OBJECTIVE Cavitation in the mechanical heart valve (MHV) was first detected in Edwards-Duromedics (ED) clinical explants. Early studies indicated that the pitted surface of the valve leaflet was due to cavitation phenomena occurring during valve closing. Cavitation is seen as the transient appearance of bubbles on the MHV surface on valve closure. The cavitation bubbles occur due to abrupt pressure changes in the vicinity of the valve on valve closing. Hence, analysis of the recorded field pressure can provide useful information relating to cavitation potential. In the present study, MHV cavitation potential was evaluated by counting bubble appearance probability and measuring bubble-size using an image-processing method. A simple and reliable technique using wavelet packet analysis (WPA) to evaluate cavitation potential was also investigated. METHODS A single-valve, pneumatic-driven burst tester system with adjustable pressure control unit, was used to simulate the closing process in the heart mitral valve at three driving pressures: 200, 500 and 1,000 mmHg, using three valve models. A triggering and imaging system was developed within the burst tester system to capture images of cavitation bubbles at predetermined time delays on valve closing. Transient pressure signals were recorded on both sides of the MHV occluder, using a high-frequency piezoelectric pressure transducer and a physiological pressure transducer. The pictures recorded were analyzed using image processing software to determine bubble appearance probability and bubble size. WPA was applied to the transient closing pressure signals to evaluate cavitation potential. RESULTS Cavitation intensity index (Cii) and bubble size-based cavitation index (BS-Ci) were measured by analyzing images captured at different time delays on valve closing at different driving pressures. WPA was used to analyze transient pressure signals at the inflow side of the MHV occluder at valve closure to calculate the WPA-based cavitation index (WPA-Ci). The three methods showed a similar trend for cavitation potential in the valves tested. CONCLUSIONS In the present study, two new approaches to evaluate MHV cavitation were investigated, namely WPA and BS-Ci. The results obtained produced a similar trend to that seen with an earlier method based on counting the probability that cavitation bubbles occur. As cavitation is primarily a function of the transient pressure within the vicinity of the closing valve occluder, the WPA method can be an effective method for future investigation of cavitation potential of mechanical heart valves.