Cementitious composites incorporating well-dispersed carbon nanotubes (CNTs) have demonstrated significant mechanical performance enhancements, however, there has only been limited investigation into ...the nanocomposite microstructure and pore structure. In this study, the effects of (i) CNT dispersion with and without the assistance of a dispersant, (ii) CNT dose at 0.05–0.25 wt% of cement, and (iii) CNT dispersion quality upon the composite microstructure after 7- and 28-days' hydration were investigated using quantitative image analysis of backscattered electron microscopy and X-ray computed microtomographic datasets. Results show that dispersed CNTs promoted the formation of low- and higher-density hydration products within 7 days of hydration and reduced the pore size distribution by 28 days, although 3D X-ray results showed some large CNT agglomerations formed in the nanocomposite. Further, poorly-dispersed CNTs increased the 28-day pore size distribution, but had a beneficial effect, similar to well-dispersed CNTs, upon the microstructural composition of the hydrated phases.
The American lobster, Homarus americanus, supports the most valuable single-species fishery in North America; however to-date, a reliable and robust method to determine age does not exist, and thus ...some of the more traditional catch-at-age stock assessment methods cannot be used to determine status. In lieu of this, the Atlantic States Marine Fisheries Commission’s American Lobster Stock Assessment Model uses a probabilistic growth transition matrix to determine how lobsters of different sizes will grow and recruit into the fishery. Developing and updating the growth transition matrix to reflect current growth dynamics requires estimates of molt increment and interval. Tagging studies are an inexpensive method to collect this type of information, but relying on fishing industry reported recapture data and buy-in can be challenging, resulting in varied return rates and uncertainty in measurement accuracy. Here, we report on a subset of recaptured lobsters released in the Gulf of Maine and Georges Bank regions between 2015 and 2020, where in addition to recapture locations, harvesters were encouraged to submit images of recaptured lobsters alongside a standard lobster gauge as a scale to estimate carapace length. Some fishermen were able to provide both direct measurements and images. Images were analyzed in ImageJ to estimate individual lobster carapace length (CL); and each image was assigned a quality score. For all groups, image-derived lengths were correlated with measured observations, regardless of the overall image quality, with only a slight underestimation at the maximum end of CL. This image-based method for length estimation provides high-quality length predictions regardless of image quality and can significantly increase the likelihood of harvesters contributing data with broad potential to engage commercial fisherman in collaborative science. Using this method we were able to expand the dataset of length records from this project by 251 or 33 %, providing additional data at a rate similar to tagging an additional 6000 lobsters.
The author proposes a method for simultaneous registration and segmentation of multi-source images, using the multivariate mixture model (MvMM) and maximum of log-likelihood (LL) framework. ...Specifically, the method is applied to the problem of myocardial segmentation combining the complementary information from multi-sequence (MS) cardiac magnetic resonance (CMR) images. For the image misalignment and incongruent data, the MvMM is formulated with transformations and is further generalized for dealing with the hetero-coverage multi-modality images (HC-MMIs). The segmentation of MvMM is performed in a virtual common space, to which all the images and misaligned slices are simultaneously registered. Furthermore, this common space can be divided into a number of sub-regions, each of which contains congruent data, thus the HC-MMIs can be modeled using a set of conventional MvMMs. Results show that MvMM obtained significantly better performance compared to the conventional approaches and demonstrated good potential for scar quantification as well as myocardial segmentation. The generalized MvMM has also demonstrated better robustness in the incongruent data, where some images may not fully cover the region of interest, and the full coverage can only be reconstructed combining the images from multiple sources.
Smartphone based devices (SBDs) have the potential to revolutionize food safety control by empowering citizens to perform screening tests. To achieve this, it is of paramount importance to understand ...current research efforts and identify key technology gaps. Therefore, a systematic review of optical SBDs in the food safety sector was performed. An overview of reviewed SBDs is given focusing on performance characteristics as well as image analysis procedures. The state-of-the-art on commercially available SBDs is also provided. This analysis revealed several important technology gaps, the most prominent of which are: (i) the need to reach a consensus regarding optimal image analysis, (ii) the need to assess the effect of measurement variation caused by using different smartphones and (iii) the need to standardize validation procedures to obtain robust data. Addressing these issues will drive the development of SBDs and potentially unlock their massive potential for citizen-based food control.
•Optical smartphone based sensors in the food safety field are systematically reviewed.•Recommendations on image analysis optimization are given.•The analytical performance of smartphone based sensors is discussed.•Available commercial devises are critically compared.
Artificial intelligence (AI) systems for computer-aided diagnosis and image-based screening are being adopted worldwide by medical institutions. In such a context, generating fair and unbiased ...classifiers becomes of paramount importance. The research community of medical image computing is making great efforts in developing more accurate algorithms to assist medical doctors in the difficult task of disease diagnosis. However, little attention is paid to the way databases are collected and how this may influence the performance of AI systems. Our study sheds light on the importance of gender balance in medical imaging datasets used to train AI systems for computer-assisted diagnosis. We provide empirical evidence supported by a large-scale study, based on three deep neural network architectures and two well-known publicly available X-ray image datasets used to diagnose various thoracic diseases under different gender imbalance conditions. We found a consistent decrease in performance for underrepresented genders when a minimum balance is not fulfilled. This raises the alarm for national agencies in charge of regulating and approving computer-assisted diagnosis systems, which should include explicit gender balance and diversity recommendations. We also establish an open problem for the academic medical image computing community which needs to be addressed by novel algorithms endowed with robustness to gender imbalance.
Analysis of the corrosion distribution and composition of corrosion products on steel surfaces using supervised machine learning of optical microscopic images was investigated. The accuracy of the ...artificial intelligence in evaluating the composition of iron compound reference samples was affected by the illumination intensity and surface roughness during image capture. The evaluation accuracy was high for compounds with a wide distribution of R value such as Fe2O3 and FeOOH, but low for compounds with a narrow distribution such as Fe3O4. The results of wet-dry cycling tests on weathering steel with NaCl particles on the surface showed that the transition of corrosion products during the corrosion progress can be analyzed from optical microscope images.
•We discuss different forms of supervision in medical image analysis.•Over 140 papers using semi-supervised, multi-instance or transfer learning are covered.•We discuss connections between these ...scenarios and further opportunities for research.
Display omitted
Machine learning (ML) algorithms have made a tremendous impact in the field of medical imaging. While medical imaging datasets have been growing in size, a challenge for supervised ML algorithms that is frequently mentioned is the lack of annotated data. As a result, various methods that can learn with less/other types of supervision, have been proposed. We give an overview of semi-supervised, multiple instance, and transfer learning in medical imaging, both in diagnosis or segmentation tasks. We also discuss connections between these learning scenarios, and opportunities for future research. A dataset with the details of the surveyed papers is available via https://figshare.com/articles/Database_of_surveyed_literature_in_Not-so-supervised_a_survey_of_semi-supervised_multi-instance_and_transfer_learning_in_medical_image_analysis_/7479416.
Recent advancements in wearable electronics offer seamless integration with the human body for extracting various biophysical and biochemical information for real-time health monitoring, clinical ...diagnostics, and augmented reality. Enormous efforts have been dedicated to imparting stretchability/flexibility and softness to electronic devices through materials science and structural modifications that enable stable and comfortable integration of these devices with the curvilinear and soft human body. However, the optical properties of these devices are still in the early stages of consideration. By incorporating transparency, visual information from interfacing biological systems can be preserved and utilized for comprehensive clinical diagnosis with image analysis techniques. Additionally, transparency provides optical imperceptibility, alleviating reluctance to wear the device on exposed skin. This review discusses the recent advancement of transparent wearable electronics in a comprehensive way that includes materials, processing, devices, and applications. Materials for transparent wearable electronics are discussed regarding their characteristics, synthesis, and engineering strategies for property enhancements. We also examine bridging techniques for stable integration with the soft human body. Building blocks for wearable electronic systems, including sensors, energy devices, actuators, and displays, are discussed with their mechanisms and performances. Lastly, we summarize the potential applications and conclude with the remaining challenges and prospects.
•Summarize deep learning methods for semantic segmentation of remote sensing images.•Identify three major challenges faced by researchers.•Summarize the innovative development to address them.
...Semantic segmentation of remote sensing imagery has been employed in many applications and is a key research topic for decades. With the success of deep learning methods in the field of computer vision, researchers have made a great effort to transfer their superior performance to the field of remote sensing image analysis. This paper starts with a summary of the fundamental deep neural network architectures and reviews the most recent developments of deep learning methods for semantic segmentation of remote sensing imagery including non-conventional data such as hyperspectral images and point clouds. In our review of the literature, we identified three major challenges faced by researchers and summarize the innovative development to address them. As tremendous efforts have been devoted to advancing pixel-level accuracy, the emerged deep learning methods demonstrated much-improved performance on several public data sets. As to handling the non-conventional, unstructured point cloud and rich spectral imagery, the performance of the state-of-the-art methods is, on average, inferior to that of the satellite imagery. Such a performance gap also exists in learning from small data sets. In particular, the limited non-conventional remote sensing data sets with labels is an obstacle to developing and evaluating new deep learning methods.