Hemolysis, the breakdown of red blood cells (RBCs), releases hemoglobin (Hb) into plasma, impairing gas exchange and causing oxidative stress. Hb converted to methemoglobin (mHb), a non-oxygen-binding derivative, further worsens the clinical condition. Current diagnostic methods are invasive or lack real-time capabilities for accurately assessing degree and phases of hemolysis. This study introduces a novel approach utilizing photoacoustics (PA). A low-cost, laser diode based PA system with 447, 638, and 905 nm illumination was developed to investigate the distinguishability of lysed blood at different phases RBCs in Hb solution (intermediate phase) and RBCs in mHb solution (final phase) across various equivalent-hematocrit (eqHCT) levels and degrees of lysis. Later, a two-step artificial neural network model accurately classified the hemolysis phase (sensitivity = 1.00 and specificity = 1.00) and estimated eqHCT and lysis degree with 1.14% and 3.17% error, respectively, for the intermediate phase and 0.67% and 1.34% error for the final phase.
Banerjee et al. (Sun,) studied this question.