Combining Image Analysis Research Articles

Representativity of morphological measurements and 2-d shape descriptors on mineral admixtures

Using mineral admixtures has been proved to be significant for the production of cement-based composites. Burnt Clay Powder (BCP), Rice Husk Ash (RHA), Coconut Shell Ash (CSA) and Gum Arabic (GA) can be used as cement substitutes in cement-based composites whilst Cassava Starch (CS) can be used as an admixture in cement-based composites. Understanding shape parameters, microstructure and morphology of mineral admixtures and thus utilising them more efficiently for practical applications is critically required. Although research has been conducted to characterise mineral admixtures, limited studies have documented the combined Scanning Electron Microscopy (SEM) and image analysis. Thanks to image analysis-based methodology, it was possible to establish circularity (Circ), roundness (RD), solidity (SLD), axial ratio (AR), surface porosity (SP) and surface roughness (SR) of pozzolans and admixtures. This involved usage of SEM images of mineral admixtures to establish shape descriptors, porosity and surface roughness of mineral admixtures using image-based analysis and assessing the correlation and interdependence among the parameters. The mineral admixtures in this study were characterised by average values of solidity, roundness, circularity and axial ratio of greater than 0.7. Evaluation of correlation between roundness and circularity illustrated weak R2 values ranging from 0.30 to 0.40 for all mineral admixtures. The axial ratio also weakly correlated with solidity with a wider range of R2 values from 0.27 to 0.45. Importantly, the relationships were noticed to be significant in the interpretation of morphology and shape data.

Effects of straw returning to movement characteristics of soil preferential flow in sugarcane fields

The measures of returning straw to the field can change soil environment in the field and affect crop growth. In order to explore the impacts of different straw returning measures on the preferential flow in the field, we investigated the movement characteristics of preferential flow in sugarcane fields under different straw returning measures [straw mulching (CM), straw burning mulching (BM), and non-straw mulching (CK)] in Guangxi by combining image analysis techniques with ecological landscape pattern analysis methods. The results showed that under the same external water supply environment, soil staining morphology of non-straw mulching sugarcane fields differentiated into finger-like and clump-like. Both straw mulching and straw burning mulching were mainly differentiated into clump-like. The mean total staining area ratio was significantly higher than that of non-straw mulching (26.0%). The ratio of surface staining area decreased rapidly under non-straw mulching. The water movement range decreased rapidly under straw mulching and straw burning mulching in the depth of 15-30 cm. The change rate of straw mulching staining area ratio was the lowest in the deep soil. The shape index of non-straw mulching was 15.54. The water flow movement was more concentrated. The average sub-circular index of straw mulching and straw burning mulching was 0.67 in the sugarcane field, with a higher degree of preferential flow bending, an average proximity index of 0.87, and better water flow connectivity. The average substrate flow depth (3.52 cm) and edge density index (11.51) were the smallest under non-straw mulching, but the priority flow ratio was the largest (73.2%), and the spatial development degree of preferential flow was the highest. The straw returning measures reduced the preferential flow in the sugarcane fields, with positive effects on water conservation.

Automating fractional flow reserve (FFR) calculation from CT scans: A rapid workflow using unsupervised learning and computational fluid dynamics.

Fractional flow reserve (FFR) provides the functional relevance of coronary atheroma. The FFR-guided strategy has been shown to reduce unnecessary stenting, improve overall health outcome, and to be cost-saving. The non-invasive, coronary computerised tomography (CT) angiography-derived FFR (cFFR) is an emerging method in reducing invasive catheter based measurements. This computational fluid dynamics-based method is laborious as it requires expertise in multidisciplinary analysis of combining image analysis and computational mechanics. In this work, we present a rapid method, powered by unsupervised learning, to automatically calculate cFFR from CT scans without manual intervention.

Assessing the Physiologic Relevance of Red Blood Cell Deformability in Iron Deficiency Anemia

Background: Iron deficiency anemia (IDA), which affects individuals of all ages worldwide, is an often overlooked and undertreated component of chronic disease, despite data correlating its association with adverse outcomes in patients with cardiovascular disease (von Haehling, Nat Rev Cardiol, 2015). While red blood cells (RBCs) in IDA are known to be smaller and contain less hemoglobin than healthy RBCs, how RBC deformability is altered in IDA remains poorly understood; some ektacytometry studies have observed impaired deformability in iron deficient RBCs (idRBCs), while others described either unchanged or increased deformability (Brandão, Clin Hemorheol Microcirc, 2009). Here we ask: can single cell biophysical techniques definitively determine whether idRBCs are less deformable than healthy RBCs and how heterogenous that phenomena may be? Recent investigations into IDA's role in cardiovascular disease have generally focused on the myocardium and coronary vasculature, yet much regarding other physiologic implications remains unknown, including whether idRBCs cause microvascular obstruction or vasculopathy. To address such questions, we leveraged a suite of microvascular models we developed.Methods: We first coupled our microfluidic capillary model with μEXACT, our customized automated particle tracking program for hematologic cell-based assays, to collect high-throughput velocity tracking of single RBCs from a healthy control and 2 IDA patients (anemic for age, ferritin <10 ng/mL) to create a single cell deformability index (sDI) for each RBC (Fig 1). Next, whole blood samples collected in EDTA tubes from the control and IDA patients were perfused into both straight 100μm wide channels (mimicking large venules) and branching 30μm wide microfluidic devices (mimicking smaller venules) at a constant shear rate for 30 minutes to observe if any occlusions or obvious alterations in flow were observed (Fig 2). Finally, using the straight 100μm channel microfluidic devices, human umbilical vein endothelial cells (HUVECs) were cultured throughout each microchannel and RBCs from a healthy control and 3 IDA patients were perfused in parallel microchannels for 4 hours. The endothelialized models were then fixed, permeabilized, and immunostained with antibodies against VCAM-1 and E-selectin, known markers of endothelial inflammation. Mean fluorescence intensity was measured to quantify endothelial inflammation (Fig 3).Results: sDI distribution histograms were obtained for healthy and IDA patient RBCs. The mean sDIs for IDA patient RBCs were decreased in comparison to the healthy RBCs. Additionally, both IDA patient's RBCs lacked a subpopulation of highly deformable RBCs, likely reticulocytes, seen in the healthy RBCs (Fig 1C). There was no evidence of microchannel occlusion for the healthy control or IDA patient whole blood samples in either the straight 100μm microchannels or branching 30μm microfluidic devices (Fig 2D). Finally, in our endothelialized microfluidic model, endothelium exposed to IDA patient RBCs exhibited increased VCAM-1 and E-selectin expression over endothelium exposed to healthy RBCs (Fig 3B).Conclusions: By utilizing an array of microfluidic models we can develop a more comprehensive understanding of the role idRBCs play systemically on microvasculature. Our combined microfluidic and image analysis system demonstrated decreased deformability in idRBCs and can offer detection of subpopulation differences that cannot be fully characterized with bulk techniques such as ektacytometry. So far, our data demonstrates that while no microvascular occlusion occurs, idRBCs contribute to endothelial inflammation. Additionally, the observation that physical interactions between endothelial cells and idRBCs are sufficient to cause endothelial inflammation warrants further investigation, as generally idRBCs had not been viewed as pro-inflammatory. Ongoing studies will couple unique sDI distribution curves with the degree of endothelial inflammation, as well as elucidate how these changes are associated with the degree of IDA or clinical events such as the initiation of iron supplementation. Utilizing atomic force microscopy to better understand how the idRBC membrane impacts deformability and developing biophysical computer simulations to determine if increased idRBC-endothelium interactions are observed in silico are also planned. [Display omitted] DisclosuresLam: Sanguina, Inc.: Current holder of individual stocks in a privately-held company.

Characterization of an APP/tau rat model of Alzheimer\u2019s disease by positron emission tomography and immunofluorescent labeling

BackgroundTo better understand the etiology and pathomechanisms of Alzheimer’s disease, several transgenic animal models that overexpress human tau or human amyloid-beta (Aβ) have been developed. In the present study, we generated a novel transgenic rat model by cross-breeding amyloid precursor protein (APP) rats with tau rats. We characterized this model by performing positron emission tomography scans combined with immunofluorescent labeling and cerebrospinal fluid analyses.MethodsAPP/Tau rats were generated by cross-breeding male McGill-R-Thy1-APP transgenic rats with female hTau-40/P301L transgenic rats. APP/Tau double transgenic rats and non-transgenic (ntg) littermates aged 7, 13, and 21 months were subjected to dynamic [11C] PiB scan and dynamic [18F]THK-5317 scans. For regional brain analysis, a template was generated from anatomical MR images of selected animals, which was co-registered with the PET images. Regional analysis was performed by application of the simplified reference tissue model ([11C]PiB data), whereas [18F]THK-5317 data were analyzed using a 2-tissue compartment model and Logan graphical analysis. In addition, immunofluorescent labeling (tau, amyloid) and cerebrospinal fluid analyses were performed.Results[11C]PiB binding potential (BPND) and [18F]THK-5317 volume of distribution (VT) showed an increase with age in several brain regions in the APP/Tau group but not in the ntg control group. Immunohistochemical analysis of brain slices of PET-scanned animals revealed a positive correlation between Aβ labeling and [11C]PiB regional BPND. Tau staining yielded a trend towards higher levels in the cortex and hippocampus of APP/Tau rats compared with ntg littermates, but without reaching statistical significance. No correlation was found between tau immunofluorescence labeling results and the respective [18F]THK-5317 VT values.ConclusionsWe thoroughly characterized a novel APP/Tau rat model using combined PET imaging and immunofluorescence analysis. We observed an age-related increase in [11C]PiB and [18F]THK-5317 binding in several brain regions in the APP/Tau group but not in the ntg group. Although we were able to reveal a positive correlation between amyloid labeling and [11C]PiB regional brain uptake, we observed relatively low human tau and amyloid fibril expression levels and a somewhat unstable brain pathology which questions the utility of this animal model for further studies.

Early Estimation of Olive Production from Light Drone Orthophoto, through Canopy Radius

Background: The present work aims at obtaining an approximate early production estimate of olive orchards used for extra virgin olive oil production by combining image analysis techniques with light drone images acquisition and photogrammetric reconstruction. Methods: In May 2019, an orthophoto was reconstructed through a flight over an olive grove to predict oil production from segmentation of plant canopy surfaces. The orchard was divided into four plots (three considered as training plots and one considered as a test plot). For each olive tree of the considered plot, the leaf surface was assessed by segmenting the orthophoto and counting the pixels belonging to the canopy. At harvesting, the olive production per plant was measured. The canopy radius of the plant (R) was automatically obtained from the pixel classification and the measured production was plotted as a function of R. Results: After applying a k-means-classification to the four plots, two distinct subsets emerged in association with the year of loading (high-production) and unloading. For each plot of the training set the logarithm of the production curves against R were fitted with a linear function considering only four samples (two samples belonging to the loading region and two samples belonging to the unloading one) and the total production estimate was obtained by integrating the exponent of the fitting-curve over R. The three fitting curves obtained were used to estimate the total production of the test plot. The resulting estimate of the total production deviates from the real one by less than 12% in training and less than 18% in tests. Conclusions: The early estimation of the total production based on R extracted by the orthophotos can allow the design of an anti-fraud protocol on the declared production.

Clinicopathologic and Transcriptomic Analysis of Radiation-Induced Lung Injury in Nonhuman Primates.

Radiation-induced lung injury (RILI) is a progressive condition with an early phase (radiation pneumonitis) and a late phase (lung fibrosis). RILI may occur after partial-body ionizing radiation exposures or internal radioisotope exposure, with wide individual variability in timing and extent of lung injury. This study aimed to provide new insights into the pathogenesis and progression of RILI in the nonhuman primate (NHP) rhesus macaque model. We used an integrative approach to understand RILI and its evolution at clinical and molecular levels in 17 NHPs exposed to 10 Gy of whole-thorax irradiation in comparison with 3 sham-irradiated control NHPs. Clinically, we monitored respiratory rates, computed tomography (CT) scans, plasma cytokine levels, and bronchoalveolar lavage (BAL) over 8 months and lung samples collected at necropsy for molecular and histopathologic analyses using RNA sequencing and immunohistochemistry. Elevated respiratory rates, greater CT density, and more severe pneumonitis with increased macrophage content were associated with early mortality. Radiation-induced lung fibrosis included polarization of macrophages toward the M2-like phenotype, TGF-β signaling, expression of CDKN1A/p21 in epithelial cells, and expression of α-SMA in lung stroma. RNA sequencing analysis of lung tissue revealed SERPINA3, ATP12A, GJB2, CLDN10, TOX3, and LPA as top dysregulated transcripts in irradiated animals. In addition to transcriptomic data, we observed increased protein expression of SERPINA3, TGF-β1, CCL2, and CCL11 in BAL and plasma samples. Our combined clinical, imaging, histologic, and transcriptomic analysis provides new insights into the early and late phases of RILI and highlights possible biomarkers and potential therapeutic targets of RILI. Activation of TGF-β and macrophage polarization appear to be key mechanisms involved in RILI.

Comparison of Early Imaging and Imaging 60 min Post-Injection after Forced Diuresis with Furosemide in the Assessment of Local Recurrence in Prostate Cancer Patients with Biochemical Recurrence Referred for 68Ga-PSMA-11 PET/CT.

Background: 68Ga-PSMA-11 PET/CT is a promising method for the assessment of local recurrence (LR) in prostate cancer (PCa) patients. The aim of this study was to evaluate the diagnostic performance of early 68Ga-PSMA-11 PET imaging in comparison to 68Ga-PSMA-11 PET imaging 60 min post-injection (p.i.) in the detection of LR in patients with biochemical recurrence (BR) of prostate carcinoma. Materials and Methods: 190 image sets of patients with BR in PCa who underwent 68Ga-PSMA-11 PET/CT were assessed retrospectively (median prostate specific antigen (PSA) value, 0.70 ng/mL (range, 0.1–105.6 ng/mL)). Patients received an early static scan of the pelvic area (median, 248 s p.i. (range, 56–923 s)) and a whole-body scan 60 min p.i. (median, 64 min p.i. (range, 45–100 min)) with intravenous administration of 20 mg furosemide i.v. at the time of tracer application, followed by intravenous hydration with 500 mL of sodium chloride (NaCl 0.9%). Assessment was based on visual analysis and calculation of the maximum standardized uptake value (SUVmax) of the pathologic lesions present in the prostate fossa found in the early PET imaging and 60 min PET scans. The scans were characterized as negative, positive, or equivocal. The results were compared, and the combination of early and 60 min p.i. imaging was evaluated. Results: Image assessment resulted in 30 (15.8%) positive, 17 (8.9%) equivocal, and 143 (75.3%) negative findings in early scans, and 28 (14.7%) positive, 25 (13.2%) equivocal, and 137 (72.1%) negative findings of LR in 60 min p.i. images. For combined image analysis, 33 (17.4%) cases were positive and 20 (10.5%) were equivocal. There was no statistical significance between the number of positive (p = 0.815), negative (p = 0.327), and equivocal (p = 0.152) findings. Furthermore, the combination of both scans showed no statistically significant differences for the positive and negative findings (p = 0.063). The median SUVmax was 4.9 (range, 2.0–55.2) for positive lesions in the early scans and 8.0 (range, 2.1–139.9) in the scans 60 min p.i. The median SUVmax for bladder activity was 2.5 (range, 0.9–12.2) in the early scans and 8.2 (range, 1.8–27.6) in the scans 60 min p.i. Conclusion: Early static imaging additional to 68Ga-PSMA-11 PET images acquired 60 min p.i. has limited value in patients prepared with furosemide and hydration, and showed no statistically significant change in the detection rate (DR) of LR and the number of equivocal findings. Based on our results, in departments following a protocol with forced diuresis, including furosemide, additional early static imaging cannot be routinely recommended for the assessment of BR in PCa patients.

The value of dynamic contrast-enhanced magnetic resonance imaging combined with apparent diffusion coefficient in the differentiation of benign and malignant diseases of the breast.

The value of combined dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and apparent diffusion coefficient (ADC) histogram analysis for the diagnosis of breast cancer has not been evaluated in previous studies. To investigate the diagnostic value of DCE-MRI combined with ADC in benign and malignant breast lesions. The clinicopathological imaging data included 168 patients (177 lesions) with breast lesions who underwent convention breast MRI, DCE-MRI, and diffusion-weighted imaging (DWI); they were divided into the benign lesion group (n = 39) and malignant lesion group (n = 129) based on pathology. Using the type III outflow curve as a diagnostic criterion for malignant breast lesions, the diagnostic sensitivity was 76.9%, the specificity was 80%, the correct rate was 72.2%, and its area under the curve (AUC) was 0.823. Using an enhancement ratio > 100% as a diagnostic criterion for malignant breast lesions, the sensitivity was 61.5%, specificity was 80%, and AUC was 0.723. Using > 3 ipsilateral vessels as a diagnostic criterion for malignant lesions in the breast resulted in a diagnostic sensitivity of 81.6%, a specificity of 80.8%, and an AUC of 0.805. The type of time intensity curve DCE-MRI, the early enhancement rate in the first phase, the number of ipsilateral vessels, and the ADC full volume histogram of the blood supply score and DWI are valuable in the diagnosis of benign and malignant breast lesions.

Quality Control in Fiore Sardo PDO Cheese: Detection of Heat Treatment Application and Production Chain by MRI Relaxometry and Image Analysis

Fiore Sardo (FS), a traditional Italian cheese, is present in the market as a heterogeneous variety of products. The use of heat-treated (HT) milk is forbidden by the official production protocol, but no official analytical method able to detect heat application is yet available. Here, a combined magnetic resonance imaging (MRI) relaxometry and image analysis approach to recognize FS made from raw milk is presented. Artisanal FS cheeses were produced from raw milk (RC) by five shepherds in accordance with the official protocol. They were compared to HT-milk counterparts (HTC). Additionally, industrially manufactured commercial FS cheeses (I) were also purchased and compared to RC and HTC. Relaxometry data of FS indicated the presence of two water populations; the ratio of characteristic relaxation time constant T2 and area fraction (Score, Ṩ) of the fastest relaxing population was used to compare RC, HTC and I samples. RC from HTC were successfully discriminated, the latter exhibiting lower Ṩ (enhanced protein hydration). I cheeses exhibited the lowest Ṩ values, sometimes comparable to HTC. Since visual appearance of RC and HTC is appreciably different, an image analysis deep learning approach using MRI and photographic pictures was adopted to discriminate the two productions, with promising percentages (&gt;93%).

Detection of Potential Vernal Pools on the Canadian Shield (Ontario) Using Object-Based Image Analysis in Combination with Machine Learning

Vernal pools are small, temporary, forested wetlands of ecological importance with a high sensitivity to changing climate and land-use patterns. These ecosystems are under considerable development pressure in southeastern Georgian Bay, where mapping techniques are required to inform wise land-use decisions. Our mapping approach combines common machine learning techniques [random forest, support vector machines (SVMs)] with object-based image analysis. Using multispectral image segmentation on high-resolution orthoimagery, we first created objects and assigned classes based on field collected data. We then supplied machine learning algorithms with data from freely available sources (Ontario orthoimagery and Sentinel 2) and tested accuracy on a reserved dataset. We achieved producer’s accuracies of 85 and 79% and user’s accuracies of 78 and 84% for random forest and SVMs models, respectively. Difficulty differentiating between small, dark shadows and small, obscured pools accounted for many of the omission and commission errors. Our automated approach of vernal pool classification provides a relatively accurate, consistent, and fast mapping strategy compared to manual photointerpretation. Our models can be applied on a regional basis to help verify the locations of pools in an area of Ontario that is in critical need of more detailed ecological information.

Microscale intrinsic properties of hybrid unidirectional/woven composite laminates: Part I experimental tests

Understanding of the failure of composites is critical for evaluating their safety and reliability when they are in service. In this research, a combined image analysis technique using computed microtomography (micro-CT) and scanning electron microscopy (SEM) is proposed to study the microstructures, damage and failure of a hybrid unidirectional/woven composite laminate (HUWCL). Micro-CT is used to extract the 3D microscale morphology and identify the internal failure behaviours of the HUWCL. A series of material parameters, such as fiber volume fraction, void ratio and layer mode, are acquired by the statistical analysis of the micro-CT images. In addition, microscale surface damage and failure mechanism of woven and unidirectional (UD) laminae are further investigated by SEM. It is found that the matrix damage around the interface always occurs in both woven and unidirectional (UD) laminae, and distinct delamination is found between the UD laminae and the woven composites. The study shows a great potential of combining micro-CT with SEM in revealing the local damage and intrinsic failure mechanism in HUWCLs at microscale.

Prediction of olive ripening degree combining image analysis and FT-NIR spectroscopy for virgin olive oil optimisation

This work proposes classification models for the prediction of olive maturity index based on Fourier Transform-Near Infrared (FT-NIR) spectra of intact drupes. An image analysis (IA) method was purposely developed for the objective evaluation of the maturity index. Thirteen cultivars at different ripening stages were harvested along three years. The reliability of the IA method was confirmed by the highly significant correlation with the common visual evaluation of maturity index. Classification models were developed with Partial Least Square-Discriminant Analysis (PLS-DA), using IA results and FT-NIR spectra of olives collected in diffuse reflectance. Most PLS-DA models calculated separately for olive origin gave sensitivity and specificity values in prediction higher than 81%. The global model performed slightly worse (sensitivity, 79%; specificity, 75%), but it is definitely more robust and can provide the olive sector with a fast, green and non-destructive olive sorting method for the production of high quality virgin oil.

Simultaneous automated image analysis and Raman spectroscopy of powders at an individual particle level

Solid form diversity of raw materials can be critical for the performance of the final drug product. In this study, Raman spectroscopy, image analysis and combined Raman and image analysis were utilized to characterize the solid form composition of a particulate raw material. Raman spectroscopy provides chemical information and is complementary to the physical information provided by image analysis. To demonstrate this approach, binary mixtures of two solid forms of carbamazepine with a distinct shape, an anhydrate (prism shaped) and a dihydrate (needle shaped), were characterized at an individual particle level. Partial least squares discriminant analysis classification models were developed and tested with known, gravimetrically mixed test samples, followed by analysis of unknown, commercially supplied carbamazepine raw material samples. Classification of several thousands of particles was performed, and it was observed that with the known binary mixtures, the minimum number of particles needed for the combined Raman spectroscopy – image analysis classification model was approximately 100 particles per solid form. The carbamazepine anhydrate and dihydrate particles were detected and classified with a classification error of 1 % using the combined model. Further, this approach allowed the identification of raw material solid form impurity in unknown raw material samples. Simultaneous automated image analysis and Raman spectroscopy of powders at an individual particle level has its potential in accurate detection of low amounts of unwanted solid forms in particulate raw material samples.

Analysis Technology of Tennis Sports Match Based on Data Mining and Image Feature Retrieval

Tennis game technical analysis is affected by factors such as complex background and on-site noise, which will lead to certain deviations in the results, and it is difficult to obtain scientific and effective tennis technical training strategies through a few game videos. In order to improve the performance of tennis game technical analysis, based on machine learning algorithms, this paper combines image analysis to identify athletes’ movement characteristics and image feature recognition processing with image recognition technology, realizes real-time tracking of athletes’ dynamic characteristics, and records technical characteristics. Moreover, this paper combines data mining technology to obtain effective data from massive video and image data, uses mathematical statistics and data mining technology for data processing, and scientifically analyzes tennis game technology with the support of ergonomics. In addition, this paper designs a controlled experiment to verify the technical analysis effect of the tennis match and the performance of the model itself. The research results show that the model constructed in this paper has certain practical effects and can be applied to actual competitions.

In-field hyperspectral imaging: An overview on the ground-based applications in agriculture

The measurement of vegetation indexes that characterise the plants growth, assessing the fruit ripeness or detecting the presence of defects and diseases, is a key factor to gain high quality of fruit or vegetables. Such evaluation can be carried out using both destructive and non destructive techniques. Among non-destructive techniques, hyperspectral imaging (HSI), combining image analysis and visible/near-infrared spectroscopy, looks particularly useful. Many studies have been published concerning the use of hyperspectral cameras in the agronomic and food field, especially in controlled laboratory conditions. Conversely, few studies described the application of HSI technology directly in field, especially involving ground-based systems. Results suggest that this technique could be particularly useful, even if the role of environmental variables has to be considered (e.g., intensity and incidence of solar radiation, wind or the soil in the background). In this paper, recent in-field HSI applications based on ground systems are reviewed.

Direct Atomic‐Level Imaging of Zeolites: Oxygen, Sodium in Na‐LTA and Iron in Fe‐MFI

AbstractZeolites are becoming more versatile in their chemical functions through rational design of their frameworks. Therefore, direct imaging of all atoms at the atomic scale, basic units (Si, Al, and O), heteroatoms in the framework, and extra‐framework cations, is needed. TEM provides local information at the atomic level, but the serious problem of electron‐beam damage needs to be overcome. Herein, all framework atoms, including oxygen and most of the extra‐framework Na cations, are successfully observed in one of the most electron‐beam‐sensitive and lowest framework density zeolites, Na‐LTA. Zeolite performance, for instance in catalysis, is highly dependent on the location of incorporated heteroatoms. Fe single atomic sites in the MFI framework have been imaged for the first time. The approach presented here, combining image analysis, electron diffraction, and DFT calculations, can provide essential structural keys for tuning catalytically active sites at the atomic level.

Direct Atomic-Level Imaging of Zeolites: Oxygen, Sodium in Na-LTA and Iron in Fe-MFI.

Zeolites are becoming more versatile in their chemical functions through rational design of their frameworks. Therefore, direct imaging of all atoms at the atomic scale, basic units (Si, Al, and O), heteroatoms in the framework, and extra‐framework cations, is needed. TEM provides local information at the atomic level, but the serious problem of electron‐beam damage needs to be overcome. Herein, all framework atoms, including oxygen and most of the extra‐framework Na cations, are successfully observed in one of the most electron‐beam‐sensitive and lowest framework density zeolites, Na‐LTA. Zeolite performance, for instance in catalysis, is highly dependent on the location of incorporated heteroatoms. Fe single atomic sites in the MFI framework have been imaged for the first time. The approach presented here, combining image analysis, electron diffraction, and DFT calculations, can provide essential structural keys for tuning catalytically active sites at the atomic level.

Habitat maps to enhance monitoring and management of the Great Barrier Reef

The Great Barrier Reef (GBR) is of immense biological, cultural and economic importance, but has also rapidly degraded over the last 30 years. Improved spatial information on reef geomorphic zonation and benthic cover type (including coral type) is critical to support scientific work to understand how the GBR is changing, and to support resource management decisions that enable conservation of the reef and its essential ecosystem services. Yet, no comprehensive maps exist that detail the geomorphic zonation or benthic cover for the GBR’s ~ 3000 reefs. This study presents three new types of shallow reef maps for 237 reefs in the central Cairns Management Region of the GBR Marine Park (GBRMP), explores how the detailed habitat maps created compared to current maps and posits how the new maps may support and refine current critical key science outputs and management challenges. Geomorphic Zonation, Benthic Cover and Coral Type habitat maps were created using a unique combined object-based image analysis and ecological modelling approach that incorporated satellite imagery, limited field data and key reef physical attributes (depth, slope, waves) using a previously peer-reviewed mapping approach developed for the Capricorn Bunker Group reefs, Southern GBR. The mapping approach was consistent and repeatable, suggesting applicability to mapping all 3000 reefs in the GBRMP. Compared to existing maps that only outline each reef, the increase in detail provided by these new habitat maps enabled discrete characterisation of each reef’s geomorphology and benthic composition. With the new habitat maps, areas within each reef can be identified as either coral habitat or not coral habitat. This has not been possible previously. As such, a model of coral ecological and biophysical processes that depends on bottom reflectance of sand and coral areas can be fine-tuned. Similarly, for reef restoration, nursery structures are commonly placed in non-coral habitats, and/or coral larvae are dispersed in areas of known coral habitat. The new habitat maps presented permit more accurate identification of these areas such that restoration projects can be targeted more effectively. These findings confirm the need to now apply this mapping approach to the full extent of the GBR.

SpikeSegNet-a deep learning approach utilizing encoder-decoder network with hourglass for spike segmentation and counting in wheat plant from visual imaging

BackgroundHigh throughput non-destructive phenotyping is emerging as a significant approach for phenotyping germplasm and breeding populations for the identification of superior donors, elite lines, and QTLs. Detection and counting of spikes, the grain bearing organs of wheat, is critical for phenomics of a large set of germplasm and breeding lines in controlled and field conditions. It is also required for precision agriculture where the application of nitrogen, water, and other inputs at this critical stage is necessary. Further, counting of spikes is an important measure to determine yield. Digital image analysis and machine learning techniques play an essential role in non-destructive plant phenotyping analysis.ResultsIn this study, an approach based on computer vision, particularly object detection, to recognize and count the number of spikes of the wheat plant from the digital images is proposed. For spike identification, a novel deep-learning network, SpikeSegNet, has been developed by combining two proposed feature networks: Local Patch extraction Network (LPNet) and Global Mask refinement Network (GMRNet). In LPNet, the contextual and spatial features are learned at the local patch level. The output of LPNet is a segmented mask image, which is further refined at the global level using GMRNet. Visual (RGB) images of 200 wheat plants were captured using LemnaTec imaging system installed at Nanaji Deshmukh Plant Phenomics Centre, ICAR-IARI, New Delhi. The precision, accuracy, and robustness (F1 score) of the proposed approach for spike segmentation are found to be 99.93%, 99.91%, and 99.91%, respectively. For counting the number of spikes, “analyse particles”—function of imageJ was applied on the output image of the proposed SpikeSegNet model. For spike counting, the average precision, accuracy, and robustness are 99%, 95%, and 97%, respectively. SpikeSegNet approach is tested for robustness with illuminated image dataset, and no significant difference is observed in the segmentation performance.ConclusionIn this study, a new approach called as SpikeSegNet has been proposed based on combined digital image analysis and deep learning techniques. A dedicated deep learning approach has been developed to identify and count spikes in the wheat plants. The performance of the approach demonstrates that SpikeSegNet is an effective and robust approach for spike detection and counting. As detection and counting of wheat spikes are closely related to the crop yield, and the proposed approach is also non-destructive, it is a significant step forward in the area of non-destructive and high-throughput phenotyping of wheat.