Terms Of Patches Research Articles

Effectiveness of transfer learning for enhancing tumor classification with a convolutional neural network on frozen sections

Fast and accurate confirmation of metastasis on the frozen tissue section of intraoperative sentinel lymph node biopsy is an essential tool for critical surgical decisions. However, accurate diagnosis by pathologists is difficult within the time limitations. Training a robust and accurate deep learning model is also difficult owing to the limited number of frozen datasets with high quality labels. To overcome these issues, we validated the effectiveness of transfer learning from CAMELYON16 to improve performance of the convolutional neural network (CNN)-based classification model on our frozen dataset (N = 297) from Asan Medical Center (AMC). Among the 297 whole slide images (WSIs), 157 and 40 WSIs were used to train deep learning models with different dataset ratios at 2, 4, 8, 20, 40, and 100%. The remaining, i.e., 100 WSIs, were used to validate model performance in terms of patch- and slide-level classification. An additional 228 WSIs from Seoul National University Bundang Hospital (SNUBH) were used as an external validation. Three initial weights, i.e., scratch-based (random initialization), ImageNet-based, and CAMELYON16-based models were used to validate their effectiveness in external validation. In the patch-level classification results on the AMC dataset, CAMELYON16-based models trained with a small dataset (up to 40%, i.e., 62 WSIs) showed a significantly higher area under the curve (AUC) of 0.929 than those of the scratch- and ImageNet-based models at 0.897 and 0.919, respectively, while CAMELYON16-based and ImageNet-based models trained with 100% of the training dataset showed comparable AUCs at 0.944 and 0.943, respectively. For the external validation, CAMELYON16-based models showed higher AUCs than those of the scratch- and ImageNet-based models. Model performance for slide feasibility of the transfer learning to enhance model performance was validated in the case of frozen section datasets with limited numbers.

Biophysical Interactions in Fragmented Marine Canopies: Fundamental Processes, Consequences, and Upscaling

Spatial fragmentation is a near-ubiquitous characteristic of marine canopies. Biophysical interactions with fragmented canopies are multi-faceted and have many significant implications at multiple scales. The aims of this paper are to review research on biophysical interactions in fragmented marine canopies, identify current gaps in knowledge and understanding, and propose ways forward. Although fragmented landscapes may be analyzed with relatively sophisticated approaches based on gradients and non-linear ecosystem interactions, it is argued that a simpler conceptualization in terms of patches and gaps is appropriate in this context. The review starts at the patch/gap scale and focuses initially on hydrodynamic interactions. It then considers the consequences of these interactions for particulate and dissolved material, and distributions of canopy-associated organisms. Finally, it addresses issues of upscaling to landscape scale and ways in which this research can be applied to marine landscape management, conservation and restoration. Work on a broad range of canopy types is considered, including micro-algal biofilms and turf algae; macro-algae, seagrasses and coral reefs; saltmarsh vegetation and mangroves. Although the focus is on marine canopies, insights from studies of fragmented canopies in other contexts are drawn on where relevant. These include freshwater environments and terrestrial forests, grasslands, crop canopies and urban areas. Specific areas requiring greater attention are highlighted. The review finds that more studies are needed of: wave interactions with patches and gaps; three-dimensional biophysical interactions with gaps; and biophysical interactions with fuzzy or indistinct patch/gap boundaries. More cross-comparisons of studies of hydrodynamic interactions with different types of canopies are also required. The primary obstacle to progress in this field is found to be the current gap between the most common scale for measurement and modelling – the patch/gap scale – and landscape scale, at which the insights delivered by research may be applied most usefully. Therefore, upscaling is identified as the most important current problem in this field. To address this, a combination of the recently-founded fields of eco-mechanics and hydrodynamics of aquatic ecosystems that balances physical, biological and other perspectives, rather than viewing one as being served by the others, would be the best way forward.

Prediction of Multidrug-Resistant TB from CT Pulmonary Images Based on Deep Learning Techniques.

While tuberculosis (TB) disease was discovered more than a century ago, it has not been eradicated yet. Quite contrary, at present, TB constitutes one of the top 10 causes of death and has shown signs of increasing. To complement the conventional diagnostic procedure of applying microbiological culture that takes several weeks and remains expensive, high resolution computer tomography (CT) of pulmonary images has been resorted to not only for aiding clinicians to expedite the process of diagnosis but also for monitoring prognosis when administering antibiotic drugs. This research undertakes the investigation of predicting multidrug-resistant (MDR) patients from drug-sensitive (DS) ones based on CT lung images to monitor the effectiveness of treatment. To contend with smaller data sets (i.e., hundreds) and the characteristics of CT TB images with limited regions capturing abnormities, patch-based deep convolutional neural network (CNN) allied to support vector machine (SVM) classifier is implemented on a collection of data sets from 230 patients obtained from the ImageCLEF 2017 competition. As a result, the proposed architecture of CNN + SVM + patch performs the best with classification accuracy rate at 91.11% (79.80% in terms of patches). In addition, a hand-crafted SIFT based approach accomplishes 88.88% in terms of subject and 83.56% with reference to patches, the highest in this study, which can be explained away by the fact that the data sets are in small numbers. Significantly, during the Tuberculosis Competition at ImageCLEF 2017, the authors took part in the task of classification of 5 types of TB disease and achieved the top one with regard to averaged classification accuracy (i.e., ACC = 0.4067), which is also premised on the approach of CNN + SVM + patch. On the other hand, when the whole slices of 3D TB data sets are applied to train a CNN network, the best result is achieved through the application of CNN coupled with orderless pooling and SVM at 64.71% accuracy rate.

Exemplar-based color image inpainting: a fractional gradient function approach

Image inpainting is the art of recovering a plausible image from images which are generally incomplete due to various factors, including degradation due to ageing, damage due to wear and tear and missing image details due to occlusion. In such situations, there is a need to predict the missing image information without introducing undesirable artifacts. Original contribution in this direction is due to a seminal paper by Criminisi et al. This has led to a number of novel contributions in terms of patch filling prioritization and associated metrics to measure color and structure. In this paper, we propose a fast and simple technique based on a novel gradient function and its generalization via fractional derivatives to evaluate the filling order prioritization. Results demonstrate superior and robust performance over all the recent advances in the domain of exemplar-based methods quoted in the literature.

Spatial analysis by distance indices: an alternative local clustering index for studying spatial patterns

Summary1. The spatial analysis by distance indices (SADIE) methodology for data analysis is valuable for quantifying spatial patterns of organisms in terms of patches and gaps. Previous research showed that the calculation of the local clustering indices, key SADIE statistics, does not adequately adjust for the absolute location or the magnitude of the counts.2. We present a new definition of a local clustering index, which overcomes the problem associated with the original cluster indices related to sampling position and count size. The new index is calculated without breaking the link between the observed count and its original position and quantifies the contribution of an observed count at this particular position to the local gaps or patches for the observed pattern relative to the expected under the assumption of spatial independence amongst observed counts. Randomisation‐based testing for statistical significance of an individual local clustering index follows naturally from the definition of the new index.3. New indices, calculated for several simulated and observed data sets, showed that the original indices overestimated the number of points (sites, locations) contributing to the gaps/patches in a spatial grid. Results indicate that the significance (or interpretation) of individual local clustering indices cannot be made based on its magnitude only and needs to be supported by statistical testing.4. The newly developed index will provide a valuable tool for quantifying the local pattern and testing for its significance and enhance the value of SADIE methodology in analysing spatial patterns. It can also be used in conjunction with other approaches that test for global clustering.

Characterisation of Agri-Landscape Systems at a Regional Level: A Case Study in Northern Tuscany

Preserving our landscape in sustainable development processes is now widely considered as fundamental. It is a complex and evolving issue that can be tackled from several perspectives. Agronomy can contribute to analyzing the relationships between agricultural production systems (cropping, farming and agricultural systems) at different levels (field, farm, and region) and the agricultural landscape (in terms of patches, matrixes, dynamics, etc). This is of particular interest where the relationships between “what and how” are produced by agricultural activities and the landscape are changing. In this case their own reciprocity may represent an opportunity to analyze complex systems, such as the characterization of agri-landscapes at a regional level. We propose a case study developed as an up-scaling analytical process from a farm to a regional level. The result was the identification of six main agri-landscape systems highlighting the landscape drivers that are changing the traditional landscape of a rural region in Northern Tuscany (Lunigiana).

SmPL: A Domain-Specific Language for Specifying Collateral Evolutions in Linux Device Drivers

Collateral evolutions are a pervasive problem in large-scale software development. Such evolutions occur when an evolution that affects the interface of a generic library entails modifications, i.e., collateral evolutions, in all library clients. Performing these collateral evolutions requires identifying the affected files and modifying all of the code fragments in these files that in some way depend on the changed interface.We have studied the collateral evolution problem in the context of Linux device drivers. Currently, collateral evolutions in Linux are mostly done manually using a text editor, possibly with the help of tools such as grep. The large number of Linux drivers, however, implies that this approach is time-consuming and unreliable, leading to subtle errors when modifications are not done consistently.In this paper, we propose a transformation language, SmPL, to specify collateral evolutions. Because Linux programmers are accustomed to exchanging, reading, and manipulating program modifications in terms of patches, we build our language around the idea and syntax of a patch, extending patches to semantic patches.

Landscape ecology planning principles in Korean Feng-Shui, Bi-bo woodlands and ponds

The purpose of this study is to suggest practical planning principles, which are absent in landscape ecology planning, according to which the spatial completeness of the watershed is derived from native Korean Feng-Shui ( ). Landscape ecology planning principles are used widely in contemporary planning projects in terms of patch, corridor, matrix and network of landscape pattern. However, landscape indices for planning principles are complicated and constrained, and so are limited to applications for site and eco-village plannings. Native Korean Feng-Shui is different from Chinese Feng-Shui in that it is aimed at theoretical completeness in terms of aspect and topographical shape, based on the concept of ideal Myung-dang ( ) to complete the space, according to local conditions in the physiognomy of the watershed. The complementary method is called Bi-bo ( ) in native Korean Feng-Shui. These principles have been applied in traditional Korean villages, leading to consistent location choices and fractal patterns in land use. Furthermore, Bi-bo woodlands and ponds have been introduced to achieve spatial completeness in the landscape structure of the village.

Interrelationships Among SADIE Indices for Characterizing Spatial Patterns of Organisms

ABSTRACT The SADIE (spatial analysis by distance indices) methodology for data analysis is a useful approach for quantifying the patterns of organisms (in terms of patches and gaps) and testing for randomness of the patterns. We investigated the interrelationship among key SADIE indices: index for distance to regularity for a data set (I(a)), a global measure of aggregation or clustering; the local clustering indices (v(i) and v(j)), scaled distances to regularity for each individual sampling unit; and the averages of v(i) and v(j) across all sampling units, which are additional global measures of aggregation. We demonstrated that v(i) and v(j) are mathematically related to I(a) and showed conditions when I(a) and mean local clustering indices give very similar results. Overall differences in average v(i) and Iv(j) I values, and between I(a) and these averages, decreased with increasing size of the sampling grid in a simulation study. This was because one component of v(i) and v(j) (iY)-a measure of the distance to regularity under randomness for a given location (not a given count)-was found generally to vary little with location, except for locations near corners of the sampling grid. Nevertheless, because distance to regularity for individual observed counts was location-dependent, and this location effect varied with the observed counts value as well, a new-scaled index for each count x location combination may be warranted. The implications of these findings on epidemiological research are discussed.

On final states of two-dimensional decaying turbulence

Numerical and analytical studies of final states of two-dimensional (2D) decaying turbulence are carried out. The first part of this work is trying to give a definition for final states of 2D decaying turbulence. The functional relation of ω-ψ, which is frequently adopted as the characterization of those final states, is merely a sufficient but not necessary condition; moreover, it is not proper to use it as the definition. It is found that the method through the value of the effective area S covered by the scatter ω-ψ plot, initially suggested by Read, Rhines, and White [“Geostrophic scatter diagrams and potential vorticity dynamics,” J. Atmos. Sci. 43, 3226 (1986)] is more general and suitable for the definition. Based on this concept, a definition is presented, which covers all existing results in late states of decaying 2D flows (including some previous unexplainable weird double-valued ω-ψ scatter plots). The remaining part of the paper is trying to further study 2D decaying turbulence with the assistance of this definition. Some numerical results, leading to “bar” final states and further verifying the predictive ability of statistical mechanics [Yin, Montgomery, and Clercx, “Alternative statistical-mechanical descriptions of decaying two-dimensional turbulence in terms of patches and points,” Phys. Fluids 15, 1937 (2003)], are reported. It is realized that some simulations with narrow-band energy spectral initial conditions result in some final states that cannot be very well interpreted by the statistical theory (meanwhile, those final states are still in the scope of the definition).

The application of local measures of spatial autocorrelation for describing pattern in north Australian landscapes

This paper tests the use of a spatial analysis technique, based on the calculation of local spatial autocorrelation, as a possible approach for modelling and quantifying structure in northern Australian savanna landscapes. Unlike many landscapes in the world, northern Australian savanna landscapes appear on the surface to be intact. They have not experienced the same large-scale land clearance and intensive land management as other landscapes across Australia. Despite this, natural resource managers are beginning to notice that processes are breaking down and declines in species are becoming more evident. With future declines of species looking more imminent it is particularly important that models are available that can help to assess landscape health, and quantify any structural change that takes place. GIS and landscape ecology provide a useful way of describing landscapes both spatially and temporally and have proved to be particularly useful for understanding vegetation structure or pattern in landscapes across the world. There are many measures that examine spatial structure in the landscape and most of these are now available in a GIS environment (e.g. FRAGSTATS*ARC, r.le, and Patch Analyst). All these methods depend on a landscape described in terms of patches, corridors and matrix. However, since landscapes in northern Australia appear to be relatively intact they tend to exist as surfaces of continuous variation rather than in clearly defined homogeneous units. As a result they cannot be easily described using entity-based models requiring patches and other essentially cartographic approaches. This means that more appropriate methods need to be developed and explored. The approach examined in this paper enables clustering and local pattern in the data to be identified and forms a generic method for conceptualising the landscape structure where patches are not obvious and where boundaries between landscape features are difficult to determine. Two sites are examined using this approach. They have been exposed to different degrees of disturbance by fire and grazing. The results show that savanna landscapes are very complex and that even where there is a high degree of disturbance the landscape is still relatively heterogeneous. This means that treating savanna landscapes as being made up of homogeneous units can limit analysis of pattern, as it can over simplify the structure present, and that methods such as the autocorrelation approach are useful tools for quantifying the variable nature of these landscapes.