How artificial intelligence tools can be used to assess individual patient risk in cardiovascular disease: problems with the current methods

Table 1 Examples of artificial neural networks analyses in the cardiovascular field.

Diagnosis
	Year	No. Pts	Disease	Variables	Results
Selker	1995	3453	Ischemia	Clinical indicators	ANN superior vs LogR
Ellenius et al	1997	88	MI	Biochemical variables	ANNs give added value
Baxt et al	2002	2204	MI	History, clinical, biochemical EGC	High sensitivity (95%) and specificity (96%)
Prognosis
Baldassarre et al	2004	949	CV event	biochemical, carotid US clinical indicators	ANN superior vs LDA
Voss et al	2002	5159	CV event	Clinical, biochemical indicators	ANN superior vs LogR
Bigi et al	2005	496	Outcome after MI	Clinical, exercise ECG and stress echo	ANN superior vs LDA

MI: Miocardial infarction; CV: cardiovascular; ANN Artificial neural networks; LogR: logistic regression; LDA: linear discriminant analysis

ISSN: 1471-2261