In the present study an existing, improved in-house finite element head model is used to reconstruct 70 well-documented real-world head trauma cases for the development of robust model-based brain injury criteria under Visual Performance Solution (VPS, ESI group). A combined strain and strain rate parameter, named the brain injury score (BIS), has been proposed to predict AIS2+ brain injury accurately based on statistical analysis. Further head kinematic-based Head Injury Criteria (HIC), Brain Injury Criteria (BrIC), Universal Brain Injury Criterion (UBrIC) and Convolution of Impulse response for Brain Injury Criterion (CIBIC) are computed and tissue- level injury criteria maximum principal strain, von Mises stress and brain pressure are extracted from the simulation. The injury risk probability for each case was computed for different parameters and the delta percentage risk was calculated for comparison between these injury criteria. In addition six real-world head trauma cases were reconstructed to validate the prediction capability of BIS, HIC, BrIC, UBrIC and CIBIC. It was concluded that BIS predicts the AIS2+ brain injury more accurately when compared to the others. The HIC and BrIC typically over predict the brain injury risk.