Artificial Neural Networks (ANNs) are very popular as classification or regression mechanisms in medical decision support systems despite the fact that they are unstable predictors. This instability means that small changes in the training data used to build the model (i.e. train the ANN) may result in very different models. A central implication of this is that different sets of training data may produce models with very different generalisation accuracies. In this paper we show in detail how this can happen in a prediction system for use in In-Vitro Fertilisation. We argue that claims for the generalisation performance of ANNs used in such a scenario should only be based on k-fold cross validation tests. We also show how the accuracy of such a predictor can be improved by aggregating the output of several predictors.
We sought to identify mechanisms for chronic dysfunction in hibernating myocardium. Pigs were instrumented with a left anterior descending artery stenosis for 3 mo. Angiography demonstrated high-grade stenoses and hibernating myocardium with 1) severe anterior hypokinesis (P < 0.001 vs. shams), 2) reduced subendocardial perfusion [0.73 +/- 0.05 (SE) vs. 1.01 +/- 0.06 ml. min(-1). g(-1) in normal, P < 0.001], and 3) critically reduced adenosine flow (1.0 +/- 0.17 vs. 3.84 +/- 0.26 ml. min(-1). g(-1) in normal, P < 0.001). Histology did not reveal necrosis. Northern blot analysis of hibernating myocardium demonstrated regional downregulation in mRNAs for sarcoplasmic reticulum (SR) proteins phospholamban (0.76 +/- 0.08 vs. 1.07 +/- 0.06, P < 0.02) and SR Ca(2+)-ATPase (0.83 +/- 0.06 vs. 1.02 +/- 0.06, P < 0.05) with no change in calsequestrin (1.08 +/- 0.06 vs. 0.96 +/- 0.05, P = not significant). Heat shock protein (HSP)-70 mRNA was regionally induced in hibernating myocardium (2.4 +/- 0.3 vs. 1.0 +/- 0.11, P < 0.01). Directionally similar changes were confirmed by Western blot analysis of respective proteins. Our results indicate that hibernating myocardium exhibits a molecular phenotype that on a regional basis is similar to end-stage ischemic cardiomyopathy. This supports the hypothesis that SR dysfunction from reversible ischemia may be an early defect in the progression of left ventricular dysfunction.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.