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.

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.

The alchemy of junk: patent law and non-coding DNA

This article considers the recent international controversy over the patents held by a Melbourne firm, Genetic Technologies Limited (GTG), in respect of non-coding DNA and genomic mapping. It explores the ramifications of the GTG dispute in terms of licensing, litigation, and policy reform, and—as a result of this dispute—the perceived conflict between law and science. GTG has embarked upon an ambitious licensing program with twenty seven commercial licensees and five research licensees. Most significantly, GTG has obtained an exclusive licence from Myriad Genetics to use and exploit its medical diagnostics in Australia, New Zealand, and the Asia-Pacific region. In the US, GTG brought a legal action for patent infringement against the Applera Corporation and its subsidiaries. In response, Applera counterclaimed that the patents of GTG were invalid because they failed to comply with the requirements of US patent law, such as novelty, inventive step, and written specifications. In New Zealand, the Auckland District Health Board brought legal action in the High Court, seeking a declaration that the patents of GTG were invalid, and that, in any case, the Board has not infringed them. The New Zealand Ministry of Health and the Ministry of Economic Development have reported to Cabinet on the issues relating to the patenting of genetic material. Similarly, the Australian Law Reform Commission (ALRC) has also engaged in an inquiry into gene patents and human health; and the Advisory Council on Intellectual Property (ACIP) has considered whether there should be a new defence in respect of experimental use and research.

Mitochondrial DNA sequence analysis from multiple gene fragments reveals genetic heterogeneity of Crassostrea ariakensis in East Asia

The native Asian oyster, Crassostrea ariakensis is one of the most common and important Crassostrea species that occur naturally along the coast of East Asia. Molecular species diagnosis is a prerequisite for population genetic analysis of wild oyster populations because oyster species cannot be discriminated reliably using external morphological characters alone due to character ambiguity. To date there have been few phylogeographic studies of natural edible oyster populations in East Asia, in particular this is true of the common species in Korea C. ariakensis. We therefore assessed the levels and patterns of molecular genetic variation in East Asian wild populations of C. ariakensis from Korea, Japan, and China using DNA sequence analysis of five concatenated mtDNA regions namely; 16S rRNA, cytochrome oxidase I, cytochrome oxidase II, cytochrome oxidase III, and cytochrome b. Two divergent C. ariakensis clades were identified between southern China and remaining sites from the northern region. In addition, hierarchical AMOVA and pairwise UST analyses showed that genetic diversity was discontinuous among wild populations of C. ariakensis in East Asia. Biogeographical and historical sea level changes are discussed as potential factors that may have influenced the genetic heterogeneity of wild C. ariakensis stocks across this region.

Super DNAging - New insights into DNA integrity, genome stability and telomeres in the oldest old

Reductions in DNA integrity, genome stability, and telomere length are strongly associated with the aging process, age-related diseases as well as the age-related loss of muscle mass. However, in people reaching an age far beyond their statistical life expectancy the prevalence of diseases, such as cancer, cardiovascular disease, diabetes or dementia, is much lower compared to “averagely” aged humans. These inverse observations in nonagenarians (90–99 years), centenarians (100–109 years) and super-centenarians (110 years and older) require a closer look into dynamics underlying DNA damage within the oldest old of our society. Available data indicate improved DNA repair and antioxidant defense mechanisms in “super old” humans, which are comparable with much younger cohorts. Partly as a result of these enhanced endogenous repair and protective mechanisms, the oldest old humans appear to cope better with risk factors for DNA damage over their lifetime compared to subjects whose lifespan coincides with the statistical life expectancy. This model is supported by study results demonstrating superior chromosomal stability, telomere dynamics and DNA integrity in “successful agers”. There is also compelling evidence suggesting that life-style related factors including regular physical activity, a well-balanced diet and minimized psycho-social stress can reduce DNA damage and improve chromosomal stability. The most conclusive picture that emerges from reviewing the literature is that reaching “super old” age appears to be primarily determined by hereditary/genetic factors, while a healthy lifestyle additionally contributes to achieving the individual maximum lifespan in humans. More research is required in this rapidly growing population of super old people. In particular, there is need for more comprehensive investigations including short- and long-term lifestyle interventions as well as investigations focusing on the mechanisms causing DNA damage, mutations, and telomere shortening.

The IFITM5 mutation c.-14C > T results in an elongated transcript expressed in human bone; And causes varying phenotypic severity of osteogenesis imperfecta type V

Background The genetic mutation resulting in osteogenesis imperfecta (OI) type V was recently characterised as a single point mutation (c.-14C > T) in the 5' untranslated region (UTR) of IFITM5, a gene encoding a transmembrane protein with expression restricted to skeletal tissue. This mutation creates an alternative start codon and has been shown in a eukaryotic cell line to result in a longer variant of IFITM5, but its expression has not previously been demonstrated in bone from a patient with OI type V. Methods Sanger sequencing of the IFITM5 5' UTR was performed in our cohort of subjects with a clinical diagnosis of OI type V. Clinical data was collated from referring clinicians. RNA was extracted from a bone sample from one patient and Sanger sequenced to determine expression of wild-type and mutant IFITM5. Results: All nine subjects with OI type V were heterozygous for the c.-14C > T IFITM5 mutation. Clinically, there was heterogeneity in phenotype, particularly in the manifestation of bone fragility amongst subjects. Both wild-type and mutant IFITM5 mRNA transcripts were present in bone. Conclusions The c.-14C > T IFITM5 mutation does not result in an RNA-null allele but is expressed in bone. Individuals with identical mutations in IFITM5 have highly variable phenotypic expression, even within the same family.

Expression profiling in spondyloarthropathy synovial biopsies highlights changes in expression of inflammatory genes in conjunction with tissue remodelling genes

Background: In the spondyloarthropathies, the underlying molecular and cellular pathways driving disease are poorly understood. By undertaking a study in knee synovial biopsies from spondyloarthropathy (SpA) and ankylosing spondylitis (AS) patients we aimed to elucidate dysregulated genes and pathways. Methods RNA was extracted from six SpA, two AS, three osteoarthritis (OA) and four normal control knee synovial biopsies. Whole genome expression profiling was undertaken using the Illumina DASL system, which assays 24000 cDNA probes. Differentially expressed candidate genes were then validated using quantitative PCR and immunohistochemistry. Results: Four hundred and sixteen differentially expressed genes were identified that clearly delineated between AS/SpA and control groups. Pathway analysis showed altered gene-expression in oxidoreductase activity, B-cell associated, matrix catabolic, and metabolic pathways. Altered «myogene» profiling was also identified. The inflammatory mediator, MMP3, was strongly upregulated (5-fold) in AS/SpA samples and the Wnt pathway inhibitors DKK3 (2.7-fold) and Kremen1 (1.5-fold) were downregulated. Conclusions: Altered expression profiling in SpA and AS samples demonstrates that disease pathogenesis is associated with both systemic inflammation as well as local tissue alterations that may underlie tissue damaging modelling and remodelling outcomes. This supports the hypothesis that initial systemic inflammation in spondyloarthropathies transfers to and persists in the local joint environment, and might subsequently mediate changes in genes directly involved in the destructive tissue remodelling.

Induction of a Th1 immune response and suppression of IgE via immunotherapy with a recombinant hybrid molecule encapsulated in liposome-protamine-DNA nanoparticles in a model of experimental allergy

Liposome-protamine-DNA nanoparticles (LPD) are safe, effective, and non-toxic adjuvants that induce Th1-like immune responses. We hypothesized that encapsulation of allergens into liposomes could be an appropriate option for immunotherapy. The present study evaluated the immunotherapeutic potential of a recombinant hybrid molecule (rHM) encapsulated in LPD nanoparticles in a murine model of Chenopodium album allergy. BALB/c mice were sensitized with the allergen in alum, and the immunotherapy procedure was performed by subcutaneous injections of LPD-rHM, rHM, or empty LPD at weekly intervals. Sensitized mice developed a Th2-biased immune response characterized by strong specific IgG1 and IgE production, IL-4, and the transcription factor GATA3 in spleen cell cultures. Treatment with the LPD-rHM resulted in a reduction in IgE and a marked increase in IgG2a. The LPD-rHM induced allergen-specific responses with relatively high interferon-gamma production, as well as expression of the transcription factor T-bet in stimulated splenocytes. In addition, lymphoproliferative responses were higher in the LPD-rHM-treated mice than in the other groups. Removal of the nanoparticles from the rHM resulted in a decrease in the allergen's immunogenicity. These results indicate that the rHM complexed with LPD nanoparticles has a marked suppressive effect on the allergic response and caused a shift toward a Th1 pathway.

A polymerase chain reaction-based method for isolating clones from a complimentary DNA library in sheep

The sheep (Ovis aries) is favored by many musculoskeletal tissue engineering groups as a large animal model because of its docile temperament and ease of husbandry. The size and weight of sheep are comparable to humans, which allows for the use of implants and fixation devices used in human clinical practice. The construction of a complimentary DNA (cDNA) library can capture the expression of genes in both a tissue- and time-specific manner. cDNA libraries have been a consistent source of gene discovery ever since the technology became commonplace more than three decades ago. Here, we describe the construction of a cDNA library using cells derived from sheep bones based on the pBluescript cDNA kit. Thirty clones were picked at random and sequenced. This led to the identification of a novel gene, C12orf29, which our initial experiments indicate is involved in skeletal biology. We also describe a polymerase chain reaction-based cDNA clone isolation method that allows the isolation of genes of interest from a cDNA library pool. The techniques outlined here can be applied in-house by smaller tissue engineering groups to generate tools for biomolecular research for large preclinical animal studies and highlights the power of standard cDNA library protocols to uncover novel genes.

Electrochemical detection of benzo(a)pyrene and related DNA damage using DNA/hemin/nafion-graphene biosensor

A novel electrochemical biosensor, DNA/hemin/nafion–graphene/GCE, was constructed for the analysis of the benzo(a)pyrene PAH, which can produce DNA damage induced by a benzo(a)pyrene (BaP) enzyme-catalytic product. This biosensor was assembled layer-by-layer, and was characterized with the use of cyclic voltammetry, electrochemical impedance spectroscopy (EIS) and atomic force microscopy. Ultimately, it was demonstrated that the hemin/nafion–graphene/GCE was a viable platform for the immobilization of DNA. This DNA biosensor was treated separately in benzo(a)pyrene, hydrogen peroxide (H2O2) and in their mixture, respectively, and differential pulse voltammetry (DPV) analysis showed that an oxidation peak was apparent after the electrode was immersed in H2O2. Such experiments indicated that in the presence of H2O2, hemin could mimic cytochrome P450 to metabolize benzo(a)pyrene, and a voltammogram of its metabolite was recorded. The DNA damage induced by this metabolite was also detected by electrochemical impedance and ultraviolet spectroscopy. Finally, a novel, indirect DPV analytical method for BaP in aqueous solution was developed based on the linear metabolite versus BaP concentration plot; this method provided a new, indirect, quantitative estimate of DNA damage.

VIP Barcoding: Composition vector-based software for rapid species identification based on DNA barcoding

Species identification based on short sequences of DNA markers, that is, DNA barcoding, has emerged as an integral part of modern taxonomy. However, software for the analysis of large and multilocus barcoding data sets is scarce. The Basic Local Alignment Search Tool (BLAST) is currently the fastest tool capable of handling large databases (e.g. >5000 sequences), but its accuracy is a concern and has been criticized for its local optimization. However, current more accurate software requires sequence alignment or complex calculations, which are time-consuming when dealing with large data sets during data preprocessing or during the search stage. Therefore, it is imperative to develop a practical program for both accurate and scalable species identification for DNA barcoding. In this context, we present VIP Barcoding: a user-friendly software in graphical user interface for rapid DNA barcoding. It adopts a hybrid, two-stage algorithm. First, an alignment-free composition vector (CV) method is utilized to reduce searching space by screening a reference database. The alignment-based K2P distance nearest-neighbour method is then employed to analyse the smaller data set generated in the first stage. In comparison with other software, we demonstrate that VIP Barcoding has (i) higher accuracy than Blastn and several alignment-free methods and (ii) higher scalability than alignment-based distance methods and character-based methods. These results suggest that this platform is able to deal with both large-scale and multilocus barcoding data with accuracy and can contribute to DNA barcoding for modern taxonomy. VIP Barcoding is free and available at http://msl.sls.cuhk.edu.hk/vipbarcoding/.

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.