Neural networks for mining the associations between diseases and symptoms in clinical notes

Abstract

There are challenges for analyzing the narrative clinical notes in Electronic Health Records (EHRs) because of their unstructured nature. Mining the associations between the clinical concepts within the clinical notes can support physicians in making decisions, and provide researchers evidence about disease development and treatment. In this paper, in order to model and analyze disease and symptom relationships in the clinical notes, we present a concept association mining framework that is based on word embedding learned through neural networks. The approach is tested using 154,738 clinical notes from 500 patients, which are extracted from the Indiana University Health's Electronic Health Records system. All patients are diagnosed with more than one type of disease. The results show that this concept association mining framework can identify related diseases and symptoms. We also propose a method to visualize a patients' diseases and related symptoms in chronological order. This visualization can provide physicians an overview of the medical history of a patient and support decision making. The presented approach can also be expanded to analyze the associations of other clinical concepts, such as social history, family history, medications, etc.

DSR: A Collection for the Evaluation of Graded Disease-Symptom Relations

“…Here, we first describe two recently proposed methods [10,14] for the extraction of disease-symptom relations as our baselines. Afterwards, we describe limitations of the method described in [14] and propose an adapted version in which the limitations are addressed.…”

“…Embedding: Proposed by Shah et al [10], the method is based on the cosine similarity of the vector representations of a disease and a symptom. We first apply MetaMap [1], a tool for the identification of medical concepts within a given text, to the full text of all PMC articles to substitute the identified diseases/symptoms by their unique names.…”

“…We first apply MetaMap [1], a tool for the identification of medical concepts within a given text, to the full text of all PMC articles to substitute the identified diseases/symptoms by their unique names. Then, we train a word2vec model [8] with 300 dimensions and a window size of 15, following the parameter setting in [10]. Using the word embedding, the disease-symptom relation is defined as λ(x, s) = cos(e x , e s ), where e refers to the vector representation of a word.…”

“…For example, abdominal pain and nausea are both symptoms of an appendicitis, but while abdominal pain is a key differentiating factor, nausea does only little to distinguish appendicitis from other diseases of the digestive system. While several disease-symptom extraction methods have been proposed that retrieve a ranked list of symptoms for a disease [14,10,7,13], no collection is available to systematically evaluate the performance of such methods [11]. While these method are extensively used in downstream tasks, e.g., to increase the accuracy of computer-assisted medical diagnosis [9], their effectiveness for disease-symptom extraction remains unclear.…”

“…As baselines, we implement two methods from previous studies to compute graded disease-symptom relations: In the first method [10], the relation is the cosine similarity between the word vectors of a disease and a symptom, taken from a word embedding model. In the second method [14], the relation between a disease and symptom is calculated based on their co-occurrence in the MeSH-keywords 3 of medical articles.…”