Dataset Information

A deep learning approach to identify missing is-a relations in SNOMED CT.

ABSTRACT:

Objective

SNOMED CT is the largest clinical terminology worldwide. Quality assurance of SNOMED CT is of utmost importance to ensure that it provides accurate domain knowledge to various SNOMED CT-based applications. In this work, we introduce a deep learning-based approach to uncover missing is-a relations in SNOMED CT.

Materials and methods

Our focus is to identify missing is-a relations between concept-pairs exhibiting a containment pattern (ie, the set of words of one concept being a proper subset of that of the other concept). We use hierarchically related containment concept-pairs as positive instances and hierarchically unrelated containment concept-pairs as negative instances to train a model predicting whether an is-a relation exists between 2 concepts with containment pattern. The model is a binary classifier leveraging concept name features, hierarchical features, enriched lexical attribute features, and logical definition features. We introduce a cross-validation inspired approach to identify missing is-a relations among all hierarchically unrelated containment concept-pairs.

Results

We trained and applied our model on the Clinical finding subhierarchy of SNOMED CT (September 2019 US edition). Our model (based on the validation sets) achieved a precision of 0.8164, recall of 0.8397, and F1 score of 0.8279. Applying the model to predict actual missing is-a relations, we obtained a total of 1661 potential candidates. Domain experts performed evaluation on randomly selected 230 samples and verified that 192 (83.48%) are valid.

Conclusions

The results showed that our deep learning approach is effective in uncovering missing is-a relations between containment concept-pairs in SNOMED CT.

SUBMITTER: Abeysinghe R

PROVIDER: S-EPMC9933066 | biostudies-literature | 2023 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A deep learning approach to identify missing is-a relations in SNOMED CT.

Abeysinghe Rashmie R Zheng Fengbo F Bernstam Elmer V EV Shi Jay J Bodenreider Olivier O Cui Licong L

Journal of the American Medical Informatics Association : JAMIA 20230201 3

<h4>Objective</h4>SNOMED CT is the largest clinical terminology worldwide. Quality assurance of SNOMED CT is of utmost importance to ensure that it provides accurate domain knowledge to various SNOMED CT-based applications. In this work, we introduce a deep learning-based approach to uncover missing is-a relations in SNOMED CT.<h4>Materials and methods</h4>Our focus is to identify missing is-a relations between concept-pairs exhibiting a containment pattern (ie, the set of words of one concept b ...[more]

PMID: 36539234

Similar Datasets

Project description:ObjectiveChildhood hearing loss has well-known, lifelong consequences. Infection-related hearing loss disproportionately affects underserved communities yet can be prevented with early identification and treatment. This study evaluates the utility of machine learning in automating tympanogram classifications of the middle ear to facilitate layperson-guided tympanometry in resource-constrained communities.DesignDiagnostic performance of a hybrid deep learning model for classifying narrow-band tympanometry tracings was evaluated. Using 10-fold cross-validation, a machine learning model was trained and evaluated on 4810 pairs of tympanometry tracings acquired by an audiologist and layperson. The model was trained to classify tracings into types A (normal), B (effusion or perforation), and C (retraction), with the audiologist interpretation serving as reference standard. Tympanometry data were collected from 1635 children from October 10, 2017, to March 28, 2019, from two previous cluster-randomized hearing screening trials (NCT03309553, NCT03662256). Participants were school-aged children from an underserved population in rural Alaska with a high prevalence of infection-related hearing loss. Two-level classification performance statistics were calculated by treating type A as pass and types B and C as refer.ResultsFor layperson-acquired data, the machine-learning model achieved a sensitivity of 95.2% (93.3, 97.1), specificity of 92.3% (91.5, 93.1), and area under curve of 0.968 (0.955, 0.978). The model's sensitivity was greater than that of the tympanometer's built-in classifier [79.2% (75.5, 82.8)] and a decision tree based on clinically recommended normative values [56.9% (52.4, 61.3)]. For audiologist-acquired data, the model achieved a higher AUC of 0.987 (0.980, 0.993), had an equivalent sensitivity of 95.2 (93.3, 97.1), and a higher specificity of 97.7 (97.3, 98.2).ConclusionsMachine learning can detect middle ear disease with comparable performance to an audiologist using tympanograms acquired either by an audiologist or a layperson. Automated classification enables the use of layperson-guided tympanometry in hearing screening programs in rural and underserved communities, where early detection of treatable pathology in children is crucial to prevent the lifelong adverse effects of childhood hearing loss.

Project description:BackgroundData collected by an actigraphy device worn on the wrist or waist can provide objective measurements for studies related to physical activity; however, some data may contain intervals where values are missing. In previous studies, statistical methods have been applied to impute missing values on the basis of statistical assumptions. Deep learning algorithms, however, can learn features from the data without any such assumptions and may outperform previous approaches in imputation tasks.ObjectiveThe aim of this study was to impute missing values in data using a deep learning approach.MethodsTo develop an imputation model for missing values in accelerometer-based actigraphy data, a denoising convolutional autoencoder was adopted. We trained and tested our deep learning-based imputation model with the National Health and Nutrition Examination Survey data set and validated it with the external Korea National Health and Nutrition Examination Survey and the Korean Chronic Cerebrovascular Disease Oriented Biobank data sets which consist of daily records measuring activity counts. The partial root mean square error and partial mean absolute error of the imputed intervals (partial RMSE and partial MAE, respectively) were calculated using our deep learning-based imputation model (zero-inflated denoising convolutional autoencoder) as well as using other approaches (mean imputation, zero-inflated Poisson regression, and Bayesian regression).ResultsThe zero-inflated denoising convolutional autoencoder exhibited a partial RMSE of 839.3 counts and partial MAE of 431.1 counts, whereas mean imputation achieved a partial RMSE of 1053.2 counts and partial MAE of 545.4 counts, the zero-inflated Poisson regression model achieved a partial RMSE of 1255.6 counts and partial MAE of 508.6 counts, and Bayesian regression achieved a partial RMSE of 924.5 counts and partial MAE of 605.8 counts.ConclusionsOur deep learning-based imputation model performed better than the other methods when imputing missing values in actigraphy data.

Dataset Information

A deep learning approach to identify missing is-a relations in SNOMED CT.

Objective

Materials and methods

Results

Conclusions

Publications

A deep learning approach to identify missing is-a relations in SNOMED CT.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets