Diagnostic performance of photoplethysmography for early vascular compromise in a customisablein vitroflap model

Free flap reconstruction is a standard procedure in surgery, yet vascular thrombosis occurs in 3%-5% of cases. Early detection within the first few hours is essential for successful salvage, as the success rate of surgical re-exploration decreases over time. Since current clinical assessment remains subjective and depends on experience, objective and continuous monitoring is required. Photoplethysmography (PPG), which monitors blood volume changes non-invasively, is a candidate for this application. However, the influence of specific waveform changes during early haemodynamic shifts is not fully understood.

This study evaluated the diagnostic capacity of PPG for detecting early vascular changes using a custom silicone phantom. Synthetic vessels mimicking human arterial and venous mechanics were embedded at depths of 3, 9, 15, and 21 mm. A perfusion system simulated normal, early ischaemic, and early congested states. Signal quality was assessed using the signal-to-noise ratio (SNR), and only signals with SNR > 15 dB were used for morphological analysis. Over 50 parameters, including time, area, and slope, were extracted from the waveforms to identify those that characteristically respond to each haemodynamic state.

The custom-made free flap phantom was successfully validated. Signal quality assessments limited morphological evaluation to depths up to 15 mm. Analysis showed that Intensity and Area-based parameters were the most effective indicators at all depths. At shallow positions, Time-related features showed clear changes during ischaemia, while Slope and Second Derivative (SDPPG) features emerged as key indicators at depth. Red light was useful for superficial monitoring at 3 mm, whereas Infrared was necessary for assessing deeper states.

A custom phantom capable of replicating early haemodynamic compromise was developed. Identifying specific feature variations across depths provides a framework for objective, continuous monitoring. These findings suggest that combining multiple morphological features can improve the reliability of flap assessment.