Enhancing an enterprise data warehouse for research with data extracted using natural language processing

Objective

This study aims to develop a generalizable architecture for enhancing an enterprise data warehouse for research (EDW4R) with results from a natural language processing (NLP) model, which allows discrete data derived from clinical notes to be made broadly available for research use without need for NLP expertise. The study also quantifies the additional value that information extracted from clinical narratives brings to EDW4R.

Materials and methods

Clinical notes written during one month at an academic health center were used to evaluate the performance of an existing NLP model and to quantify its value added to the structured data. Manual review was utilized for performance analysis. The architecture for enhancing the EDW4R is described in detail to enable reproducibility.

Results

Two weeks were needed to enhance EDW4R with data from 250 million clinical notes. NLP generated 16 and 39% increase in data availability for two variables.

Discussion

Our architecture is highly generalizable to a new NLP model. The positive predictive value obtained by an independent team showed only slightly lower NLP performance than the values reported by the NLP developers. The NLP showed significant value added to data already available in structured format.

Conclusion

Given the value added by data extracted using NLP, it is important to enhance EDW4R with these data to enable research teams without NLP expertise to benefit from value added by NLP models.

Determination of Nitrate in Forages by Using Selective Ion Electrode: Collaborative Study

Each of 10 collaborating laboratories analyzed 4 blind duplicate pairs of forage samples for nitrate, by using a potentiometric method. Two forage controls and a 100 000 mg KNO3/L standard were also provided. Nitrate was extracted into an aqueous Al2(SO4)3 solution containing 70 mg KNO3/L and quantitated with a nitrate-selective electrode. Standards were prepared using extracting solution as diluent. Nitrate concentrations in forage samples ranged from <0.50 to 4.35% KNO3. Repeatability coefficients of variation (CV0) ranged from 1.74 to 3.61%, and reproducibility coefficients of variation (CV1) ranged from 6.92 to 7.66%. Mean recovery of a 0.55% KNO3 spike was 94.5%. The method has been adopted official first action.

Screening Nitrate in Forages with a Test Strip: Collaborative Study

A collaborative study was conducted for screening nitrate in forages with a commercially available test strip. The method involves extracting a finely ground sample with deionized water. The test strip is dipped in the sample extract. The color of the reaction zone on the test strip changes from white to pink or purple depending on the nitrate concentration in sample extract. The nitrate present in the extract is determined by comparing the color of the test strip to the color scale on the test strip container. Six blind quintuplicates of forage samples were analyzed by 20 collaborators. Nitrate concentrations in forage samples tested ranged from <1000 ppm nitrate to >10 000 ppm nitrate on dry matter basis. Each collaborator was asked to assign each sample to one of the 4 following nitrate concentration ranges: (1) <1000 ppm, (2) 1000 to 5000 ppm, (3)>5000 ppm to 10 000 ppm, and (4) >10 000 ppm. Nineteen of 20 collaborators reported results. Results from 2 laboratories were rejected as outliers by inspection and χ2 test. Sensitivity rates (p+) ranged from 0.965 to 0.998, with standard errors of 0.006 to 0.16. Specificity rates (p−) ranged from 0.991 to 0.997 for the 4 ranges, with standard errors of 0.003 to 0.006. False-positive rates (pf+) ranged from 0.006 to 0.046, with standard errors of 0.006 to 0.025. False-negative rates (pf−) ranged from 0.003 to 0.007, with standard errors of 0.003 to 0.006. Screening nitrate in forages with a test strip has been adopted first action by AOAC INTERNATIONAL.

pH Method for Determination of Cholinesterase in Whole Blood: Collaborative Study

A collaborative study of an indirect method for determination of cholinesterase activity was conducted with cattle blood. Ten collaborators analyzed 8 samples of varying degrees of cholinesterase inhibition. The sample is diluted with a buffer, and the pH change that results from the hydrolysis of acetylcholine is monitored. Repeatability relative standard deviations (RSDr) ranged from 6.6 to 8.6%, and reproducibility relative standard deviations (RSDR) ranged from 10.7 to 17.3% for mean activities of 0.60-0.40 ΔpH/h. All samples that were completely inhibited were detected. The method was adopted first action by AOAC International for determination of cholinesterase activity in whole bovine blood.

Pigmentary keratitis in pugs in the United Kingdom: prevalence and associated features

Background

Pigmentary keratitis (PK) is commonly recognised in Pugs, but its aetiology is not completely understood. The aim of this study was to determine the prevalence and associated features of PK in Pugs in the United Kingdom (UK).

Results

A total of 210 Pugs (420 eyes) were recruited from 12 UK dog shows and social events. The median age of Pugs recruited was 2.50 years (range 0.25–16.25 years). Pigmentary keratitis was detected in 369/420 (87.8%) eyes and in at least one eye 193/210 (91.9%) Pugs, of which 17/193 (8.8%) were affected unilaterally and 176/193 (91.2%) bilaterally. Pigmentary keratitis was typically mild to moderate (46.3 and 49.9% of eyes, respectively). Detection of PK was significantly associated with increased age (P = 0.002) and the presence of medial entropion of the lower eyelid (MELE) (P = 0.001). Severity of PK was significantly associated with the grade of MELE (P < 0.001). There was also a correlation between the presence of limbal pigment and PK (P = 0.036) that warrants further study.

Conclusions

This study estimated a high disease prevalence of PK in UK Pugs, and demonstrated significant associations with age and the presence of MELE. These associations, which have not been previously reported, offer an insight into the underlying pathophysiology of this condition in Pugs. The results encourage further population research, such as prospective longitudinal studies. These findings also support the development of clinical and breeding strategies based on the reduction of MELE and, possibly, limbal pigment.

Comparison of a needle-free high-pressure injection system with needle-tipped injection of intracavernosal alprostadil for erectile dysfunction

Patients identified from hospital records as using alprostadil injections for erectile dysfunction were invited to take part in this open crossover study. On alternate weeks eight patients were given intracavernosal needle injections and transdermal needle-free injection of alprostadil in a randomized order. Efficacy of injection and associated pain were assessed and compared for the two methods. Pain produced during injection was significantly greater with the needle-free system than with the needle-tipped injection whilst efficacy was significantly less. Bruising was reported in all except one patient following needle-free injection only. Patient ratings of the needle-free injector were significantly lower than ratings for needle-tipped alprostadil delivery and when asked to express a preference, every patient chose the needle-tipped injection over the needle-free device.

Independent component analysis: A flexible nonlinearity and decorrelating manifold approach

Independent component analysis (ICA) finds a linear transformation to variables that are maximally statistically independent. We examine ICA and algorithms for finding the best transformation from the point of view of maximizing the likelihood of the data. In particular, we discuss the way in which scaling of the unmixing matrix permits a "static" nonlinearity to adapt to various marginal densities. We demonstrate a new algorithm that uses generalized exponential functions to model the marginal densities and is able to separate densities with light tails. We characterize the manifold of decorrelating matrices and show that it lies along the ridges of high-likelihood unmixing matrices in the space of all unmixing matrices. We show how to find the optimum ICA matrix on the manifold of decorrelating matrices, and as an example we use the algorithm to find independent component basis vectors for an ensemble of portraits.

Separating spatially distributed response to stimulation from background. I. Optical imaging

We consider the problem of estimating a small stimulus-induced response to stimulation that is masked by a fluctuating background when measurements of the background in the absence of stimulation are available, as is common in optical imaging of the cortex and in many other experimental situations. Two related methods based on the Karhunen-Loève procedure are discussed. One seeks the function, an indicator function, that is most parallel to the response data and most orthogonal to the background data. The second removes the subspace spanned by the background from the response. Numerical investigations on simulated optical imaging data show that the first method is generally superior. Connections between the two methods and techniques for assessing the quality of the result are discussed.

Cancer of the uterine corpus after hormonal treatment for breast cancer

Among 45853 women in whom breast cancer was diagnosed after age forty-nine, from the series of the End Results Program of the National Cancer Institute, cancer of the uterine corpus subsequently developed in 203. The risk was greater among those women receiving hormones than in other treatment groups, and tended to rise with increasing interval from first treatment. One method of estimating an expected value indicated that the excess risk of corpus cancer in breast-cancer patients was restricted to those treated with hormones. Given the time period under study, it may be assumed that the hormones were primarily non-steroidal oestrogens.