Pore-throat characterization of unconsolidated porous media using watershed-segmentation algorithm

Transfer Learning Approach and Nucleus Segmentation with MedCLNet Colon Cancer Database

Machine learning has been recently used especially in the medical field. In the diagnosis of serious diseases such as cancer, deep learning techniques can be used to reduce the workload of experts and to produce quick solutions. The nuclei found in the histopathology dataset are an essential parameter in disease detection. The nucleus segmentation was performed using the colorectal histology MNIST dataset for nucleus detection in this study. The graph theory, PSO, watershed, and random walker algorithms were used for the segmentation process. In addition, we present the 10-class MedCLNet visual dataset consisting of the NCT-CRC-HE-100 K dataset, LC25000 dataset, and GlaS dataset that can be used in transfer learning studies from deep learning techniques. The study proposes a transfer learning technique using the MedCLNet database. Deep neural networks pre-trained with the proposed transfer learning method were used in the classification with the colorectal histology MNIST dataset in the experimental process. DenseNet201, DenseNet169, InceptionResNetV2, InceptionV3, ResNet152V2, ResNet101V2, and Xception deep learning algorithms were used in transfer learning and the classification studies. The proposed approach was analyzed before and after transfer learning with different methods (DenseNet169 + SVM, DenseNet169 + GRU). In the performance measurement, using the colorectal histology MNIST dataset, 94.29% accuracy was obtained in the DenseNet169 model, which was initiated with random weights in the multi-classification study, and 95.00% accuracy after transfer learning was applied. In comparison with the results obtained from empirical studies, it was demonstrated that the proposed method produced satisfactory outcomes. The application is expected to provide a secondary evaluation for physicians in colon cancer detection and the segmentation.

Stereolithography 3D Printer for Micromodel Fabrications with Comprehensive Accuracy Evaluation by Using Microtomography

Micromodels are important for studying various pore-scale phenomena in hydrogeology. However, the fabrication of a custom micromodel involves complicated steps with cost-prohibitive equipment. The direct fabrication of micromodels with a 3D printer can accelerate the fabrication steps and reduce the cost. A stereolithography (SLA) 3D printer is one of the best options because it has sufficient printing performance for micromodel fabrication and is relatively inexpensive. However, it is not without drawbacks. In this report, we explored the capability of an SLA 3D printer for micromodel fabrication. Various parameters affecting the printing results, such as the effects of geometries, dimensions, printing axis configurations, printing thickness resolutions, and pattern thicknesses were investigated using microtomography for the first time. Eventually, the most optimal printing configuration was then also discussed. In the end, a complete micromodel was printed, assembled, and used for fluid displacement experiments. As a demonstration, viscous and capillary fingerings were successfully performed using this micromodel design.

DEM investigation of the effect of particle breakage on compact properties

Particle breakage during compaction affects compaction behavior and the quality of the formed compact. This work conducted a numerical study based on the discrete element method (DEM) to investigate the effect of particle breakage on compaction dynamics and compact properties, including particle size and density distributions, and pore properties. A force-based breakage criterion and Apollonian sphere packing algorithm were employed to characterize particle breakage behavior. The pore structures of the compacts were characterized by the watershed pore segmentation method. Calibrated with experimental data, the model was able to simulate the stress-strain relation comparable with experimental observation. During compaction, the particles were gradually broken from top to bottom with increasing pressure. Both density and pore size of the compacts had relatively uniform distribution at larger stress, while the pore size decreased sharply when the particles started to break, indicating that the smaller fragments in the compact system have a significant effect on the pore size distribution.