Plaque progression and vulnerability are influenced by many risk factors. Our goal is to find simple methods to combine multiple risk factors for better plaque development predictions. A sample size of 374 intravascular ultrasound (IVUS) slices with matched follow-up was obtained from 9 patients (Mean age 59, 7 m) with informed consent obtained. 3D fluid-structure interaction models were constructed to obtain plaque stress/strain conditions. Four morphological and biomechanical factors (plaque burden (PB), cap thickness (CT), lipid percent (LP) and average plaque wall stress (PWS)) were chosen to predict plaque burden increase defined as PBI = (PB at follow-up) - (PB at baseline). For a given slice Si, the ground truth Y PBI is define as Y PBI (Si)=1 if PBI(Si)>0; Y PBI =0 if PBI(Si)≤0. For a single predictor W, a threshold value Wc was used to assign the binary prediction outcome: Y W (Si)=1 if W>Wc; Y W (Si)=0 if W≤Wc. Wc was chosen to get optimal agreement between Y W and Y PBI . To use multiple predictors (say, W1, W2, W3) to PBI, a new predictor Combo(W1,W2,W3) was created with its values defined as Combo(W1,W2,W3)=Y W1 +Y W2 +Y W3 , where Y W1 , Y W2 and Y W3 were evaluated the same way as before. Combo was then treated as a single predictor and a threshold value was determined to achieve best agreement with Y PBI . Table 1 summarizes the optimal thresholds and agreement rates for all 15 strategies. Agreement rate using PB alone was 57.5%. PWS was the best single predictor for PBI with agreement rate 62.6%. Combining CT and PWS achieved 66.5% agreement rate, 9% better over PB, which was also obtained by combining 4 risk factors. The method presented here could be used to combine predictors from different sources (stenosis, cap, lipid, inflammation, macrophage, hemorrhage, stress, strain, flow shear stress, FFR, smoking, diabetes, cholesterol, alcohol, hypertension, pro-rupture genes, etc.) to improve prediction accuracy and help decision-making in clinical practice.