Predicting mental conditions based on "history of present illness" in psychiatric notes with deep neural networks

被引:47
作者
Tung Tran [1 ]
Kavuluru, Ramakanth [1 ,2 ]
机构
[1] Univ Kentucky, Dept Comp Sci, 329 Rose St, Lexington, KY 40506 USA
[2] Univ Kentucky, Dept Internal Med, Div Biomed Informat, 725 Rose St, Lexington, KY 40536 USA
关键词
Psychiatric condition prediction; Multi-label text classification; Convolutional and recurrent neural networks; Hierarchical attention networks; RECORDS;
D O I
10.1016/j.jbi.2017.06.010
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Background: Applications of natural language processing to mental health notes are not common given the sensitive nature of the associated narratives. The CEGS N-GRID 2016 Shared Task in Clinical Natural Language Processing (NLP) changed this scenario by providing the first set of neuropsychiatric notes to participants. This study summarizes our efforts and results in proposing a novel data use case for this dataset as part of the third track in this shared task. Objective: We explore the feasibility and effectiveness of predicting a set of common mental conditions a patient has based on the short textual description of patient's history of present illness typically occurring in the beginning of a psychiatric initial evaluation note. Materials and methods: We clean and process the 1000 records made available through the N-GRID clinical NLP task into a key-value dictionary and build a dataset of 986 examples for which there is a narrative for history of present illness as well as Yes/No responses with regards to presence of specific mental conditions. We propose two independent deep neural network models: one based on convolutional neural networks (CNN) and another based on recurrent neural networks with hierarchical attention (ReHAN), the latter of which allows for interpretation of model decisions. We conduct experiments to compare these methods to each other and to baselines based on linear models and named entity recognition (NER). Results: Our CNN model with optimized thresholding of output probability estimates achieves best overall mean micro-F score of 63.144% for 11 common mental conditions with statistically significant gains (p < 0.05) over all other models. The ReHAN model with interpretable attention mechanism scored 61.904% mean micro-F1 score. Both models' improvements over baseline models (support vector machines and NER) are statistically significant. The ReHAN model additionally aids in interpretation of the results by surfacing important words and sentences that lead to a particular prediction for each instance. Conclusions: Although the history of present illness is a short text segment averaging 300 words, it is a good predictor for a few conditions such as anxiety, depression, panic disorder, and attention deficit hyperactivity disorder. Proposed CNN and RNN models outperform baseline approaches and complement each other when evaluating on a per-label basis. (C) 2017 Published by Elsevier Inc.
引用
收藏
页码:S138 / S148
页数:11
相关论文
共 42 条
[1]  
Abadi Martin, 2015, TENSORFLOW LARGE SCA
[2]   Text mining applications in psychiatry: a systematic literature review [J].
Abbe, Adeline ;
Grouin, Cyril ;
Zweigenbaum, Pierre ;
Falissard, Bruno .
INTERNATIONAL JOURNAL OF METHODS IN PSYCHIATRIC RESEARCH, 2016, 25 (02) :86-100
[3]  
[Anonymous], UN MED LANG SYST REF
[4]   An overview of MetaMap: historical perspective and recent advances [J].
Aronson, Alan R. ;
Lang, Francois-Michel .
JOURNAL OF THE AMERICAN MEDICAL INFORMATICS ASSOCIATION, 2010, 17 (03) :229-236
[5]  
Bahdanau D, 2016, Arxiv, DOI [arXiv:1409.0473, DOI 10.48550/ARXIV.1409.0473]
[6]  
Bengio Y, 2001, ADV NEUR IN, V13, P932
[7]  
Benton A, 2017, PROCEEDINGS OF THE 1, P152, DOI DOI 10.18653/V1/E17-1015
[8]  
Bergstra J, 2010, P PYTH SCI COMP C SC, V4, P1, DOI [10.25080/Majora-92bf1922-003, DOI 10.25080/MAJORA-92BF1922-003]
[9]   Multitask learning [J].
Caruana, R .
MACHINE LEARNING, 1997, 28 (01) :41-75
[10]  
Collobert J., 2008, P 25 ICML, P160, DOI [10.1145/1390156.1390177, DOI 10.1145/1390156.1390177]