Deep Learning-based Inaccuracy Compensation in Reconstruction of High Resolution XCT Data

Using X-ray computed tomography in pore structure characterization for a Berea sandstone: Resolution effect

Good dispersion of compounded ingredients in a rubber formulation is important for mechanical performance. After mixing, certain materials can remain undispersed within the rubber matrix, which could lead to critical flaws, influencing performance according to the Griffith failure criteria. High resolution X-ray computed tomography (XCT) offers a unique opportunity to measure phase domain size and distributions. Fillers such as carbon black or silica can be differentiated from sulfur or zinc oxide, providing an opportunity to determine dispersion characteristics of the various phases. The XCT technique has become an important characterization tool for three-dimensional and higher dimension material science due to the availability of polychromatic micro-focus X-ray sources and efficient high spatial resolution detectors with superior scintillators. High resolution XCT provides very rich data quantifying mixing efficiency of particulates in a matrix, such as insoluble sulfur or silica particles in rubber. Imaging with X-rays provides attenuation, phase, or scattering contrast and will prove to be a critical method for evaluating the field of rubber crosslinking, considering realistic environments in situ. This paper highlights methodology development and validation and provides insight on the dispersion of polymeric (insoluble) sulfur in rubber formulations. Dispersion assessment is compared using three techniques: high resolution XCT, population survival analysis in tensile testing, and optical microscopy.

The significance of enamel thickness in hominoid evolution has been plagued by the absence of nondestructive quantitative analysis. The aim of this investigation was to use a nondestructive method of analysis to document the volume of enamel and dentin, thereby providing a quantitative method for comparing both extant and extinct hominoid dentition. High-resolution X-ray computer tomography (HRXCT) is a nondestructive technique for visualizing and quantifying the interior of objects such as bone, teeth and minerals. HRXCT is also capable of obtaining digital information on their 3D geometries and volumetric properties. HRXCT differs from conventional medical CT-scanning in its ability to resolve details as small as a few microns in size, even when imaging objects made of high density materials like enamel and dentin. HRXCT also differs from micro-CT in its ability to examine large specimens up to 1.5 meters, with higher energy sources (typically 125–450 keV) that make the instrument capable of penetrating much denser objects including teeth and very heavily mineralized fossils. HRXCT offers several advantages over both the medical and micro-CT systems. Hominoid teeth used in this study consisted of a collection of extant hominoids as well as a number of fossil hominoids (Proconsul and Sivapithecus). HRXCT was used to obtain a data set of serially sectioned digitized images at a slice thickness of approximately 50 micrometers per section. The digital images were analyzed, and 3D reconstructions allowed for the collection of volumetric data for coronal enamel and dentin. HRXCT provides researchers with capabilities not found in other CT systems, which allows for a wider range of specimens sizes, with customizable scanning parameters unique to each specimen type. Our results demonstrate that HRXCT is an effective means by which volumetric data and 3D reconstruction of dental hard tissues are obtained from both extant and extinct hominoid dentitions.

Recommending music based on a user’s music preference is a way to improve user listening experience. Finding the correlation between the user data (e.g., location, time of the day, music listening history, emotion, etc.) and the music is a challenging task. In this paper, we propose an emotion-aware personalized music recommendation system (EPMRS) to extract the correlation between the user data and the music. To achieve this correlation, we combine the outputs of two approaches: the deep convolutional neural networks (DCNN) approach and the weighted feature extraction (WFE) approach. The DCNN approach is used to extract the latent features from music data (e.g., audio signals and corresponding metadata) for classification. In the WFE approach, we generate the implicit user rating for music to extract the correlation between the user data and the music data. In the WFE approach, we use the term-frequency and inverse document frequency (TF-IDF) approach to generate the implicit user ratings for the music. Later, the EPMRS recommends songs to the user based on calculated implicit user rating for the music. We use the million songs dataset (MSD) to train the EPMRS. For performance comparison, we take the content similarity music recommendation system (CSMRS) as well as the personalized music recommendation system based on electroencephalography feedback (PMRSE) as the baseline systems. Experimental results show that the EPMRS produces better accuracy of music recommendations than the CSMRS and the PMRSE. Moreover, we build the Android and iOS APPs to get realistic data of user experience on the EPMRS. The collected feedback from anonymous users also show that the EPMRS sufficiently reflect their preference on music.

Breast density is one of the most significant factors that is associated with cancer risk. In this study, our purpose was to develop a supervised deep learning approach for automated estimation of percentage density (PD) on digital mammograms (DMs). The input ‘for processing’ DMs was first log-transformed, enhanced by a multi-resolution preprocessing scheme, and subsampled to a pixel size of 800 µm × 800 µm from 100 µm × 100 µm. A deep convolutional neural network (DCNN) was trained to estimate a probability map of breast density (PMD) by using a domain adaptation resampling method. The PD was estimated as the ratio of the dense area to the breast area based on the PMD. The DCNN approach was compared to a feature-based statistical learning approach. Gray level, texture and morphological features were extracted and a least absolute shrinkage and selection operator was used to combine the features into a feature-based PMD. With approval of the Institutional Review Board, we retrospectively collected a training set of 478 DMs and an independent test set of 183 DMs from patient files in our institution. Two experienced mammography quality standards act radiologists interactively segmented PD as the reference standard. Ten-fold cross-validation was used for model selection and evaluation with the training set. With cross-validation, DCNN obtained a Dice’s coefficient (DC) of 0.79 ± 0.13 and Pearson’s correlation (r) of 0.97, whereas feature-based learning obtained DC = 0.72 ± 0.18 and r = 0.85. For the independent test set, DCNN achieved DC = 0.76 ± 0.09 and r = 0.94, while feature-based learning achieved DC = 0.62 ± 0.21 and r = 0.75. Our DCNN approach was significantly better and more robust than the feature-based learning approach for automated PD estimation on DMs, demonstrating its potential use for automated density reporting as well as for model-based risk prediction.

Additive manufacturing (AM) provides design flexibility and allows rapid fabrications of parts with complex geometries. The presence of internal defects, however, can lead to deficit performance of the fabricated part. X-ray Computed Tomography (XCT) is a non-destructive inspection technique often used for AM parts. Although defects within AM specimens can be identified and segmented by manually thresholding the XCT images, the process can be tedious and inefficient, and the segmentation results can be ambiguous. The variation in the shapes and appearances of defects also poses difficulty in accurately segmenting defects. This paper describes an automatic defect segmentation method using U-Net based deep convolutional neural network (CNN) architectures. Several models of U-Net variants are trained and validated on an AM XCT image dataset containing pores and cracks, achieving a best mean intersection over union (IOU) value of 0.993. Performance of various U-Net models is compared and analyzed. Specific to AM porosity segmentation with XCT images, several techniques in data augmentation and model development are introduced. This work demonstrates that, using XCT images, U-Net can be effectively applied for automatic segmentation of AM porosity with high accuracy. The method can potentially help improve quality control of AM parts in an industry setting.

Ore characterization is crucial for efficient and profitable production of mineral products from an ore deposit. Analysis is typically performed at various scales (meter to microns) in a sequential fashion, where sample volume is reduced with increasing spatial resolution due to the increasing costs and run times of analysis. Thus, at higher resolution, sampling and data quality become increasingly important to represent the entire ore deposit. In particular, trace metal mineral characterization requires high-resolution analysis, due to the typical very fine grain sizes (sub-millimeter) of trace metal minerals. Automated Mineralogy (AM) is a key technique in the mining industry to quantify process-relevant mineral parameters in ore samples. Yet the limitation to two-dimensional analysis of flat sample surfaces constrains the sampling volume, introduces an undesired stereological error, and makes spatial interpretation of textures and structures difficult. X-ray computed tomography (XCT) allows three-dimensional imaging of rock samples based on the x-ray linear attenuation of the constituting minerals. Minerals are visually differentiated though not chemically classified. In this study, decimeter to millimeter large ore samples were analyzed at resolutions from 45 to 1 μm by AM and XCT to investigate the potential of multi-scale correlative analysis between the two techniques. Mineralization styles of Au, Bi-minerals, scheelite, and molybdenite were studied. Results show that AM can aid segmentation (mineralogical classification) of the XCT data, and vice versa, that XCT can guide (sub-)sampling (e.g., for heavy trace minerals) for AM analysis and provide three-dimensional context to the two-dimensional quantitative AM data. XCT is particularly strong for multi-scale analysis, increasingly higher resolution scans of progressively smaller volumes (e.g., by mini-coring), while preserving spatial reference between (sub-)samples. However, results also reveal challenges and limitations with the segmentation of the XCT data and the data integration of AM and XCT, particularly for quantitative analysis, due to their different functionalities. In this study, no stereological error could be quantified as no proper grain separation of the segmented XCT data was performed. Yet, some well-separated grains exhibit a potential stereological effect. Overall, the integration of AM with XCT improves the output of both techniques and thereby ore characterization in general.

Ore characterization is crucial for efficient and profitable production of mineral products from an ore deposit. Analysis is typically performed at various scales (meter to microns) in a sequential fashion, where sample volume is reduced with increasing spatial resolution due to the increasing costs and run times of analysis. Thus, at higher resolution, sampling and data quality become increasingly important to represent the entire ore deposit. In particular, trace metal mineral characterization requires high-resolution analysis, due to the typical very fine grain sizes (sub-millimeter) of trace metal minerals. Automated Mineralogy (AM) is a key technique in the mining industry to quantify process-relevant mineral parameters in ore samples. Yet the limitation to two-dimensional analysis of flat sample surfaces constrains the sampling volume, introduces an undesired stereological error, and makes spatial interpretation of textures and structures difficult. X-ray computed tomography (XCT) allows three-dimensional imaging of rock samples based on the x-ray linear attenuation of the constituting minerals. Minerals are visually differentiated though not chemically classified. In this study, decimeter to millimeter large ore samples were analyzed at resolutions from 45 to 1 μm by AM and XCT to investigate the potential of multi-scale correlative analysis between the two techniques. Mineralization styles of Au, Bi-minerals, scheelite, and molybdenite were studied. Results show that AM can aid segmentation (mineralogical classification) of the XCT data, and vice versa, that XCT can guide (sub-)sampling (e.g., for heavy trace minerals) for AM analysis and provide three-dimensional context to the two-dimensional quantitative AM data. XCT is particularly strong for multi-scale analysis, increasingly higher resolution scans of progressively smaller volumes (e.g., by mini-coring), while preserving spatial reference between (sub-)samples. However, results also reveal challenges and limitations with the segmentation of the XCT data and the data integration of AM and XCT, particularly for quantitative analysis, due to their different functionalities. In this study, no stereological error could be quantified as no proper grain separation of the segmented XCT data was performed. Yet, some well-separated grains exhibit a potential stereological effect. Overall, the integration of AM with XCT improves the output of both techniques and thereby ore characterization in general.

Breast density is one of the most significant factors for cancer risk. In this study, we proposed a supervised deep learning approach for automated estimation of percentage density (PD) on digital mammography (DM). The deep convolutional neural network (DCNN) was trained to estimate a probability map of breast density (PMD). PD was calculated as the ratio of the dense area to the breast area based on the probability of each pixel belonging to dense region or fatty region at a decision threshold of 0.5. The DCNN estimate was compared to a feature-based statistical learning approach, in which gray level, texture and morphological features were extracted from each ROI and the least absolute shrinkage and selection operator (LASSO) was used to select and combine the useful features to generate the PMD. The reference PD of each image was provided by two experienced MQSA radiologists. With IRB approval, we retrospectively collected 347 DMs from patient files at our institution. The 10-fold cross-validation results showed a strong correlation r=0.96 between the DCNN estimation and interactive segmentation by radiologists while that of the feature-based statistical learning approach vs radiologists’ segmentation had a correlation r=0.78. The difference between the segmentation by DCNN and by radiologists was significantly smaller than that between the feature-based learning approach and radiologists (p < 0.0001) by two-tailed paired t-test. This study demonstrated that the DCNN approach has the potential to replace radiologists’ interactive thresholding in PD estimation on DMs.

Deep learning has emerged as a state-of-the-art learning technique across a wide range of applications, including image recognition, object detection and localisation, natural language processing, prediction and forecasting systems. With significant applicability, deep learning could be used in new and broader areas of applications, including remanufacturing. Remanufacturing is a process of taking used products through disassembly, inspection, cleaning, reconditioning, reassembly and testing to ascertain that their condition meets new products conditions with warranty. This process is complex and requires a good understanding of the respective stages for proper analysis. Inspection is a critical process in remanufacturing, which guarantees the quality of the remanufactured products. It is currently an expensive manual operation in the remanufacturing process that depends on operator expertise, in most cases. This research investigates the application of deep learning algorithms to inspection in remanufacturing, towards automating the inspection process. This paper presents a novel vision-based inspection system based on deep convolution neural network (DCNN) for eight types of defects, namely pitting, rust, cracks and other combination faults. The materials used for this feasibility study were 100 cm × 150 cm mild steel plate material, purchased locally, and captured using a USB webcam of 0.3 megapixels. The performance of this preliminary study indicates that the DCNN can classify with up to 100% accuracy on validation data and above 96% accuracy on a live video feed, by using 80% of the sample dataset for training and the remaining 20% for testing. Therefore, in the remanufacturing parts inspection, the DCNN approach has high potential as a method that could surpass the current technologies used in the design of inspection systems. This research is the first to apply deep learning techniques in remanufacturing inspection. The proposed method offers the potential to eliminate expert judgement in inspection, save cost, increase throughput and improve precision. This preliminary study demonstrates that deep learning techniques have the potential to revolutionise inspection in remanufacturing. This research offers valuable insight into these opportunities, serving as a starting point for future applications of deep learning algorithms to remanufacturing.

This paper presents a model of Deep Convolutional Neural Networks (CNN) based on transfer learning for image recognition. This means to use a Deep CNN system pretrained on the large ImageNet dataset of 14 million images and 1000 classes in order to learn feature selection. The results of the pretraining phase are transferred to the problem of classification for the images belonging to the UC Merced Land Use dataset with 21 classes. As benchmark, we have considered a Deep CNN trained with a fraction of the same UC Merced Land Use dataset containing the test images for classification. The experimental results have pointed out the obvious advantage of the Deep CNN with transfer learning (accuracy of 0.87 using pretraining over 0.46 for fully training on the same dataset).

Weathering of silicate-rich industrial wastes such as slag can reduce emissions from the steelmaking industry. During slag weathering, different minerals spontaneously react with atmospheric CO2 to produce calcite. Here, we evaluate the CO2 uptake during slag weathering using image-based analysis. The analysis was applied to an X-ray computed tomography (XCT) dataset of a slag sample associated with the former Ravenscraig steelworks in Lanarkshire, Scotland. The element distribution of the sample was studied using scanning electron microscopy (SEM), coupled with energy-dispersive spectroscopy (EDS). Two advanced image segmentation methods, namely trainable WEKA segmentation in the Fiji distribution of ImageJ and watershed segmentation in Avizo ® 9.3.0, were used to segment the XCT images into matrix, pore space, calcite, and other precipitates. Both methods yielded similar volume fractions of the segmented classes. However, WEKA segmentation performed better in segmenting smaller pores, while watershed segmentation was superior in overcoming the partial volume effect presented in the XCT data. We estimate that CO2 has been captured in the studied sample with an uptake between 20 and 17 kg CO2/1,000 kg slag for TWS and WS, respectively, through calcite precipitation.

Investigations of biphasic monodisperse soft (rubber) and stiff (glass) particle mixtures under hydrostatic conditions show an interesting behavior with regard to the effective stiffness. P‐wave modulus measured by acoustic wave propagation at ultrasonic frequencies showed a significant decline while more soft particles are added, that is, higher rubber volume fractions, due to a change in the microstructure of the granular medium. However, for small volume fractions of soft particles, it could be observed that the P‐wave modulus is increasing. This result cannot be explained by classical mixture rules or effective medium theories. For the understanding of those effects, a detailed insight into the microstructure of the granular medium is necessary. To gain this information and link it later back to the measured effective mechanical properties, high‐resolution micro X‐ray computed tomography (XRCT) imaging is a well‐established tool. With XRCT imaging, the granular microstructure can be visualized in 3D and characterized subsequently. Combining classical effective characterization methods with XRCT imaging can help to solve a variety of multi‐scale problems. Performing the characterization step in situ, meaning inside the laboratory‐based XRCT scanner, has the advantage that exactly the same samples are mechanically characterized and visualized. To address the mentioned observation above, we designed a low X‐ray absorbing oedometer cell with integrated broadband piezoelectric P‐wave transducers which enables this kind of investigation inside a laboratory‐based XRCT scanner. The focus of this contribution is on the general experimental methodology which can be transferred to other multi‐scale problems. It starts with a description of the image acquisition and ends with the post‐processing of the in situ acquired image data. To demonstrate this, cylindrical samples consisting of the same monodisperse rubber and glass particle mixtures that were studied before under hydrostatic stress conditions are considered. Selected results are presented to explain the single steps.

The relationship between the behavior of water in snow and its microstructure is crucial to improve the prediction of wet snow disasters. X-ray computed tomography (X-ray CT) is frequently used to observe snow microscopically. However, distinguishing between ice and water in the X-ray images is difficult because ice exhibits an X-ray absorption coefficient similar to that of water. In contrast, magnetic resonance imaging (MRI) acquires nuclear magnetic resonance (NMR) signals of protons in a liquid and visualizes the NMR signal intensity, enabling discrimination between water and ice signals. However, snow grains and pore spaces cannot be distinguished in MRI images because they do not generate NMR signals. To investigate the relationship between the microstructure of snow and the distribution of liquids in snow, we developed a novel method that combines X-ray CT and MRI images to compensate for the disadvantages associated with each method. Using this method, we successfully visualized where liquid (C12H24) occupied pore spaces. We also showed the possibility of using C12H24 instead of water to obtain water retention curve of snow cover, which is a fundamental aspect of hydraulic properties. Although there is room for improvement in the visualization of water in snow, such as shortening the imaging time to escape snow metamorphosis and image superimposition methods, this method is expected to effectively elucidate the behavior of water in snow and clarify the characteristics of wet snow.

A statistical description of X-ray CT (computerized tomography) imaging, from the projection data to the reconstructed image, is presented. The Gaussianity of the pixel image generated by the convolution (image reconstruction) algorithm is justified. The conditions for two pixel images to be statistically independent (for a given probability) and the conditions for a group of pixel images to be a spatial stationary random process and ergodic in mean and autocorrelations are derived. These properties provide the basis for establishing the stochastic image model and conducting the statistical image analysis of X-ray CT images.