Nuclear magnetic resonance spectrum aided diagnosis based on DNN neural network

The nuclear magnetic resonance spectrum contains a variety of effective information, and most of the current clinical medicine uses nuclear magnetic resonance images as a diagnostic basis, but the spectral information is still not effectively explored. In order to improve the diagnostic results of nuclear magnetic resonance spectrum, this study uses DNN neural network as a technical support to extract effective information of nuclear magnetic resonance spectrum. Simultaneously, in order to improve the ability to describe the local features of the image, the traditional Crow algorithm is improved, and a similar target localization algorithm based on F-CroW is proposed. In addition, starting from the animal model of nasopharyngeal carcinoma and the serum and urine samples of patients with clinical nasopharyngeal carcinoma, this study designed a comparative study to study the performance of the proposed algorithm. According to the research and analysis, the DNN neural network proposed in this study has certain effects in the nuclear magnetic resonance spectrum analysis, which can be applied to clinical practice.

Optimization of Lignin Extraction Variables by Response Surface Methodology from Pine Saw Dust, and Quantification of Major Structural Units in Isolated Lignin Fraction

A synergistic combination of dioxane, acetic acid, and HCl was investigated for lignin extraction from pine wood biomass. After initial screening of reagent combination, response surface methodology (RSM) was used to optimize the lignin yield with respect to the variables of time 24–72 h, solids loading 5–15%, and catalyst dose 5–15 mL. A quadratic model predicted 8.33% of the lignin yield, and it was further confirmed experimentally and through the analysis of variance (ANOVA). Lignin at optimum combination exhibited features in terms of derivatization followed by reductive cleavage (DFRC) with a value of (305 µmol/gm), average molecular weights of 4358 and polydispersity of 1.65, and 2D heteronuclear single quantum coherence nuclear magnetic resonance spectrum (2D-HSQC NMR) analysis showing relative β-O-4 linkages (37.80%). From here it can be suggested that this fractionation can be one option for high quality lignin extraction from lignocellulosic biomass.

Магнитные свойства композита частиц DyMn-=SUB=-2-=/SUB=-O-=SUB=-5-=/SUB=--Mn-=SUB=-3-=/SUB=-O-=SUB=-4-=/SUB=-

The paper presents the results of the experimental studies of magnetic and resonance properties of the DyMn2O5 – Mn3O4 particles composite. There are two magnetic transitions at temperatures T1  65 K and T2  230 K. T1 differs from the transition temperatures for the initial materials, and this fact is associated with the interparticle interaction. T2 corresponds to the impurity phase DyMnO3 (1 at.%). Three peaks of microwave absorption are observed in the magnetic resonance spectrum, which is explained in the framework of the model of a two-phase system in magnetic relation. One resonance is attributed to Mn3O4 and the other two peaks refer to the ensemble of strongly anisotropic DyMn2O5 particles with a random distribution of anisotropy axes. Patrin Gennady Semionovich1,3 Mataev Mukhametkali Musagalievich2 Abdraimova Moldir Rashidovna2 Tursinova Zhanar Iliyasovna2 Kezdikbaeva Asel Taupykovna4 Shiyan Yaroslav Germanovich1,3 Plekhanov Vasil Granitovich1 1 Siberian Federal University, Svobodny pr. 79, Krasnoyarsk, 660041, Russia 2 Kazakh State Women’s Teacher Training University, Aiteke bi, 99, Almaty, 050000, Kazakhstan 3 L.V. Kirensky Institute of Physics, Federal Research Center KSC, Siberian Branch, Russian Academy of Sciences, Krasnoyarsk, 660036 Russia 4 E.A. Buketov Karaganda State University, st. University 28, City of Karaganda, 100028, Kazakhstan