Big data driven co-occurring evidence discovery in chronic obstructive pulmonary disease patients

Abstract

Background: Chronic Obstructive Pulmonary Disease (COPD) is a chronic lung disease that affects airflow to the lungs. Discovering the co-occurrence of COPD with other diseases, symptoms, and medications is invaluable to medical staff. Building co-occurrence indexes and finding causal relationships with COPD can be difficult because often times disease prevalence within a population influences results. A method which can better separate occurrence within COPD patients from population prevalence would be desirable. Large hospital systems may potentially have tens of millions of patient records spanning decades of collection and a big data approach that is scalable is desirable. The presented method, Co-Occurring Evidence Discovery (COED), presents a methodology and framework to address these issues.Methods: Natural Language Processing methods are used to examine 64,371 deidentified clinical notes and discover associations between COPD and medical terms. Apache cTAKES is leveraged to annotate and structure clinical notes. Several extensions to cTAKES have been written to parallelize the annotation of large sets of clinical notes. A co-occurrence score is presented which can penalize scores based on term prevalence, as well as a baseline method traditionally used for finding co-occurrence. These scoring systems are implemented using Apache Spark. Dictionaries of ground truth terms for diseases, medications, and symptoms have been created using clinical domain knowledge. COED and baseline methods are compared using precision, recall, and F1 score. Results:The highest scoring diseases using COED are lung and respiratory diseases. In contrast, baseline methods for co-occurrence rank diseases with high population prevalence highest. Medications and symptoms evaluated with COED share similar results. When evaluated against ground truth dictionaries, the maximum improvements in recall for symptoms, diseases, and medications were 0.212, 0.130, and 0.174. The maximum improvements in precision for symptoms, diseases, and medications were 0.303, 0.333, and 0.180. Median increase in F1 score for symptoms, diseases, and medications were 38.1%, 23.0%, and 17.1%. A paired t-test was performed and F1 score increases were found to be statistically significant, where p < 0.01. Conclusion:Penalizing terms which are highly frequent in the corpus results in better precision and recall performance. Penalizing frequently occurring terms gives a better picture of the diseases, symptoms, and medications co-occurring with COPD. Using a mathematical and computational approach rather than purely expert driven approach, large dictionaries of COPD related terms can be assembled in a short amount of time.

Natural Language Processing of Clinical Notes on Chronic Diseases: Systematic Review

“…IE from clinical notes based on NLP was also used to (1) screen computed tomography reports for invasive pulmonary mold [103], (2) discover the co-occurrences of chronic obstructive pulmonary disease with other medical terms [104], (3) quantify the relationship between aggregated preoperative risk factors and cataract surgery complications [105], (4) detect patients with multiple sclerosis from the clinical notes prior to the initial recognition by their health care providers [106], and (5) identify patients on dialysis in the Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II) publicly available dataset [107].…”

Big data as a value generator in decision support systems: a literature review

“…The last group of articles classified in this study was that with those papers that developed or used some information system as a technology application. Baechle, Agarwal & Zhu (2017) described the benefits of improved accuracy of services provided over the short term by applying large dictionaries of chronic obstructive pulmonary disease–related terms. Gillingham (2019b) presents through the principles of algorithmic responsibility, and combination with ethical codes of social work, the possibility for social workers to identify when wrong or biased recommendations are made by the deployed DSS and thereby develop defense strategies for social service users.…”

Machine Learning to Assist in Pneumonia Decision Making: A Systematic Review of the Literature

“…Exemplos de trabalhos que se destacam com maior nota nos CQ neste quesito incluem [Naydenova et al 2015], [DeLisle et al 2013 e [Liao et al 2020]. Outros trabalhos que se destacaram nas demais categorias com maior nota nos índices de qualidade foram: [Baechle et al 2017], [Wu et al 2014] e [Caruana et al 2015].…”

“…Para trabalhos focados nas categorias de Mortalidade e Complicações e Tempo de Internação e Readmissão, existe uma divisão dos resultados entre Tarefas de Classificação, em que os autores focam apenas na sinalização desses riscos à equipe médica [Wu et al 2014] [Shimizu et al 2019] [Lai et al 2018] e Regressão, para estimar a probabilidade de tais riscos ocorrerem [Caruana et al 2015] [Villiers et al 2018]. No que se refere à identificação de Atributos ou Biomarcadores que influenciam no tratamento de pneumonia, autores utilizaram unicamente técnicas de Agrupamento e Regras de Associação, incluindo entre os dados analisados scores médicos e comparando ocorrências de atributos com Diagnósticos [Baechle et al 2017] [Lin et al 2010] [Ubaid et al 2010].…”