Information Technology Reference
In-Depth Information
Computer Aided Detection of Bone
Metastases in the Thoracolumbar Spine
Jianhua Yao, Joseph E. Burns and Ronald M. Summers
Abstract Computer-aided detection (CAD) techniques and algorithms for radiologic
applications are rapidly growing in scope and sophistication. One important appli-
cation of CAD techniques in medicine is in the detection and assessment of metastatic
disease to the bone. Bonemetastases affect approximately 400,000 patients per year in
the United States. Early detection of bone metastases is important clinically, as the
prognosis can change and the treatment regimen can at that point be altered from one
of curative therapy to one of palliative treatment. Both lytic and sclerotic metastatic
disease can act to biomechanicallyweaken the bone, and potentially lead to pathologic
fractures. This chapter presents a framework for computer-aided detection of lytic and
sclerotic metastatic lesions in the thoracolumbar spine using computed tomography
(CT). State-of-art techniques are described in detail in each module of the framework.
Thorough validation experiments are designed and results are presented. We also
discuss the clinical signi
cance and limitation of the CAD system.
1 Background of Bone Metastases
Bone metastases are a frequent occurrence of cancer, affecting approximately
400,000 patients per year in the United States [
1
]. In advanced breast and prostate
cancer alone, bone metastases are seen in up to 70 % of patients, and in patients with
carcinoma in the lung, colon, stomach, bladder, uterus, rectum, thyroid or kidney,
