There was a mistake in the Methods section on page 844 in the 1 December 1991 issue (1). The sentence should have read as follows. Patients whose pattern, when entered into the network, yielded an output of more than 0.55 were deemed to have had a myocardial infarction, and patients whose pattern resulted in an output of less than or equal to 0.55 were deemed not to have had a myocardial infarction.
Artificial Neural Networks. Ann Intern Med. ;116:94. doi: 10.7326/0003-4819-116-1-94_2
Download citation file:
Published: Ann Intern Med. 1992;116(1):94.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use