1
|
Buchlak QD, Esmaili N, Bennett C, Farrokhi F. Natural Language Processing Applications in the Clinical Neurosciences: A Machine Learning Augmented Systematic Review. ACTA NEUROCHIRURGICA. SUPPLEMENT 2022; 134:277-289. [PMID: 34862552 DOI: 10.1007/978-3-030-85292-4_32] [Citation(s) in RCA: 6] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 10/19/2022]
Abstract
Natural language processing (NLP), a domain of artificial intelligence (AI) that models human language, has been used in medicine to automate diagnostics, detect adverse events, support decision making and predict clinical outcomes. However, applications to the clinical neurosciences appear to be limited. NLP has matured with the implementation of deep transformer models (e.g., XLNet, BERT, T5, and RoBERTa) and transfer learning. The objectives of this study were to (1) systematically review NLP applications in the clinical neurosciences, and (2) explore NLP analysis to facilitate literature synthesis, providing clear examples to demonstrate the potential capabilities of these technologies for a clinical audience. Our NLP analysis consisted of keyword identification, text summarization and document classification. A total of 48 articles met inclusion criteria. NLP has been applied in the clinical neurosciences to facilitate literature synthesis, data extraction, patient identification, automated clinical reporting and outcome prediction. The number of publications applying NLP has increased rapidly over the past five years. Document classifiers trained to differentiate included and excluded articles demonstrated moderate performance (XLNet AUC = 0.66, BERT AUC = 0.59, RoBERTa AUC = 0.62). The T5 transformer model generated acceptable abstract summaries. The application of NLP has the potential to enhance research and practice in the clinical neurosciences.
Collapse
Affiliation(s)
- Quinlan D Buchlak
- School of Medicine, The University of Notre Dame Australia, Sydney, NSW, Australia.
| | - Nazanin Esmaili
- School of Medicine, The University of Notre Dame Australia, Sydney, NSW, Australia
- Faculty of Engineering and Information Technology, University of Technology Sydney, Ultimo, NSW, Australia
| | - Christine Bennett
- School of Medicine, The University of Notre Dame Australia, Sydney, NSW, Australia
| | - Farrokh Farrokhi
- Neuroscience Institute, Virginia Mason Medical Center, Seattle, WA, USA
| |
Collapse
|
2
|
Lee YJ, Boyd AD, Li JJ, Gardeux V, Kenost C, Saner D, Li H, Abraham I, Krishnan JA, Lussier YA. COPD Hospitalization Risk Increased with Distinct Patterns of Multiple Systems Comorbidities Unveiled by Network Modeling. AMIA ... ANNUAL SYMPOSIUM PROCEEDINGS. AMIA SYMPOSIUM 2014; 2014:855-64. [PMID: 25954392 PMCID: PMC4419951] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Subscribe] [Scholar Register] [Indexed: 06/04/2023]
Abstract
Earlier studies on hospitalization risk are largely based on regression models. To our knowledge, network modeling of multiple comorbidities is novel and inherently enables multidimensional scoring and unbiased feature reduction. Network modeling was conducted using an independent validation design starting from 38,695 patients, 1,446,581 visits, and 430 distinct clinical facilities/hospitals. Odds ratios (OR) were calculated for every pair of comorbidity using patient counts and compared their tendency with hospitalization rates and ED visits. Network topology analyses were performed, defining significant comorbidity associations as having OR≥5 & False-Discovery-Rate≤10(-7). Four COPD-associated comorbidity sub-networks emerged, incorporating multiple clinical systems: (i) metabolic syndrome, (ii) substance abuse and mental disorder, (iii) pregnancy-associated conditions, and (iv) fall-related injury. The latter two have not been reported yet. Features prioritized from the network are predictive of hospitalizations in an independent set (p<0.004). Therefore, we suggest that network topology is a scalable and generalizable method predictive of hospitalization.
Collapse
Affiliation(s)
- Young Ji Lee
- Department of Medicine, University of Illinois at Chicago, Chicago, IL
| | - Andrew D Boyd
- Institute for Translational Health Informatics, University of Illinois at Chicago, Chicago, IL ; Departments of Biomedical and Health Information Sciences, University of Illinois at Chicago, Chicago, IL ; University of Illinois Hospital and Health Science System, University of Illinois at Chicago, Chicago, IL
| | - Jianrong John Li
- Department of Medicine, The University of Arizona, Tucson, AZ, USA
| | - Vincent Gardeux
- Department of Medicine, The University of Arizona, Tucson, AZ, USA
| | - Colleen Kenost
- Department of Medicine, The University of Arizona, Tucson, AZ, USA ; Biomedical Informatics Service Group, Arizona Health Science Center, The University of Arizona, Tucson, AZ, USA
| | - Don Saner
- Cancer Center, The University of Arizona, Tucson, AZ, USA ; Biomedical Informatics Service Group, Arizona Health Science Center, The University of Arizona, Tucson, AZ, USA
| | - Haiquan Li
- Department of Medicine, The University of Arizona, Tucson, AZ, USA
| | - Ivo Abraham
- Department of Pharmacy Practice and Science, The University of Arizona, Tucson, AZ, USA
| | - Jerry A Krishnan
- Department of Medicine, University of Illinois at Chicago, Chicago, IL ; University of Illinois Hospital and Health Science System, University of Illinois at Chicago, Chicago, IL
| | - Yves A Lussier
- Department of Medicine, The University of Arizona, Tucson, AZ, USA ; Cancer Center, The University of Arizona, Tucson, AZ, USA ; Biomedical Informatics Service Group, Arizona Health Science Center, The University of Arizona, Tucson, AZ, USA ; Interdisciplinary Program in Statistics, The University of Arizona, Tucson, AZ, USA
| |
Collapse
|