BazTech - Yadda

1

Automated detection of multi-class urinary sediment particles: An accurate deep learning approach

Lyu He, Xu Fanxin, Jin Tao, Zheng Siyi, Zhou Chenchen, Cao Yang, Luo Bin, Huang Qinzhen, Xiang Wei, Li Dong

Biocybernetics and Biomedical Engineering

|

2023

|

Vol. 43, no. 4

672--683

EN

Urine microscopy is an essential diagnostic tool for kidney and urinary tract diseases, with automated analysis of urinary sediment particles improving diagnostic efficiency. However, some urinary sediment particles remain challenging to identify due to individual variations, blurred boundaries, and unbalanced samples. This research aims to mitigate the adverse effects of urine sediment particles while improving multi-class detection performance. We proposed an innovative model based on improved YOLOX for detecting urine sediment particles (YUS-Net). The combination of urine sediment data augmentation and overall pre-trained weights enhances model optimization potential. Furthermore, we incorporate the attention module into the critical feature transfer path and employ a novel loss function, Varifocal loss, to facilitate the extraction of discriminative features, which assists in the identification of densely distributed small objects. Based on the USE dataset, YUS-Net achieves the mean Average Precision (mAP) of 96.07%, 99.35% average precision, and 96.77% average recall, with a latency of 26.13 ms per image. The specific metrics for each category are as follows: cast: 99.66% AP; cryst: 100% AP; epith: 92.31% AP; epithn: 100% AP; eryth: 92.31% AP; leuko: 99.90% AP; mycete: 99.96% AP. With a practical network structure, YUS-Net achieved efficient, accurate, end-to-end urinary sediment particle detection. The model takes native high-resolution images as input without additional steps. Finally, a data augmentation strategy appropriate for the urinary microscopic image domain is established, which provides a novel approach for applying other methods in urine microscopic images.

2

Improving the plane‑wave destruction method by parameterizing the local slope for VSP wavefield separation

Guo Yinling, Peng Suping, Du Wenfeng, Li Dong, Li Chuangjian

Acta Geophysica

|

2023

|

Vol. 71, no. 4

1659--1673

EN

High-precision seismic exploration has attracted a great deal of attention. Vertical seismic profiles have the advantages of high exploration accuracy and ideal imaging effects, making them an effective tool for mine exploration. The separation of upgoing and downgoing waves is a vital step in vertical seismic profile processing. However, wavefield separation processing in the frequency–wavenumber domain often leads to aliasing in the time–distance domain. In addition, it is likely to result in incomplete wavefield separation in the sparse domain because of inaccurate threshold selection. As an effective time–distance domain wavefield separation method, the plane-wave destruction method can effectively avoid these problems. However, the traditional local slope estimation in plane-wave destruction is data-driven, making it difficult to distinguish linear upgoing and downgoing waves. Therefore, the velocity-based local slope parameterization method is proposed to construct a plane-wave destruction filter that can be used to separate the vertical seismic profile wavefield. Synthetic and field data show that the proposed separating strategy has higher separation accuracy than traditional methods when separating the upgoing and downgoing waves of vertical seismic profile data.

3

Effect of Die Wall Lubrication on High Velocity Compaction Behavior and Sintering Properties of Fe-Based PM Alloy

Liu Zili, Li Dong, Liu Xiqin, Li Haohao, Huang Xin, Tang Zhihao, Zou Yuwen

Archives of Metallurgy and Materials

|

2020

|

Vol. 65, iss. 2

677--684

EN

Fe-based PM alloy powder of Fe-2.5Ni-0.5Mo-2Cu-0.4C was pressed by high velocity compaction combined with die wall lubrication, and the effect of die wall lubrication on high velocity compaction behavior and sintering properties of the Fe-based PM alloy were studied. The results indicate that the impact force, green density, sintered density of samples increase with the augment of the impact velocity and die wall lubrication. Compared with that without die wall lubrication, the green density and sintered density of the sample with die wall lubrication are about 0.07-0.12 g/cm3 and 0.08~0.11 g/cm3 higher at the same impact velocity, respectively, while the ejection force of the die wall lubricated sample is much smaller, and reduced about 26%~36%. The green compact with die wall lubrication has much fewer porosity than that without die wall lubrication, and more mechanical bonding and cold welding regions are observed. The sintered samples mainly consists of gray pearlite and white ferrite, and more pearlite is observed in the sintered sample with die wall lubrication.

4

Effect and mechanism of citric acid on flotation separation of siderite and hematite

Han Huili, Yin Wanzhong, Yao Jin, Li Dong

Physicochemical Problems of Mineral Processing

|

2019

|

Vol. 55, iss. 1

311--323

EN

Heterocoagulation can occur between fine siderite and hematite particles, which would result in the low efficiency of their separation during the flotation process. To date, there have been no mature methods to increase their separation efficiency. In this paper, citric acid was used as a regulator to enhance the slurry dispersion efficiency. Micro-flotation, scanning electron microscopy (SEM) analysis, settling tests, particle size measurements, zeta potential measurements and E-DLVO theoretical calculations were conducted in the investigations. A maximum recovery difference (53.98%) between siderite and hematite in their mixtures flotation was obtained. Settling tests confirmed that citric acid contributed to improving the dispersion degree of the slurry. SEM analysis indicated that citric acid could clean the surface of particles and weaken the coagulation between siderite and hematite, which were in line with the results of particle size measurements. The zeta potential measurements and Extended-Derjaguin-Landau-Verwey-Overbeek (E-DLVO) theoretical calculations indicated that the citric acid could adsorb on the siderite and hematite surfaces and decreased the surface charge, resulting in a visible increase of the repulsion energy between siderite and hematite particles. Therefore, citric acid can be applied to remove the easily-ground carbonate minerals first to improve the flotation performance of hematite in the separation process of carbonate-containing iron ores.