Cardiovascular disease (CVD) is still a primary reason for death worldwide emphasizing the need to develop effective and efficient automated diagnostic models. In this study, a Quantum Machine Learning based Framework (QuMLF) is introduced for Heart Disease (HD) prediction which leverages the Quantum Support Vector Classifier (QSVC), Genetic Algorithm (GA) for feature selection, and SMOTE (Synthetic Minority Over-sampling Technique) technique to manage imbalanced dataset. The dataset employed for the study is the Statlog HD dataset taken from the UCI repository. Our results have shown significant improvement in accuracy, specifically when QSVC with GA was employed in the balanced Statlog dataset, The QuMLF accuracy surpasses the VQC_S(VQC applied on selected features) model by 4% and the QSVC_S(QSVC applied on selected features) model by 8%. This study also highlights the comparative analysis of the quantum algorithms for HD prediction with and without the feature selection technique. The proposed framework not only finds the relevant features but also enhances the predictive accuracy of HD classification.