Author/Authors :
Mansourvar, Marjan university of malaya - Faculty of Computer Science and Information Technology - Department of Information System, Malaysia , Raj, Ram Gopal university of malaya - Faculty of Computer Science and Information Technology - Department of Artificial Intelligence, Malaysia , Ismail, Maizatul Akmar university of malaya - Faculty of Computer Science and Information Technology - Department of Information System, Malaysia , Kareem, Sameem Abdul university of malaya - Faculty of Computer Science and Information Technology - Department of Artificial Intelligence, Malaysia , Shanmugam, Saravanan Vellore Institute of Technology (VIT University) - School of Computing Science and Engineering, India , Wahid, Shahrom Forensic Science Society of Malaysia, Malaysia , Mahmud, Rohana university of malaya - Faculty of Computer Science and Information Technology - Department of Artificial Intelligence, Malaysia , Abdullah, Rukaini Hj. university of malaya - Faculty of Computer Science and Information Technology - Department of Artificial Intelligence, Malaysia , Nasaruddin, Fariza Hanum university of malaya - Faculty of Computer Science and Information Technology - Department of Information System, Malaysia , Idris, Norisma university of malaya - Faculty of Computer Science and Information Technology - Department of Artificial Intelligence, Malaysia
Abstract :
Bone age assessment (BAA) is often used to evaluate the growth status of children as part of the detection of hormonal problems and genetic disorders. The determination of skeletal maturity is done based on a radiological examination of the hand-wrist skeletal area.This paper introduces a novel approach for BAA that utilizes a histogram based comparison technique. This approach is executed as a web based system that uses an image repository and similarity measures based on content-based image retrieval. This study aims to overcome to the limitations of traditional methods utilized to estimate human age which were often imprecise.The system provides age prediction for hand and wrist x-ray images up till age of 18 years. The results of the system evaluation indicate this method as a reliable method for BAA with the error rates of -0.170625 years compared with BoneXpert system that have returned error rate of between ±0.46 to ±0.37 years.
Keywords :
Forensic Science , Bone Age Assessment (BAA) , ImageProcessing , Content Based Image Retrieval (CBIR) , Histogram.
