Predicting the Outcome of Patients With Subarachnoid Hemorrhage Using Machine Learning Techniques

被引:54
作者
de Toledo, Paula [1 ]
Rios, Pablo M. [1 ]
Ledezma, Agapito [1 ]
Sanchis, Araceli [1 ]
Alen, Jose F. [2 ]
Lagares, Alfonso [2 ]
机构
[1] Univ Carlos III Madrid, Control Learning & Syst Optimizat Grp, Madrid 28040, Spain
[2] Hosp Doce Octubre, Dept Neurosurg, Madrid 28041, Spain
来源
IEEE TRANSACTIONS ON INFORMATION TECHNOLOGY IN BIOMEDICINE | 2009年 / 13卷 / 05期
关键词
Data mining; knowledge discovery in databases; machine learning; prognosis; subarachnoid hemorrhage; LOGISTIC-REGRESSION; NEURAL-NETWORKS; HOSPITAL MORTALITY; CLASSIFICATION; MEDICINE; CARE; MANAGEMENT; PROGNOSIS; MODELS; SYSTEM;
D O I
10.1109/TITB.2009.2020434
中图分类号
TP [自动化技术、计算机技术];
学科分类号
080201 [机械制造及其自动化];
摘要
Background: Outcome prediction for subarachnoid hemorrhage (SAH) helps guide care and compare global management strategies. Logistic regression models for outcome prediction may be cumbersome to apply in clinical practice. Objective: To use machine learning techniques to build a model of outcome prediction that makes the knowledge discovered from the data explicit and communicable to domain experts. Material and methods: A derivation cohort (n = 441) of nonselected SAH cases was analyzed using different classification algorithms to generate decision trees and decision rules. Algorithms used were C4.5, fast decision tree learner, partial decision trees, repeated incremental pruning to produce error reduction, nearest neighbor with generalization, and ripple down rule learner. Outcome was dichotomized in favorable [Glasgow outcome scale (GOS) = I-II] and poor (GOS = III-V). An independent cohort (n = 193) was used for validation. An exploratory questionnaire was given to potential users (specialist doctors) to gather their opinion on the classifier and its usability in clinical routine. Results: The best classifier was obtained with the C4.5 algorithm. It uses only two attributes [World Federation of Neurological Surgeons (WFNS) and Fisher's scale] and leads to a simple decision tree. The accuracy of the classifier [area under the ROC curve (AUC) = 0.84; confidence interval (CI) = 0.80-0.88] is similar to that obtained by a logistic regression model (AUC = 0.86; CI = 0.83-0.89) derived from the same data and is considered better fit for clinical use.
引用
收藏
页码:794 / 801
页数:8
相关论文
共 46 条
[1]
Integrating classification trees with local logistic regression in Intensive Care prognosis [J].
Abu-Hanna, A ;
de Keizer, N .
ARTIFICIAL INTELLIGENCE IN MEDICINE, 2003, 29 (1-2) :5-23
[2]
[Anonymous], LANCET
[3]
[Anonymous], Data Mining Practical Machine Learning Tools and Techniques with Java
[4]
[Anonymous], 1989, CHOICE REV ONLINE, DOI DOI 10.5860/CHOICE.27-0936
[5]
[Anonymous], P 12 INT C MACH LEAR
[6]
[Anonymous], P 15 INT C MACH LEAR
[7]
Predictive data mining in clinical medicine: Current issues and guidelines [J].
Bellazzi, Riccardo ;
Zupan, Blaz .
INTERNATIONAL JOURNAL OF MEDICAL INFORMATICS, 2008, 77 (02) :81-97
[8]
What's wrong with hit ratio? [J].
Ben-David, Arie .
IEEE INTELLIGENT SYSTEMS, 2006, 21 (06) :68-70
[9]
An optimized experimental protocol based on neuro-evolutionary algorithms - Application to the classification of dyspeptic patients and to the prediction of the effectiveness of their treatment [J].
Buscema, M ;
Grossi, E ;
Intraligi, M ;
Garbagna, N ;
Andriulli, A ;
Breda, M .
ARTIFICIAL INTELLIGENCE IN MEDICINE, 2005, 34 (03) :279-305
[10]
Predicting hospital mortality for patients in the intensive care unit: A comparison of artificial neural networks with logistic regression models [J].
Clermont, G ;
Angus, DC ;
DiRusso, SM ;
Griffin, M ;
Linde-Zwirble, WT .
CRITICAL CARE MEDICINE, 2001, 29 (02) :291-296