Crop Agnostic Monitoring Driven by Deep Learning

Farmers require diverse and complex information to make agronomical decisions about crop management including intervention tasks. Generally, this information is gathered by farmers traversing their fields or glasshouses which is often a time consuming and potentially expensive process. In recent years, robotic platforms have gained significant traction due to advances in artificial intelligence. However, these platforms are usually tied to one setting (such as arable farmland), or algorithms are designed for a single platform. This creates a significant gap between available technology and farmer requirements. We propose a novel field agnostic monitoring technique that is able to operate on two different robots, in arable farmland or a glasshouse (horticultural setting). Instance segmentation forms the backbone of this approach from which object location and class, object area, and yield information can be obtained. In arable farmland, our segmentation network is able to estimate crop and weed at a species level and in a glasshouse we are able to estimate the sweet pepper and their ripeness. For yield information, we introduce a novel matching criterion that removes the pixel-wise constraints of previous versions. This approach is able to accurately estimate the number of fruit (sweet pepper) in a glasshouse with a normalized absolute error of 4.7% and an R2 of 0.901 with the visual ground truth. When applied to cluttered arable farmland scenes it improves on the prior approach by 50%. Finally, a qualitative analysis shows the validity of this agnostic monitoring algorithm by supplying decision enabling information to the farmer such as the impact of a low level weeding intervention scheme.

The Effect of surfactant on the drag and wall correction factor of a drop in a bounded medium

In the present article, the analytical solution for creeping motion of a drop/bubble characterized by insoluble surfactant is examined at the instant it passes the center of a spherical container filled with Newtonian fluid at low Reynolds number. The presence of surfactant characterizes the interfacial region by an axisymmetric interfacial tension gradient and coefficient of surface dilatational viscosity. Under the assumption of the small capillary number, the deformation of spherical phase interface is not taken into account. The computations not only yield information on drag force and wall correction factor, but also on interfacial velocity and flow field for different values of surface tension gradient and surface dilatational viscosity. In the limiting cases, the analytical solutions describing the drag force and wall correction factor for a drop in a bounded medium reduces to expressions previously stated by other authors in literature. The results reveal the strong influence of the surface dilatational viscosity and surface tension gradient on the motion of drop/bubble. Increasing the surface tension gradient and surface dilatational viscosity, results in linear variation of drag force. When the surface tension gradient increases, the drag force for unbounded medium increases more as compared to the bounded medium hence wall correction factor decreases with increase in surface tension gradient whereas it increases with increase in surface dilatational viscosity.

Decoding accelerometry for classification and prediction of critically ill patients with severe brain injury

Our goal is to explore quantitative motor features in critically ill patients with severe brain injury (SBI). We hypothesized that computational decoding of these features would yield information on underlying neurological states and outcomes. Using wearable microsensors placed on all extremities, we recorded a median 24.1 (IQR: 22.8–25.1) hours of high-frequency accelerometry data per patient from a prospective cohort (n = 69) admitted to the ICU with SBI. Models were trained using time-, frequency-, and wavelet-domain features and levels of responsiveness and outcome as labels. The two primary tasks were detection of levels of responsiveness, assessed by motor sub-score of the Glasgow Coma Scale (GCSm), and prediction of functional outcome at discharge, measured with the Glasgow Outcome Scale–Extended (GOSE). Detection models achieved significant (AUC: 0.70 [95% CI: 0.53–0.85]) and consistent (observation windows: 12 min–9 h) discrimination of SBI patients capable of purposeful movement (GCSm > 4). Prediction models accurately discriminated patients of upper moderate disability or better (GOSE > 5) with 2–6 h of observation (AUC: 0.82 [95% CI: 0.75–0.90]). Results suggest that time series analysis of motor activity yields clinically relevant insights on underlying functional states and short-term outcomes in patients with SBI.

Stellar Surface Inhomogeneities as a Potential Source of the Atmospheric Signal Detected in the K2-18b Transmission Spectrum

Transmission spectroscopy of transiting exoplanets is a proven technique that can yield information on the composition and structure of a planet’s atmosphere. However, transmission spectra may be compromised by inhomogeneities in the stellar photosphere. The sub-Neptune-sized habitable zone planet K2-18b has water absorption detected in its atmosphere using data from the Hubble Space Telescope (HST). Herein, we examine whether the reported planetary atmospheric signal seen from HST transmission spectroscopy of K2-18b could instead be induced by time-varying starspots. We built a time-variable spectral model of K2-18 that is designed to match the variability amplitude seen in K2 photometric data, and we used this model to simulate 1000 HST data sets that follow the K2-18b observation strategy. More than 1% of these provide a better fit to the data than the best-fitting exoplanet atmosphere model. After resampling our simulations to generate synthetic HST observations, we find that 40% of random draws would produce an atmospheric detection at a level at least as significant as that seen in the actual HST data of K2-18b. This work illustrates that the inferred detection of an atmosphere on K2-18b may alternatively be explained by stellar spectral contamination due to the inhomogeneous photosphere of K2-18. We do not rule out a detection of water in the planet’s atmosphere, but we provide a plausible alternative that should be considered and conclude that more observations are needed to fully rule out stellar contamination.

Raman and Photoluminescence Spectroscopy with a Variable Spectral Resolution

Raman and photoluminescence (PL) spectroscopy are important analytic tools in materials science that yield information on molecules’ and crystals’ vibrational and electronic properties. Here, we show results of a novel approach for Raman and PL spectroscopy to exploit variable spectral resolution by using zoom optics in a monochromator in the front of the detector. Our results show that the spectral intervals of interest can be recorded with different zoom factors, significantly reducing the acquisition time and changing the spectral resolution for different zoom factors. The smallest spectral intervals recorded at the maximum zoom factor yield higher spectral resolution suitable for Raman spectra. In contrast, larger spectral intervals recorded at the minimum zoom factor yield the lowest spectral resolution suitable for luminescence spectra. We have demonstrated the change in spectral resolution by zoom objective with a zoom factor of 6, but the perspective of such an approach is up to a zoom factor of 20. We have compared such an approach on the prototype Raman spectrometer with the high quality commercial one. The comparison was made on ZrO2 and TiO2 nanocrystals for Raman scattering and Al2O3 for PL emission recording. Beside demonstrating that Raman spectrometer can be used for PL and Raman spectroscopy without changing of grating, our results show that such a spectrometer could be an efficient and fast tool in searching for Raman and PL bands of unknown materials and, thereafter, spectral recording of the spectral interval of interest at an appropriate spectral resolution.

Cross-Borehole DC Resistivity Tomography of Sea Ice: Temporal and Spatial Variations in the Anisotropic Microstructure

<p>As an inhomogeneous mixture of pure ice, brine, air and solid salts the physical properties of sea ice depend on its highly temperature-dependent microstructure. Understanding the microstructure and the way it responds to variations in temperature and salinity is crucial in developing an improved understanding of the interaction between sea ice and the environment. However, measurements monitoring the microstructure of sea ice are difficult to obtain without disturbing its natural state. The brine fraction of sea ice is orders of magnitude more conductive than the solid ice, thus direct current resistivity techniques should yield information on sea ice microstructure. Due to the preferential vertical alignment of brine inclusions, the bulk resistivity of first-year sea ice is anisotropic, complicating interpretation of surface resistivity soundings. However, it can be shown that in a bounded anisotropic medium the resistivity structure may be resolved through in situ cross-borehole measurements. Measurement between borehole pairs, each containing one current and one potential electrode, allows the determination of the horizontal component of the anisotropic bulk resistivity (PH). Using three to four electrodes positioned at approximately the same depth in separate boreholes, provides an under-estimation of the geometric mean resistivity (Pm), and numerical modelling is required to retrieve an estimate of the true Pm. Combining these resistivities allows calculation of the vertical component of the bulk resistivity (PV). This thesis looks at results from measurements made in first year sea ice in April – June 2008 off Barrow, Alaska and in November 2009 off Ross Island, Antarctica. At Barrow, relatively quiescent conditions typically lead to a predominance of columnar ice, while more turbulent conditions and underwater ice formation in McMurdo Sound tend to produce a larger component of frazil or platelet ice. Interpretation of the resistivity measurements, aided by temperature and salinity data, shows that this measurement technique can be used to observe evolution of the ice structure, and distinguish different ice types. Basic two phase structures provide a simple picture of the brine microstructure and how it changes with depth and time. These models indicate the need for vertical connectivity of the brine inclusions even in cool ice, and that PH seems to be mostly due to connections along grain boundaries.</p>

Cross-Borehole DC Resistivity Tomography of Sea Ice: Temporal and Spatial Variations in the Anisotropic Microstructure

<p>As an inhomogeneous mixture of pure ice, brine, air and solid salts the physical properties of sea ice depend on its highly temperature-dependent microstructure. Understanding the microstructure and the way it responds to variations in temperature and salinity is crucial in developing an improved understanding of the interaction between sea ice and the environment. However, measurements monitoring the microstructure of sea ice are difficult to obtain without disturbing its natural state. The brine fraction of sea ice is orders of magnitude more conductive than the solid ice, thus direct current resistivity techniques should yield information on sea ice microstructure. Due to the preferential vertical alignment of brine inclusions, the bulk resistivity of first-year sea ice is anisotropic, complicating interpretation of surface resistivity soundings. However, it can be shown that in a bounded anisotropic medium the resistivity structure may be resolved through in situ cross-borehole measurements. Measurement between borehole pairs, each containing one current and one potential electrode, allows the determination of the horizontal component of the anisotropic bulk resistivity (PH). Using three to four electrodes positioned at approximately the same depth in separate boreholes, provides an under-estimation of the geometric mean resistivity (Pm), and numerical modelling is required to retrieve an estimate of the true Pm. Combining these resistivities allows calculation of the vertical component of the bulk resistivity (PV). This thesis looks at results from measurements made in first year sea ice in April – June 2008 off Barrow, Alaska and in November 2009 off Ross Island, Antarctica. At Barrow, relatively quiescent conditions typically lead to a predominance of columnar ice, while more turbulent conditions and underwater ice formation in McMurdo Sound tend to produce a larger component of frazil or platelet ice. Interpretation of the resistivity measurements, aided by temperature and salinity data, shows that this measurement technique can be used to observe evolution of the ice structure, and distinguish different ice types. Basic two phase structures provide a simple picture of the brine microstructure and how it changes with depth and time. These models indicate the need for vertical connectivity of the brine inclusions even in cool ice, and that PH seems to be mostly due to connections along grain boundaries.</p>

828 Quantifying perivascular immune cells in the stroma of human triple negative breast tumors using deep learning spatial analytics

BackgroundThe spatial organization and density of the immune infiltrate in the tumor microenvironment, referred to as immune contexture, can yield information relevant to prognosis and prediction of response to immunotherapy in cancer. Specifically, a distinct subset of tumor-associated macrophages (TAMs) accumulate around blood vessels where they stimulate tumor angiogenesis and limit tumor responses to frontline anti-cancer therapies like irradiation and chemotherapy.MethodsIn this study we leveraged the NeoGenomics MultiOmyx Multiplex Immunofluorescence platform alongside artificial intelligence (AI) based quantitative image analysis. This AI platform was ultimately used to investigate the distribution of perivascular (PV) TAMs, CD4+ and CD8+ T cells, and CD4+FOXP3+ regulatory T cells (Tregs) of 40 human triple negative breast carcinomas (TNBCs), and how this changed following neoadjuvant chemotherapy. During the multiplexing phase, eleven rounds of paired antibody staining were performed in sequence on tumor sections. After each round of staining, high resolution images were captured for regions of interests (ROIs) selected by a pathologist. We used AI models to segment and classify cells for each biomarker and classify regions as tumor cell islands (TCIs) or stroma. First, each nucleus was segmented out using a convolutional neural network combined with watershed thresholding on the DAPI (diamidino-2-phenylindole) immunofluorescent image. From the resulting nuclear segmentation mask, a pixel dilation on cells classified as non-tumor was employed to generate a cellular segmentation mask. A list of neighbours within a specified distance for each cell was generated by radially expanding from the cellular segmentation mask. Finally, cell neighbour information was combined with the marker expression information to quantify the cell clusters of interest.ResultsWe discovered that in the PV areas, up to 30% of PD1-LAG3-CD3+CD8+ T cells formed direct contact with both CD163+TIM3+ TAMs and CD4+FOXP3+ Tregs. Furthermore, these immune cell triads preferentially accumulated in the PV stroma regions. It is likely that close interaction with immunosuppressive TAMs and Tregs would supress the function of T cells as they enter the PV region to reach the TCIs.ConclusionsUsing an advanced analytics platform, we invented a new method to quantify clusters of cells within various regions of a tumor section. Using this platform, we detected specific immune cell triads, the frequency and location of which could correlate with the efficacy of T-cell based immunotherapies in TNBC. These analyses will enable further investigation of numerous complex cell interactions in TMEs.

Optimization of Water Quantity with Smart Agriculture in IOT

In this paper, we propose fostering a framework ideally watering horticultural harvests dependent on a remote sensor organization. This work planned to plan and foster a control framework utilizing hub sensors in the harvest field with information the executives by means of cell phone and a web application. The three parts are equipment, web application, and portable application. The principal segment was planned and executed in charge box equipment associated with gather information on the harvests. Soil dampness sensors are utilized to screen the field, associating with the control box. The subsequent segment is an electronic application that was planned and executed to control the subtleties of yield information and field data. This segment applied information mining to break down the information for anticipating appropriate temperature, stickiness, and soil dampness for ideal future administration of harvests development. The last part is essentially used to control crop watering through a versatile application in a cell phone. This permits either programmed or manual control by the client. The programmed control utilizes information from soil dampness sensors for watering. Be that as it may, the client can pick manual control of watering the harvests in the useful control mode. The framework can send notices through LINE API for the LINE application. The framework was executed and tried in Makhamtia District, Suratthani Province, Thailand. The outcomes demonstrated the execution to be valuable in agribusiness. The dampness content of the dirt was kept up with properly for vegetable development, lessening costs and expanding rural usefulness. Also, this work addresses driving farming through computerized development.

Sensitive quantification of m.3243A>G mutational proportion in non-retinal tissues and its relationship with visual symptoms

The m.3243A&gt;G mutation in the mitochondrial genome commonly causes retinal degeneration in patients with maternally inherited diabetes and deafness (MIDD) and mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes (MELAS). Like other mitochondrial mutations, m.3243A&gt;G is inherited from the mother with a variable proportion of wild type and mutant mitochondrial genomes in different cells. The mechanism by which the m.3243A&gt;G variant in each tissue relates to the manifestation of disease phenotype is not fully understood. Using a digital PCR assay we found that the % m.3243G in skin derived dermal fibroblasts was positively correlated with that of blood from the same individual. The % m.3243G detected in fibroblast cultures remained constant over multiple passages and was negatively correlated with mtDNA copy number. Although the % m.3243G present in blood was not correlated with severity of vision loss, as quantified by Goldmann visual field, a significant negative correlation between % m.3243G and the age of onset of visual symptoms was detected. Together, these results indicate that precise measurement of % m.3243G in clinically accessible tissues such as skin and blood may yield information relevant to the management of retinal m.3243A&gt;G associated disease.