Share Email Print
cover

Proceedings Paper

A novel framework for the temporal analysis of bone mineral density in metastatic lesions using CT images of the femur
Author(s): Tom H. Knoop; Loes C. Derikx; Nico Verdonschot; Cornelis H. Slump
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In the progressive stages of cancer, metastatic lesions in often develop in the femur. The accompanying pain and risk of fracture dramatically affect the quality of life of the patient. Radiotherapy is often administered as palliative treatment to relieve pain and restore the bone around the lesion. It is thought to affect the bone mineralization of the treated region, but the quantitative relation between radiation dose and femur remineralization remains unclear. A new framework for the longitudinal analysis of CT-scans of patients receiving radiotherapy is presented to investigate this relationship. The implemented framework is capable of automatic calibration of Hounsfield Units to calcium equivalent values and the estimation of a prediction interval per scan. Other features of the framework are temporal registration of femurs using elastix, transformation of arbitrary Regions Of Interests (ROI), and extraction of metrics for analysis. Build in Matlab, the modular approach aids easy adaptation to the pertinent questions in the explorative phase of the research. For validation purposes, an in-vitro model consisting of a human cadaver femur with a milled hole in the intertrochanteric region was used, representing a femur with a metastatic lesion. The hole was incrementally stacked with plates of PMMA bone cement with variable radiopaqueness. Using a Kolmogorov-Smirnov (KS) test, changes in density distribution due to an increase of the calcium concentration could be discriminated. In a 21 cm3 ROI, changes in 8% of the volume from 888 ± 57mg • ml−1 to 1000 ± 80mg • ml−1 could be statistically proven using the proposed framework. In conclusion, the newly developed framework proved to be a useful and flexible tool for the analysis of longitudinal CT data.

Paper Details

Date Published: 20 March 2015
PDF: 11 pages
Proc. SPIE 9414, Medical Imaging 2015: Computer-Aided Diagnosis, 94143A (20 March 2015); doi: 10.1117/12.2081916
Show Author Affiliations
Tom H. Knoop, Univ. Twente (Netherlands)
Loes C. Derikx, Radboud Univ. Nijmegen Medical Ctr. (Netherlands)
KU Leuven (Belgium)
Nico Verdonschot, Radboud Univ. Nijmegen Medical Ctr. (Netherlands)
Univ. of Twente (Netherlands)
Cornelis H. Slump, Univ. Twente (Netherlands)


Published in SPIE Proceedings Vol. 9414:
Medical Imaging 2015: Computer-Aided Diagnosis
Lubomir M. Hadjiiski; Georgia D. Tourassi, Editor(s)

© SPIE. Terms of Use
Back to Top