AI-Augmented Ultrasound Imaging Fused with Virtual Reality 3D Reconstruction for Early-Stage Tumor Detection and Clinical Decision Support

This research poster presents a low-cost, portable framework that combines AI-based ultrasound lesion detection with Virtual Reality (VR) volumetric visualization to support earlier tumor detection and clinical decision-making. The proposed pipeline integrates ultrasound acquisition, speckle denoising, CNN-based lesion detection and segmentation, AI-assisted segmentation, 3D volumetric reconstruction, and VR visualization. Simulation results indicate that AI-enhanced ultrasound can substantially improve sensitivity for early-stage, sub-10 mm lesions compared with conventional B-mode ultrasound, while the VR component provides an intuitive three-dimensional representation of tumor size and shape. The proposed approach is designed with accessibility and clinical practicality in mind, particularly for settings where advanced MRI/CT facilities may be limited. The work also identifies prospective validation using de-identified clinical ultrasound data and radiologist ground truth as an important next step.