Interactive Volume Visualization of Microscopic Images Using Feature Space Reduction

2 This paper proposes a method for visualizing three-dimensional (3D) microscopic images using feature space reduction. Conventional visualization methods using multi-dimensional transfer function have difficulty in displaying feature space in high dimension and in setting transfer function. Our methods use principle component analysis (PCA) for feature space reduction to support XVHU¶V H[SORUDWLRQ RI WKH feature values and design of the transfer function. We applied this method to two-photon microscopic images and confirmed that neural structures such as soma, dendrite and apical dendrite were distinguishably visualized.