Intelligent Virtual Breast Palpation Environment

Manual breast palpation is a primary examination technique for the detection of breast cancer. This skill requires rigorous training to be effective. In this study, we have developed an immersive virtual environment to assist learners in mastering breast medical palpation. This study used a Leap Motion controller to acquire the learner's three hand fingers: the middle, index, and ring finger movements. An immersive environment rendering is ensured with the Oculus Quest 2 head-mounted display. The proposed virtual environment included an intelligent agent who supervised the learner by utilizing gesture recognition by Mobile Net V3 a convolution neural network in transfer learning. This study considered gestures in three different patterns: circular, star, and sinusoidal shapes. Mobile net V3 classified and predicted images of gesture shapes with high confidence and an accuracy of 98%. It evaluated the correctness of palpation gestures in real-time, which makes it suitable for virtual reality use.