Hybrid Quantum Classical Neural Network-Based Classification of Prenatal Ventricular Septal Defect from Ultrasound Images

Prenatal ventricular septal defect (VSD) is the second most common congenital heart defect-based (CHD) cardiac anomaly among growing fetus. Diagnosis of prenatal VSD CHD is clinically accomplished from ultrasound images, which is safe but is distorted by inherent speckle noise making the diagnosis a more challenging task. In this paper, we present a hybrid quantum classical neural network model (HQCNN) executed in IBM Aer simulator to recognize the prenatal VSD from fetal cardiac 2-dimensional ultrasound images. We trained the HQCNN model by varying number of qubits and number of shots of the parameterized quantum circuit. The proposed model trained with 2 qubit and executed for 1500 shots offered superior performance comparatively by yielding a high testing accuracy of about 95.8% by accurately classifying VSD CHD.