In this paper, a novel method of concealment of facial information was proposed to protect the patient’s privacy. It consists with two steps: the construction of a standard brain template and substitution of the template to a new brain image data. Fifteen medical brain images are used to build the standard brain template. Registration and normalization procedures are applied to construct the standard brain templates by using B-Spline registration. Segmentation is used to substitute the original image for the template image and brain extraction. Canny-edge Level set segmentation algorithm is applied to segment a scalp-skull part from a whole brain volume data. New medical brain images are substituted for the standard template to construct the anonymous face models. In order to verify the validity of the anonymous face model, volumes of extracted brains from an original test brain and its anonymous brain were compared. The result shows that the actual brain part from the anonymous brain is exactly same as one which comes from an original medical brain image.
Du, Ruoyu and Jong Lee, Hyo
"Identity Concealment of Brain Images by Masking,"
Applied Mathematics & Information Sciences: Vol. 07
, Article 34.
Available at: https://dc.naturalspublishing.com/amis/vol07/iss4/34