Autonomous detection, grading, and reporting of postoperative complications using natural language processing

Selby, Luke V.; Narain, Wazim; Russo, Ashley; Strong, Vivian E.; Stetson, Peter D.

doi:10.1016/j.surg.2018.05.008

Cited by 21 publications

(45 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…00043 useful information from unstructured narrative text. 4,5 NLP algorithms have been widely applied to medical research in recent years for various tasks, including extracting protein-protein interactions, 6 predicting gene-disease associations from biomedical literature databases, 7 improving the sensitivity of screening for suicide behaviors among pregnant women from electronic health record systems, 8 and correlating mammographic imaging features with pathologic findings. 9 Our group previously used an NLP algorithm to automatically parse breast pathologic reports in both English and Chinese.…”

Section: Introductionmentioning

confidence: 99%

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Deng

Yin

Bao

et al. 2019

JCO Clinical Cancer Informatics

View full text Add to dashboard Cite

PURPOSE Quantifying the risk of cancer associated with pathogenic mutations in germline cancer susceptibility genes—that is, penetrance—enables the personalization of preventive management strategies. Conducting a meta-analysis is the best way to obtain robust risk estimates. We have previously developed a natural language processing (NLP) –based abstract classifier which classifies abstracts as relevant to penetrance, prevalence of mutations, both, or neither. In this work, we evaluate the performance of this NLP-based procedure. MATERIALS AND METHODS We compared the semiautomated NLP-based procedure, which involves automated abstract classification and text mining, followed by human review of identified studies, with the traditional procedure that requires human review of all studies. Ten high-quality gene–cancer penetrance meta-analyses spanning 16 gene–cancer associations were used as the gold standard by which to evaluate the performance of our procedure. For each meta-analysis, we evaluated the number of abstracts that required human review (workload) and the ability to identify the studies that were included by the authors in their quantitative analysis (coverage). RESULTS Compared with the traditional procedure, the semiautomated NLP-based procedure led to a lower workload across all 10 meta-analyses, with an overall 84% reduction (2,774 abstracts v 16,941 abstracts) in the amount of human review required. Overall coverage was 93%—we are able to identify 132 of 142 studies—before reviewing references of identified studies. Reasons for the 10 missed studies included blank and poorly written abstracts. After reviewing references, nine of the previously missed studies were identified and coverage improved to 99% (141 of 142 studies). CONCLUSION We demonstrated that an NLP-based procedure can significantly reduce the review workload without compromising the ability to identify relevant studies. NLP algorithms have promising potential for reducing human efforts in the literature review process.

show abstract

Section: Introductionmentioning

confidence: 99%

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Deng

Yin

Bao

et al. 2019

JCO Clinical Cancer Informatics

View full text Add to dashboard Cite

show abstract

“…This small sample size may have affected the reliability of the results. The poorer performance may also be due to the different operating mechanisms of different AI models [14][15][16] and the differences between different types of venous thrombosis. [14][15][16][17][18] Thus, the results of these subgroups need to be interpreted with caution.…”

Section: Discussionmentioning

confidence: 99%

“…As shown in Figure 2 and Supplementary Figure S1, a high-risk bias was identified in 2 studies regarding flow and timing 23,25 and in 1 study regarding patient selection. 15 The results of the QUADAS-2 literature quality evaluation suggest that most studies had a moderate risk of bias and low applicability concerns. The Cochran's Q test and Higgins's inconsistency index (I2) test statistic results revealed significant heterogeneity between the overall training and test sets.…”

Section: Quality Of Evidence Heterogeneity and Risk Of Biasmentioning

confidence: 99%

“…There were 2443 cases of DVT in the training sets and 679 cases of DVT in the test sets. Four studies 15,18,20,25 analyzed DVT and PE separately. In terms of the types of venous thrombosis, 3 studies 16,17,26 15,20,23,26 were perioperative studies.…”

Section: Characteristics Of the Included Studiesmentioning

confidence: 99%

“…Four studies 15,18,20,25 analyzed DVT and PE separately. In terms of the types of venous thrombosis, 3 studies 16,17,26 15,20,23,26 were perioperative studies. In terms of the study focus regarding diagnosis/ prediction, 4 studies 16,20,22,23 focused on prediction, and 8 studies 14,15,17,18,21,[24][25][26] focused on diagnosis.…”

Section: Characteristics Of the Included Studiesmentioning

confidence: 99%

See 2 more Smart Citations

Prediction and Diagnosis of Venous Thromboembolism Using Artificial Intelligence Approaches: A Systematic Review and Meta-Analysis

Wang

Yuan

Ding

et al. 2021

Clin Appl Thromb Hemost

View full text Add to dashboard Cite

Venous thromboembolism (VTE) is a fatal disease and has become a burden on the global health system. Recent studies have suggested that artificial intelligence (AI) could be used to make a diagnosis and predict venous thrombosis more accurately. Thus, we performed a meta-analysis to better evaluate the performance of AI in the prediction and diagnosis of venous thrombosis. PubMed, Web of Science, and EMBASE were used to identify relevant studies. Of the 741 studies, 12 met the inclusion criteria and were included in the meta-analysis. Among them, 5 studies included a training set and test set, and 7 studies included only a training set. In the training set, the pooled sensitivity was 0.87 (95% CI 0.79-0.92), the pooled specificity was 0.95 (95% CI 0.89-0.97), and the area under the summary receiver operating characteristic (SROC) curve was 0.97 (95% CI 0.95-0.98). In the test set, the pooled sensitivity was 0.87 (95% CI 0.74-0.93), the pooled specificity was 0.96 (95% CI 0.79-0.99), and the area under the SROC curve was 0.98 (95% CI 0.97-0.99). The combined results remained significant in the subgroup analyzes, which included venous thrombosis type, AI type, model type (diagnosis/prediction), and whether the period was perioperative. In conclusion, AI may aid in the diagnosis and prediction of venous thrombosis, demonstrating high sensitivity, specificity and area under the SROC curve values. Thus, AI has important clinical value.

show abstract

Racial differences in venous thromboembolism: A surveillance program in Durham County, North Carolina

Saber

Adamski

Kuchibhatla

et al. 2022

Research and Practice in Thrombosis and Haemostasis

View full text Add to dashboard Cite

Background: Venous thromboembolism (VTE) affects approximately 1-2 individuals per 1000 annually and is associated with an increased risk for pulmonary hypertension, postthrombotic syndrome, and recurrent VTE. Objective: To determine risk factors, incidence, treatments, and outcomes of VTE through a 2-year surveillance program initiated in Durham County, North Carolina (population approximately 280,000 at time of study). Patients/Methods: We performed a retrospective analysis of data actively collected from three hospitals in Durham County during the surveillance period. Results: A total of 987 patients were diagnosed with VTE, for an annual rate of 1.76 per 1000 individuals. Hospital-associated VTE occurred in 167 hospitalized patients (16.9%) and 271 outpatients who were hospitalized within 90 days of diagnosis (27.5%). Annual incidence was 1.98 per 1000 Black individuals compared to 1.25 per 1000 White individuals (p < 0.0001), and Black individuals with VTE were youngerthan White individuals (p < 0.0001). Common risk factors included active cancer, prolonged immobility, and obesity, and approximately half were still taking anticoagulant therapy 1 year later. A total of 224 patients died by 1 year (28.5% of patients for whom outcomes could be confirmed), and Black patients were more likely to have recurrent VTE than White patients during the first 6 months following initial presentation (9.4% vs. 4.1%, p = 0.01). Conclusions:Ongoing surveillance provides an effective strategy to identify patients with VTE and monitor treatment and outcomes. We demonstrated that hospitalassociated VTE continues to be a major contributor to the burden of VTE and confirmed the higher incidence of VTE in Black compared to White individuals.

show abstract

Autonomous detection, grading, and reporting of postoperative complications using natural language processing

Abstract: Natural language processing can reliably detect the presence of postoperative venous thromboembolisms, and its use should be expanded for the detection of other conditions from narrative documentation.

Cited by 21 publications

References 26 publications

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Prediction and Diagnosis of Venous Thromboembolism Using Artificial Intelligence Approaches: A Systematic Review and Meta-Analysis

Racial differences in venous thromboembolism: A surveillance program in Durham County, North Carolina

Contact Info

Product

Resources

About