Active learning: A step towards automating medical concept extraction

Objective This paper presents an automatic active learning-based system for the extraction of medical concepts from clinical free-text reports. Specifically, (1) the contribution of active learning in reducing the annotation effort, and (2) the robustness of incremental active learning framework across different selection criteria and datasets is determined. Materials and methods The comparative performance of an active learning framework and a fully supervised approach were investigated to study how active learning reduces the annotation effort while achieving the same effectiveness as a supervised approach. Conditional Random Fields as the supervised method, and least confidence and information density as two selection criteria for active learning framework were used. The effect of incremental learning vs. standard learning on the robustness of the models within the active learning framework with different selection criteria was also investigated. Two clinical datasets were used for evaluation: the i2b2/VA 2010 NLP challenge and the ShARe/CLEF 2013 eHealth Evaluation Lab. Results The annotation effort saved by active learning to achieve the same effectiveness as supervised learning is up to 77%, 57%, and 46% of the total number of sequences, tokens, and concepts, respectively. Compared to the Random sampling baseline, the saving is at least doubled. Discussion Incremental active learning guarantees robustness across all selection criteria and datasets. The reduction of annotation effort is always above random sampling and longest sequence baselines. Conclusion Incremental active learning is a promising approach for building effective and robust medical concept extraction models, while significantly reducing the burden of manual annotation.

External knowledge and query strategies in active learning: A study in clinical information extraction

This paper presents a new active learning query strategy for information extraction, called Domain Knowledge Informativeness (DKI). Active learning is often used to reduce the amount of annotation effort required to obtain training data for machine learning algorithms. A key component of an active learning approach is the query strategy, which is used to iteratively select samples for annotation. Knowledge resources have been used in information extraction as a means to derive additional features for sample representation. DKI is, however, the first query strategy that exploits such resources to inform sample selection. To evaluate the merits of DKI, in particular with respect to the reduction in annotation effort that the new query strategy allows to achieve, we conduct a comprehensive empirical comparison of active learning query strategies for information extraction within the clinical domain. The clinical domain was chosen for this work because of the availability of extensive structured knowledge resources which have often been exploited for feature generation. In addition, the clinical domain offers a compelling use case for active learning because of the necessary high costs and hurdles associated with obtaining annotations in this domain. Our experimental findings demonstrated that 1) amongst existing query strategies, the ones based on the classification model’s confidence are a better choice for clinical data as they perform equally well with a much lighter computational load, and 2) significant reductions in annotation effort are achievable by exploiting knowledge resources within active learning query strategies, with up to 14% less tokens and concepts to manually annotate than with state-of-the-art query strategies.

A reproducible method for automated extraction of brain volumes from 3D human head MR images

An automated method for extracting brain volumes from three commonly acquired three-dimensional (3D) MR images (proton density, T1 weighted, and T2-weighted) of the human head is described. The procedure is divided into four levels: preprocessing, segmentation, scalp removal, and postprocessing. A user-provided reference point is the sole operator-dependent input required. The method's parameters were first optimized and then fixed and applied to 30 repeat data sets from 15 normal older adult subjects to investigate its reproducibility. Percent differences between total brain volumes (TBVs) for the subjects' repeated data sets ranged from .5% to 2.2%. We conclude that the method is both robust and reproducible and has the potential for wide application.

Plant species’ origin predicts dominance and response to nutrient enrichment and herbivores in global grasslands

Exotic species dominate many communities; however the functional significance of species’ biogeographic origin remains highly contentious. This debate is fuelled in part by the lack of globally replicated, systematic data assessing the relationship between species provenance, function and response to perturbations. We examined the abundance of native and exotic plant species at 64 grasslands in 13 countries, and at a subset of the sites we experimentally tested native and exotic species responses to two fundamental drivers of invasion, mineral nutrient supplies and vertebrate herbivory. Exotic species are six times more likely to dominate communities than native species. Furthermore, while experimental nutrient addition increases the cover and richness of exotic species, nutrients decrease native diversity and cover. Native and exotic species also differ in their response to vertebrate consumer exclusion. These results suggest that species origin has functional significance, and that eutrophication will lead to increased exotic dominance in grasslands.

Utility of web content blocks in content extraction

Currently we are facing an overburdening growth of the number of reliable information sources on the Internet. The quantity of information available to everyone via Internet is dramatically growing each year [15]. At the same time, temporal and cognitive resources of human users are not changing, therefore causing a phenomenon of information overload. World Wide Web is one of the main sources of information for decision makers (reference to my research). However our studies show that, at least in Poland, the decision makers see some important problems when turning to Internet as a source of decision information. One of the most common obstacles raised is distribution of relevant information among many sources, and therefore need to visit different Web sources in order to collect all important content and analyze it. A few research groups have recently turned to the problem of information extraction from the Web [13]. The most effort so far has been directed toward collecting data from dispersed databases accessible via web pages (related to as data extraction or information extraction from the Web) and towards understanding natural language texts by means of fact, entity, and association recognition (related to as information extraction). Data extraction efforts show some interesting results, however proper integration of web databases is still beyond us. Information extraction field has been recently very successful in retrieving information from natural language texts, however it is still lacking abilities to understand more complex information, requiring use of common sense knowledge, discourse analysis and disambiguation techniques.

Robust Web content extraction

We present an empirical evaluation and comparison of two content extraction methods in HTML: absolute XPath expressions and relative XPath expressions. We argue that the relative XPath expressions, although not widely used, should be used in preference to absolute XPath expressions in extracting content from human-created Web documents. Evaluation of robustness covers four thousand queries executed on several hundred webpages. We show that in referencing parts of real world dynamic HTML documents, relative XPath expressions are on average significantly more robust than absolute XPath ones.

Consistent responses of soil microbial communities to elevated nutrient inputs in grasslands across the globe

Soil microorganisms are critical to ecosystem functioning and the maintenance of soil fertility. However, despite global increases in the inputs of nitrogen (N) and phosphorus (P) to ecosystems due to human activities, we lack a predictive understanding of how microbial communities respond to elevated nutrient inputs across environmental gradients. Here we used high-throughput sequencing of marker genes to elucidate the responses of soil fungal, archaeal, and bacterial communities using an N and P addition experiment replicated at 25 globally distributed grassland sites. We also sequenced metagenomes from a subset of the sites to determine how the functional attributes of bacterial communities change in response to elevated nutrients. Despite strong compositional differences across sites, microbial communities shifted in a consistent manner with N or P additions, and the magnitude of these shifts was related to the magnitude of plant community responses to nutrient inputs. Mycorrhizal fungi and methanogenic archaea decreased in relative abundance with nutrient additions, as did the relative abundances of oligotrophic bacterial taxa. The metagenomic data provided additional evidence for this shift in bacterial life history strategies because nutrient additions decreased the average genome sizes of the bacterial community members and elicited changes in the relative abundances of representative functional genes. Our results suggest that elevated N and P inputs lead to predictable shifts in the taxonomic and functional traits of soil microbial communities, including increases in the relative abundances of faster-growing, copiotrophic bacterial taxa, with these shifts likely to impact belowground ecosystems worldwide.

Relationship between living alone and food and nutrient intake

The increase in the number of individuals living alone has implications for nutrition and health outcomes. This review aimed to investigate whether there is a difference in food and nutrient intake between adults living alone and those living with others. Eight electronic databases were searched, using terms related to living alone, nutrition, food, and socioeconomic factors. Forty-one papers met the inclusion criteria, and data of interest were extracted. Results varied but suggested that, compared with persons who do not live alone, persons who live alone have a lower diversity of food intake, a lower consumption of some core foods groups (fruit, vegetables, and fish) and a higher likelihood of having an unhealthy dietary pattern. Associations between living alone and nutrient intake were unclear. Men living alone were more often observed to be at greater risk of undesirable intakes than women. The findings of this review suggest that living alone could negatively affect some aspects of food intake and contribute to the relationship between living alone and poor health outcomes, although associations could vary between socioeconomic groups. Further research is required to help to elucidate these findings.

Energy and nutrient intakes among Sri Lankan adults

Introduction The epidemic of nutrition related non-communicable diseases such as type 2 diabetes mellitus and obesity has reached to epidemic portion in the Sri Lanka. However, to date, detailed data on food consumption in the Sri Lankan population is limited. The aim of this study is to identify energy and major nutrient intake among Sri Lankan adults. Methods A nationally-representative sample of adults was selected using a multi-stage random cluster sampling technique. Results Data from 463 participants (166 Males, 297 Females) were analyzed. Total energy intake was significantly higher in males (1913 ± 567 kcal/d) than females (1514 ± 458 kcal/d). However, there was no significant gender differences in the percentage of energy from carbohydrate (Male: 72.8 ± 6.4%, Female: 73.9 ± 6.7%), fat (Male: 19.9 ± 6.1%, Female: 18.5 ± 5.7%) and proteins (Male: 10.6 ± 2.1%, Female: 10.9 ± 5.6%). Conclusion The present study provides the first national estimates of energy and nutrient intake of the Sri Lankan adult population.

Determination of tricyclazole in water using solid phase extraction and liquid chromatography

A method for determination of tricyclazole in water using solid phase extraction and high performance liquid chromatography (HPLC) with UV detection at 230nm and a mobile phase of acetonitrile:water (20:80, v/v) was developed. A performance comparison between two types of solid phase sorbents, the C18 sorbent of Supelclean ENVI-18 cartridge and the styrene-divinyl benzene copolymer sorbent of Sep-Pak PS2-Plus cartridge was conducted. The Sep-Pak PS2-Plus cartridges were found more suitable for extracting tricyclazole from water samples than the Supelclean ENVI-18 cartridges. For this cartridge, both methanol and ethyl acetate produced good results. The method was validated with good linearity and with a limit of detection of 0.008gL-1 for a 500-fold concentration through the SPE procedure. The recoveries of the method were stable at 80% and the precision was from 1.1-6.0% within the range of fortified concentrations. The validated method was also applied to measure the concentrations of tricyclazole in real paddy water.

Acoustic feature extraction using perceptual wavelet packet decomposition for frog call classification

Frog protection has become increasingly essential due to the rapid decline of its biodiversity. Therefore, it is valuable to develop new methods for studying this biodiversity. In this paper, a novel feature extraction method is proposed based on perceptual wavelet packet decomposition for classifying frog calls in noisy environments. Pre-processing and syllable segmentation are first applied to the frog call. Then, a spectral peak track is extracted from each syllable if possible. Track duration, dominant frequency and oscillation rate are directly extracted from the track. With k-means clustering algorithm, the calculated dominant frequency of all frog species is clustered into k parts, which produce a frequency scale for wavelet packet decomposition. Based on the adaptive frequency scale, wavelet packet decomposition is applied to the frog calls. Using the wavelet packet decomposition coefficients, a new feature set named perceptual wavelet packet decomposition sub-band cepstral coefficients is extracted. Finally, a k-nearest neighbour (k-NN) classifier is used for the classification. The experiment results show that the proposed features can achieve an average classification accuracy of 97.45% which outperforms syllable features (86.87%) and Mel-frequency cepstral coefficients (MFCCs) feature (90.80%).

Analysis of word embeddings and sequence features for clinical information extraction

This study investigates the use of unsupervised features derived from word embedding approaches and novel sequence representation approaches for improving clinical information extraction systems. Our results corroborate previous findings that indicate that the use of word embeddings significantly improve the effectiveness of concept extraction models; however, we further determine the influence that the corpora used to generate such features have. We also demonstrate the promise of sequence-based unsupervised features for further improving concept extraction.

What do prisoners eat? Nutrient intakes and food practices in a high secure prison

There are limited studies on the adequacy of prisoner diet and food practices, yet understanding these are important to inform food provision and assure duty of care for this group. The aim of this research was to assess the dietary intakes of prisoners to inform food and nutrition policy in this setting. This research used a cross-sectional design with convenience sampling in a 945 bed male high secure prison. Multiple methods were used to assess food available at the group level, including verification of food portion, quality, and practices. A pictorial tool supported the diet history method. Of 276 eligible prisoners, 120 dietary interviews were conducted and verified against prison records, with 106 deemed plausible. The results showed the planned food to be nutritionally adequate, with the exception of vitamin D for older males and long chain fatty acids, with sodium above Upper Limits. The Australian Dietary Targets for chronic disease risk were not achieved. High energy intakes were reported with median 13.8MJ (SE 0.3MJ). Probability estimates of inadequate intake varied with age groups: magnesium 8% (>30 years), 2.9% (<30 years); calcium 6.0% (>70 years), 1.5% (<70 years); folate 3.5%; zinc and iodine 2.7%; and vitamin A 2.3%. Nutrient intakes were greatly impacted by self-funded snacks. Results suggest nutrient intakes nutritionally favourable when compared to males in the community. This study highlights the complexity of food provision in the prison environment, and also poses questions for population level dietary guidance in delivering appropriate nutrients within energy limits.

Optimisation of in-cell accelerated solvent extraction technique for the determination of organochlorine pesticides in river sediments

Organochlorine pesticides (OCPs) are ubiquitous environmental contaminants with adverse impacts on aquatic biota, wildlife and human health even at low concentrations. However, conventional methods for their determination in river sediments are resource intensive. This paper presents an approach that is rapid and also reliable for the detection of OCPs. Accelerated Solvent Extraction (ASE) with in-cell silica gel clean-up followed by Triple Quadrupole Gas Chromatograph Mass Spectrometry (GCMS/MS) was used to recover OCPs from sediment samples. Variables such as temperature, solvent ratio, adsorbent mass and extraction cycle were evaluated and optimised for the extraction. With the exception of Aldrin, which was unaffected by any of the variables evaluated, the recovery of OCPs from sediment samples was largely influenced by solvent ratio and adsorbent mass and, to some extent, the number of cycles and temperature. The optimised conditions for OCPs extraction in sediment with good recoveries were determined to be 4 cycles, 4.5 g of silica gel, 105 ᴼC, and 4:3 v/v DCM: hexane mixture. With the exception of two compounds (α-BHC and Aldrin) whose recoveries were low (59.73 and 47.66 % respectively), the recovery of the other pesticides were in the range 85.35 – 117.97% with precision < 10 % RSD. The method developed significantly reduces sample preparation time, the amount of solvent used, matrix interference, and is highly sensitive and selective.

vSpeak: Edge detection based feature extraction for sign to text conversion

This paper presents 'vSpeak', the first initiative taken in Pakistan for ICT enabled conversion of dynamic Sign Urdu gestures into natural language sentences. To realize this, vSpeak has adopted a novel approach for feature extraction using edge detection and image compression which gives input to the Artificial Neural Network that recognizes the gesture. This technique caters for the blurred images as well. The training and testing is currently being performed on a dataset of 200 patterns of 20 words from Sign Urdu with target accuracy of 90% and above.