Deep-learning-enabled-3D-multimodal-fusion-of-cone