Spectral Analysis of Finger Photoplethysmographic Waveform Variability in a Model of Mild to Moderate Haemorrhage

• Published in: Journal of clinical monitoring and computing 2008-10, Vol.22 (5), p.343-353
• Main Authors: Middleton, Paul M., Chan, Gregory S. H., O’Lone, Emma, Steel, Elizabeth, Carroll, Rebecca, Celler, Branko G., Lovell, Nigel H.
• Format: Article
• Language: English
• Subjects: Anesthesiology; Artificial Intelligence; Critical Care Medicine; Diagnosis, Computer-Assisted - methods; Fingers - blood supply; Fingers - physiopathology; Health Sciences; Hemorrhage - diagnosis; Hemorrhage - physiopathology; Humans; Intensive; Medicine; Medicine & Public Health; Pattern Recognition, Automated - methods; Photoplethysmography - methods; Reproducibility of Results; Sensitivity and Specificity; Signal Processing, Computer-Assisted; Statistics for Life Sciences
• ISSN: 1387-1307; 1573-2614
• DOI: 10.1007/s10877-008-9140-1

Objective Slow fluctuations in cardiovascular signals such as heart rate variability (HRV) are believed to carry important clinical information. This study investigated whether frequency spectrum analysis of the finger photoplethysmographic waveform variability (PPGV) could characterize a hypovolaemic response by using a blood donation as a model of controlled mild to moderate haemorrhage. Methods This was a prospective, observational study carried out in a convenience sample of blood donors. Spectral analysis was performed on the finger infrared photoplethysmographic waveform and on the electrocardiogram- derived R–R intervals obtained from 43 healthy volunteers during blood donation. Spectral powers were calculated from low frequency (LF), mid frequency (MF) and high frequency (HF) bands of the spectrum of HRV and the coherence-weighted cross-spectrum of PPGV. Comparison was made between the four stages of blood donation: pre-donation (PRE), first half of donation (FIRST), second half of donation (SECOND) and post-donation (POST). Results A significant increase in the sum of the sympathetic-related MF and respiratory HF powers of finger PPGV (in mean-scaled units) was observed in SECOND and POST ( P < 0.01). The post-donation increase in this PPGV spectral measure occurred in 77% of the subjects, which was higher than the percentage of subjects experiencing a blood pressure drop (71% or below). Normalized LF power of HRV showed a significant rise in SECOND ( P < 0.01) but not in POST. Conclusions Spectral analysis of finger PPGV may provide valuable information in addition to vital sign measurements in characterizing a hypovolaemic response. Given the limitations of the current blood loss model, further studies are required to assess the usefulness of finger PPGV for early haemorrhage detection in the clinical setting.