作者: Meindert Niemeijer , Bram van Ginneken , Casper A. Maas , Frederik J. A. Beek , Max A. Viergever
DOI: 10.1117/12.480163
关键词: Computer-aided diagnosis 、 Region of interest 、 Active shape model 、 Artificial intelligence 、 Middle phalanx 、 Computer vision 、 Correlation 、 Computer science 、 Skeletal maturity 、 Bone age 、 Radiography
摘要: The skeletal maturity of children is usually assessed from a standard radiograph the left hand and wrist. An established clinical method to determine Tanner-Whitehouse (TW2) method. This divides development into several stages (labelled A, B, ...,I). We are developing an automated system based on this In work we focus assigning stage one region interest (ROI), middle phalanx third finger. classify each ROI as follows. A number ROIs which have been assigned certain by radiologist used construct mean image for that stage. For new input ROI, landmarks detected using Active Shape Model. These align images with image. Subsequently correlation between transformed calculated. can be highest directly, or values features in classifier. was tested 71 cases ranging E I. staged correctly 73.2% all 97.2% incorrectly error not more than