Recent studies have shown that accurate detection of heart rhythm irregularities with commercial devices that measure arterial blood volume changes by means of photoplethysmography (PPG) is now possible. Due to limited battery life, these measurements are usually intermittent. Continuous and prolonged recordings increase the chance of detecting episodes of arrhythmia that may otherwise be missed in intermittent recordings such in episodes of paroxysmal atrial fibrillation (Afib). While building a robust PPG-based Afib detector can greatly benefit from annotated training data, such data may not be readily available as it may be difficult to acquire. The abundance of SpO2 data in hospital settings may prove useful as a proxy for the required training dataset, provided they reflect the same pathophysiological dynamics as reflected by the target device. In this work, continuous blood volume pulse (BVP) recordings (13.5h±8.5h) obtained from two different PPG modalities, fingertip pulse-oximetry and radial photoplethymography, are collected, annotated, and analysed in 13 patients with ischemic stroke with a primary goal to determine if both modalities reflect the same physiological dynamics despite the different locations where they are placed and secondary goal to develop an accurate and robust continuous-analysis based Afib detector. Pulse rate variability indices, heart rate, and sample entropy were estimated in consecutive 30s-long segments which had good BVP signal quality. Complexity and pulse rate variability in both modalities were compared for agreement. Pulse rates estimated in the temporal and frequency domain were compared for agreement with heart rates. Pulse rates, measured from both BVP modalities and heart rates were highly correlated and matched within ± 5bpm (95% CI). Sample entropy and pulse rate variability metrics were statistically comparable between modalities. These results demonstrate that pulse oximetry and wrist-photoplethymography reflect the same heart rhythm dynamics, as measured with pulse rate variability, despite the difference in measuring locations.