Multimodal Palpation Sensing for Precise Prediction of Breast Lump Hardness and Size

PalpationDataset is a comprehensive multimodal dataset designed for precise characterization of breast lumps through tactile sensing. The dataset includes both controlled silicone phantom experiments and semi-realistic ex-vivo porcine tissue samples, capturing a wide range of lump properties and palpation conditions. Equipment and Materials A custom palpation system was used, integrating: Six-axis force sensor (2000 Hz) Inertial Measurement Unit (IMU) (500 Hz) Finger-shaped probe tip (10 mm diameter) Silicone Phantoms Lump diameters: 7.5 mm, 12.5 mm, 25 mm Lump stiffness: Shore 00-10, Shore 00-30, Shore A-30 Phantom base stiffness: Shore 000-34 Ecoflex Gel2, 1:1 Gel2:00-10 mixture Base volumes: 128 ml, 283 ml Embedding depths: shallow (3–5 mm), deep (8–12 mm) Porcine Tissue Ethically sourced ex-vivo porcine breast tissue Simulates heterogeneous biomechanical properties of human breast tissue