Cardiovascular disease is the leading cause of death in Western civilization. In this pilot study we evaluated a new method for the diagnosis of myocardial infarction and heart failure by determining the typical fingerprint in the infrared (IR) spectrum of 1 μL of a dried patient serum sample by Fourier transform IR spectroscopy. For classification, cluster analysis and artificial neural networks (ANN) were applied. In this study 567 subjects were enrolled, comprising 225 controls (Co) and 342 patients with myocardial infarction (MI) (n = 157) and heart failure (HF) (n = 185). By applying artificial neural network algorithms, the following sensitivities and specificities of the same spectra were determined: MI versus Co (98%, 97%), HF versus Co (98%, 100%), MI versus HF (100%, 100%), and MI plus HF versus Co (100%, 100%). Based on our data, mid-IR spectroscopy appears to be a promising new method to diagnose heart diseases from serum samples. Artificial neural network algorithms proved to be superior to cluster analysis for correct prediction.

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Log in to access OSA Member Subscription

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Log in to access OSA Member Subscription