Applied Machine Learning in Spiral Breast-CT: Can We Train a Deep Convolutional Neural Network for Automatic, Standardized and Observer Independent Classification of Breast Density?

The aim of this study was to investigate the potential of a machine learning algorithm to accurately classify parenchymal density in spiral breast-CT (BCT), using a deep convolutional neural network (dCNN). In this retrospectively designed study, 634 examinations of 317 patients were included. After image selection and preparation, 5589 images from 634 different BCT examinations were sorted by a four-level density scale, ranging from A to D, using ACR BI-RADS-like criteria. Subsequently four different dCNN models (differences in optimizer and spatial resolution) were trained (70% of data), validated (20%) and tested on a “real-world” dataset (10%). Moreover, dCNN accuracy was compared to a human readout. The overall performance of the model with lowest resolution of input data was highest, reaching an accuracy on the “real-world” dataset of 85.8%. The intra-class correlation of the dCNN and the two readers was almost perfect (0.92) and kappa values between both readers and the dCNN were substantial (0.71–0.76). Moreover, the diagnostic performance between the readers and the dCNN showed very good correspondence with an AUC of 0.89. Artificial Intelligence in the form of a dCNN can be used for standardized, observer-independent and reliable classification of parenchymal density in a BCT examination.

Systematic study of the (n,2n) reaction cross section for 121Sb and 123Sb isotopes

The cross sections of the 121Sb(n,2n) 120Sbm and 123Sb(n,2n) 122Sb reactions were measured at 12.50, 15.79 and 18.87 MeV neutron energies relative to the standard 27Al(n,α) 24Na monitor reaction using neutron activation and offline γ-ray spectrometry technique. Irradiations of the samples were performed at the BARC-TIFR Pelletron Linac Facility, Mumbai, India. The quasi-monoenergetic neutron was generated via the 7Li(p,n) reaction. Statistical model calculations were performed by nuclear reaction codes TALYS (ver. 1.9) and EMPIRE (ver. 3.2.2) using various input parameters and nuclear level density models. The cross sections of the ground and the isomeric state as well as the isomeric cross section ratio were studied theoretically from reaction threshold to 26 MeV energies. The effect of pre-equilibrium emission is also discussed in detail using different theoretical models. The present measured cross section were discussed and compared with reported experimental data and evaluation data of the JEFF-3.3, ENDF/B-VIII.0, JENDL/AD-2017 and TENDL-2019 libraries. A detailed analysis of the uncertainties in the measured cross section data was performed using the covariance analysis method. Furthermore, a systematic study of the (n,2n) reaction cross section for 121Sb and 123Sb isotopes were also performed within 14-15 MeV neutron energies using various systematic formulae. This work helps to overcome discrepancies in Sb data and illustrate a better understanding of pre-equilibrium emission in (n,2n) reaction channel.

Nuclear level densities away from line of β-stability

The variation of total nuclear level densities (NLDs) and level density parameters with proton number Z are studied around the β-stable isotope, Z0, for a given mass number. We perform our analysis for a mass range A=40 to 180 using the NLDs from popularly used databases obtained with the single-particle energies from two different microsopic mass-models. These NLDs which include microscopic structural effects such as collective enhancement, pairing and shell corrections, do not exhibit inverted parabolic trend with a strong peak at Z0 as predicted earlier. We also compute the NLDs using the single-particle energies from macroscopic-microscopic mass-model. Once the collective and pairing effects are ignored, the inverted parabolic trends of NLDs and the corresponding level density parameters become somewhat visible. Nevertheless, the factor that governs the (Z-Z0) dependence of the level density parameter, leading to the inverted parabolic trend, is found to be smaller by an order of magnitude. We further find that the (Z-Z0) dependence of NLDs is quite sensitive to the shell effects.

Effects of some level density models and γ-ray strength functions on production cross-section calculations of 16,18O and 24,26Mg radioisotopes

Oxygen and magnesium isotopes can be used in nuclear reactor materials as cooling, shielding, coating, electronics etc. They can also occur through nuclear reactions during the reactor operation. The exposure of high energy gamma can change the material and its properties, and hence its objective of selection may not remain satisfied. Thus, it is required to study the cross section of different reactions on nuclear reactor materials to understand their sustainability for the properties, for which they are chosen. In the scope of this study, theoretically, different level density model calculations and γ-ray strength functions have been performed for (γ, p) reaction for 16,18O and 24,26Mg nuclei using TALYS 1.9 and EMPI˙RE 3.2.2 codes. Also, semi empirical (γ, p) formula by Tel et al., have been calculated and compared with all results. The effect of different level density models defined in these codes on gamma strength has been studied. Finally, the consistency of these obtained data with EXFOR data have been investigated.

Shell-structure and asymmetry effects in level densities

Level density [Formula: see text] is derived for a nuclear system with a given energy [Formula: see text], neutron [Formula: see text], and proton [Formula: see text] particle numbers, within the semiclassical extended Thomas–Fermi and periodic-orbit theory beyond the Fermi-gas saddle-point method. We obtain [Formula: see text], where [Formula: see text] is the modified Bessel function of the entropy [Formula: see text], and [Formula: see text] is related to the number of integrals of motion, except for the energy [Formula: see text]. For small shell structure contribution one obtains within the micro–macroscopic approximation (MMA) the value of [Formula: see text] for [Formula: see text]. In the opposite case of much larger shell structure contributions one finds a larger value of [Formula: see text]. The MMA level density [Formula: see text] reaches the well-known Fermi gas asymptote for large excitation energies, and the finite micro-canonical limit for low excitation energies. Fitting the MMA [Formula: see text] to experimental data on a long isotope chain for low excitation energies, due mainly to the shell effects, one obtains results for the inverse level density parameter [Formula: see text], which differs significantly from that of neutron resonances.

Activation cross sections of some neutron-induced reactions in the energy range of 13.82–14.71 MeV

The cross sections of the reactions 70Zn(n,2n)69mZn,74Ge(n,α)71mZn and 90Zr(n,2n)89m,gZr in the energy range of 13.82–14.71 MeV were measured by the activation technique in combination with HPGe detector γ-ray spectroscopy. The measured results were compared with other experimental data and with the data given in the library TENDL-2019. A nuclear model calculation based on the code TALYS-1.8 was also performed after adjustment of an input parameter for the spin distribution of level density. The results of this work strengthen the database and could be useful in further evaluation of the data.

Level Density of the Pre Compound Nucleus Calculations for [28 60Ni], [28 58Ni], Using Exciton Model

The level density for the pre compound [28 60Ni],[2858Ni], at certain exciton number, are calculated for one &two component system based on Ericson's formula. The Pauli, Slight and Pairing correction are examined with the behavior of the level density for one & two components. The particle – hole state density has been calculated by means of the energy dependence of excited number in [28 60Ni], [2858Ni]. The present results are compared between one &two component with Pauli, Slight and Pairing correction. The Pairing correction has a major effect in the present calculations. Keyword: level density, Exciton model, [28 60Ni],[2858Ni], , Formula ESM.

A Data-Driven Approach to the Development and Understanding of Chiroptical Sensors for Alcohols with Remote γ-Stereocenters

Dynamic covalent chemistry-based sensors have recently emerged as powerful tools to rapidly determine the enantiomeric excess of organic small molecules. While a bevy of sensors have been developed, those for flexible molecules with stereocenters remote to the functional group that binds the chiroptical sensor remain scarce. In this study, we develop an iterative, data-driven workflow to design and analyze a chiroptical sensor capable of assessing challenging acyclic γ-stereogenic alcohols. Fol-lowing sensor optimization, the mechanism of sensing was probed with a combination of computational parameterization of the sensor molecules, statistical modeling, and high-level density functional theory (DFT) calculations. These were used to elucidate the mechanism of stereochemical recognition and revealed that competing attractive non-covalent interactions (NCIs) determine the overall performance of the sensor. It is anticipated that the data-driven workflows developed herein will be generally applicable to the development and understanding of dynamic covalent and supramolecular sensors.