: 0.89 ± 0.05) compared to models with less heterogeneous treatment conditions (RF: 0.88 ± 0.06, XG-Boost: 0.89 ± 0.07, NN: 0.98 ± 0.01). The ten most important features for each heterogeneity dataset group demonstrated their correlation with the treatments’ physical aspects and GPR prediction. Conclusion: Improvements in data generalization and model performances can be associated with datasets having similar treatment conditions. This analysis might be implemented to evaluate the dataset quality and model consistency of further ML applications in radiotherapy. Advances in knowledge: Dataset heterogeneities decrease ML model performance and reliability.

We present a practical workflow to compute the potential energy curve of the hydrogen molecule on near intermediate-scale quantum (NISQ) devices. The proposed approach uses an extrapolation scheme to deliver, with only few qubits, full configuration interaction results close to the basis-set limit. We show that despite the limitations imposed by the noisy nature of simulated quantum hardware, it is possible to recover realistic electronic correlation values, if we also estimate expectation values of the Hartree-Fock ground state energy. Using two models of noisy quantum experiments, we evaluate the performance of a scheme that requires at most a double-zeta basis set (3-21G, in this case) and compare with the most accurate Born-Oppenheimer potential energy curves available in the literature. Our flexible approach is implemented using simple variational ansatzes combined with straightforward mitigation techniques and thus we expect it to be also suitable for other energy estimation quantum schemes.

A Smolyak algorithm adapted to system-bath separation is proposed for rigorous quantum simulations. This technique combines a sparse grid method with the system-bath concept in a specific configuration without limitations on the form of the Hamiltonian, thus achieving a highly efficient convergence of the excitation transitions for the "system" part. Our approach provides a general way to overcome the perennial convergence problem for the standard Smolyak algorithm and enables the simulation of floppy molecules with more than a hundred degrees of freedom.The efficiency of the present method is illustrated on the simulation of H$_2$ caged in an sII clathrate hydrate including two kinds of cage modes. The transition energies are converged by increasing the number of normal modes of water molecules. Our results confirm the triplet splittings of both translational and rotational ($j=1$) transitions of the H$_2$ molecule. Furthermore, they show a slight increase of the translational transitions with respect to the ones in a rigid cage.

The development of actively tunable plasmonic nanostructures enables real-time reconfigurable and on-demand enhancement of optical signals. This is an essential requirement for a wide range of applications such as sensing and nanophotonic devices, for which electrically driven tunability is required. By modifying the transition energies of a material via the application of an electric field, the Stark effect offers a reliable and practical approach to achieve such tunability. In this work, we report on the use of the Stark effect to control the scattering response of a plasmonic nanogap formed between a silver nanoparticle and an extended silver film separated by a thin layer of the organic semiconductor PQT-12. The plasmonic response of such nanoscattering sources follows the quadratic Stark shift. In addition, our approach allows one to experimentally determine the polarizability of the semiconductor material embedded in the nanogap region, offering a new approach to probe the excitonic properties of extremely thin semiconducting materials such as 2D materials under applied external electric field with nanoscale resolution.

This work presents the first steps to modeling synthetic rovibrational spectra for all molecules of astrophysical interest using a new approach implemented in the Prometheus code. The goal is to create a new comprehensive source of first-principles molecular spectra, thus bridging the gap for missing data to help drive future high-resolution studies. Our primary application domain is for molecules identified as signatures of life in planetary atmospheres (biosignatures), but our approach is general and can be applied to other systems. In this work we evaluate the accuracy of our method by studying four diatomic molecules, H2, O2, N2, and CO, all of which have well-known spectra. Prometheus uses the transition-optimised shifted Hermite (TOSH) theory to account for anharmonicity for the fundamental ν = 0 → ν = 1 band, along with thermal-profile modeling for the rotational transitions. To this end, we expand TOSH theory to enable the modeling of rotational constants. We show that our simple model achieves results that are a better approximation of the real spectra than those produced through an harmonic approach. We compare our results with high-resolution HITRAN and ExoMol spectral data. We find that modeling accuracy tends to diminish for rovibrational transition away from the band origin, thus highlighting the need for the theory to be further adapted.

The chelator Bn2DT3A was used to produce a novel 68Ga complex for positron emission tomography (PET). Unusually, this system is stabilized by a coordinated hydroxide in aqueous solutions above pH 5, which confers sufficient stability for it to be used for PET. Bn2DT3A complexes Ga3+ in a hexadentate manner, forming a mer-mer complex with log K([Ga(Bn2DT3A)]) = 18.25. Above pH 5, the hydroxide ion coordinates the Ga3+ ion following dissociation of a coordinated amine. Bn2DT3A radiolabeling displayed a pH-dependent speciation, with [68Ga][Ga(Bn2DT3A)(OH)]- being formed above pH 5 and efficiently radiolabeled at pH 7.4. Surprisingly, [68Ga][Ga(Bn2DT3A)(OH)]- was found to show an increased stability in vitro (for over 2 h in fetal bovine serum) compared to [68Ga][Ga(Bn2DT3A)]. The biodistribution of [68Ga][Ga(Bn2DT3A)(OH)]- in healthy rats showed rapid clearance and excretion via the kidneys, with no uptake seen in the lungs or bones.

An entry from the Cambridge Structural Database, the world’s repository for small molecule crystal structures. The entry contains experimental data from a crystal diffraction study. The deposited dataset for this entry is freely available from the CCDC and typically includes 3D coordinates, cell parameters, space group, experimental conditions and quality measures.

Utilisation of CO2 as a chemical reagent is challenging, due to the molecule's inherent chemical stability. However, CO2 reacts promptly at high temperature (similar to 1000 degrees C) with alkaline-earth oxides to form carbonates and such reactions are used towards capture and re-utilisation. In this work, this concept is extended and CO2 is utilised as a reagent to modify the crystal structure of mixed-metal inorganic solids. Modification of the crystal structure is a "tool" used by materials scientists to tailor the physical property of solids. CO2 gas was reacted with several isostructural mixed-metal oxides Sr2CuO3, Sr1.8Ba0.2CuO3 and Ba2PdO3. These oxides are carefully selected to show anion vacancies in their crystal structure, to act as host sites for CO2 molecules, leading to the formation of carbonate anions, (CO3)(2-). The corresponding oxide carbonates were formed successfully and the favourable formation of SrCO3 as secondary phase was minimised via an innovative, yet simple synthetic procedure involving alternating of CO2 and air. We also derived a simple model to predict the kinetics of the reactions for the cuprates, using first-principles density functional theory and assimilating the reaction to a gas-surface process.

Machine learning (ML) methods have been implemented in radiotherapy to aid virtual specific-plan verification protocols, predicting gamma passing rates (GPR) based on calculated modulation complexity metrics because of their direct relation to dose deliverability. Nevertheless, these metrics might not comprehensively represent the modulation complexity, and automatically extracted features from alternative predictors associated with modulation complexity are needed. For this reason, three convolutional neural networks (CNN) based models were trained to predict GPR values (regression and classification), using respectively three predictors: (1) the modulation maps (MM) from the multi-leaf collimator, (2) the relative monitor units per control point profile (MUcp), and (3) the composite dose image (CDI) used for portal dosimetry, from 1024 anonymized prostate plans. The models' performance was assessed for classification and regression by the area under the receiver operator characteristic curve (AUC_ROC) and Spearman's correlation coefficient (r). Finally, four hybrid models were designed using all possible combinations of the three predictors. The prediction performance for the CNN-models using single predictors (MM, MUcp, and CDI) were AUC_ROC = 0.84 ± 0.03, 0.77 ± 0.07, 0.75 ± 0.04, andr= 0.6, 0.5, 0.7. Contrastingly, the hybrid models (MM + MUcp, MM + CDI, MUcp+CDI, MM + MUcp+CDI) performance were AUC_ROC = 0.94 ± 0.03, 0.85 ± 0.06, 0.89 ± 0.06, 0.91 ± 0.03, andr= 0.7, 0.5, 0.6, 0.7. The MP, MUcp, and CDI are suitable predictors for dose deliverability models implementing ML methods. Additionally, hybrid models are susceptible to improving their prediction performance, including two or more input predictors.

OBJECTIVE:High levels of beam modulation complexity (MC) and monitor units (MU) can compromise the plan deliverability of intensity-modulated radiotherapy treatments. Our study evaluates the effect of three treatment planning system (TPS) parameters on MC and MU using different multi-leaf collimator (MLC) architectures.METHODS:192 volumetric modulated arc therapy plans were calculated using one virtual prostate phantom considering three main settings: (1) three TPS-parameters (Convergence; Aperture Shape Controller, ASC; and Dose Calculation Resolution, DCR) selected from Eclipse v15.6, (2) four levels of dose-sparing priority for organs at risk (OAR), and (3) two treatment units with same nominal conformity resolution and different MLC architectures (Halcyon-v2 dual-layer MLC, DL-MLC & TrueBeam single-layer MLC, SL-MLC). We use seven complexity metrics to evaluate the MC, including two new metrics for DL-MLC, assessed by their correlation with γ passing rate (GPR) analysis.RESULTS:DL-MLC plans demonstrated lower dose-sparing values than SL-MLC plans (p<0.05). TPS-parameters did not change significantly the complexity metrics for either MLC architectures. However, for SL-MLC, significant variations of MU, target volume dose-homogeneity, and dose spillage were associated with ASC and DCR (p<0.05). MU were found to be correlated (highly or moderately) with all complexity metrics (p<0.05) for both MLC plans. Additionally, our new complexity metrics presented a moderate correlation with GPR (r<0.65). An important correlation was demonstrated between MC (plan deliverability) and dose-sparing priority level for DL-MLC.CONCLUSIONS:TPS-parameters selected do not change MC for DL-MLC architecture, but they might have a potential use to control the MU, PTV homogeneity or dose spillage for SL-MLC. Our new DL-MLC complexity metrics presented important information to be considered in future pre-treatment quality assurance programs. Finally, the prominent dependence between plan deliverability and priority applied to OAR dose sparing for DL-MLC needs to be analyzed and considered as an additional predictor of GPRs in further studies.ADVANCES IN KNOWLEDGE:Dose-sparing priority might influence in modulation complexity of DL-MLC.

With carbon dioxide (CO2) levels rising dramatically, climate change threatens marine environments. Due to increasing CO2 concentrations in the ocean, pH levels are expected to drop by 0.4 units by the end of the century. There is an urgent need to understand the impact of ocean acidification on chemical-ecological processes. To date, the extent and mechanisms by which the decreasing ocean pH influences chemical communication are unclear. Combining behaviour assays with computational chemistry, we explore the function of the predator related cue 2-phenylethylamine (PEA) for hermit crabs (Pagurus bernhardus) in current and end-of-the-century oceanic pH. Living in intertidal environments, hermit crabs face large pH fluctuations in their current habitat in addition to climate-change related ocean acidification. We demonstrate that the dietary predator cue PEA for mammals and sea lampreys is an attractant for hermit crabs, with the potency of the cue increasing with decreasing pH levels. In order to explain this increased potency, we assess changes to PEA's conformational and charge-related properties as one potential mechanistic pathway. Using quantum chemical calculations validated by NMR spectroscopy, we characterise the different protonation states of PEA in water. We show how protonation of PEA could affect receptor-ligand binding, using a possible model receptor for PEA (human TAAR1). Investigating potential mechanisms of pH-dependent effects on olfactory perception of PEA and the respective behavioural response, our study advances the understanding of how ocean acidification interferes with the sense of smell and thereby might impact essential ecological interactions in marine ecosystems.

We present a hybrid CCSD(T) + PBE-D3 approach to calculating the vibrational signatures for gas-phase benzene and benzene adsorbed on an ordered water ice surface. We compare the results of our method against experimentally recorded spectra and calculations performed using PBE-D3-only approaches (harmonic and anharmonic). Calculations use a proton ordered XIh water ice surface consisting of 288 water molecules, and results are compared against experimental spectra recorded for an ASW ice surface. We show the importance of including a water ice surface into spectroscopic calculations, owing to the resulting differences in vibrational modes, frequencies, and intensities of transitions seen in the IR spectrum. The overall intensity pattern shifts from a dominating nu(11) band in the gas-phase to several high-intensity carriers for an IR spectrum of adsorbed benzene. When used for adsorbed benzene, the hybrid approach presented here achieves an RMSD for IR active modes of 21 cm(-1), compared to 72 cm(-1) and 49 cm(-1) for the anharmonic and harmonic PBE-D3 approaches, respectively. Our hybrid model for gaseous benzene also achieves the best results when compared to experiment, with an RMSD for IR active modes of 24 cm(-1), compared to 55 cm(-1) and 31 cm(-1) for the anharmonic and harmonic PBE-D3 approaches, respectively. To facilitate assignment, we generate and provide a correspondence graph between the normal modes of the gaseous and adsorbed benzene molecules. Finally, we calculate the frequency shifts, Delta nu, of adsorbed benzene relative to its gas-phase to highlight the effects of surface interactions on vibrational bands and evaluate the suitability of our chosen dispersion-corrected density functional theory.