Quantitative assessment in vivo of trabecular bone structure
Osteoporosis is characterised by a change of the trabecular structure of cancellous bone, which may be more important for strength than the reduction of cortical bone. Until now, the 3D microstructure could only be visualised in ex vivo specimens. This project aims at studying this structure in vivo using clinical multislice CT, dual energy CT and dental cone-beam CT. 3D images are acquired and analyzed with mathematical morphology methods, resulting in measures of trabecular structure. New image analysis methods to better describe the anisotropy of bone are being developed. These methods will be compared and validated in specimens using micro-CT as the reference method and then tested in vivo in clinical studies.
Örjan Smedby , Prof
|Project leader, co-supervisor||Radiology, IMH|
Magnus Borga , Docent
|Main supervisor||Medical informatics, IMT|
Torkel Brismar , Med Dr
|Co-supervisor||Radiology, KI Huddinge|
Göran Toss , Docent
|Clinical collaborator||Orthopaedic surgery, IKE|
Gudrun Alm Carlsson , Prof
|Senior researcher||Radiation physics, IMH|
Rodrigo Moreno , PhD
|Postdoctoral researcher||Radiology, IMH|
|Doctoral student||Dental radiology, LIO|
- Former Staff:
- Project Description:
- Moreno R., Borga M., Smedby Ö. "Generalizing the mean intercept length tensor for gray-level images". Medical Physics, Vol. 39, No. 7, pp. 4599-4612, 2012. DOI:10.1118/1.4730502
- Klintström E., Moreno R., Brismar T., Smedby Ö. "Three-dimensional image processing for measuring trabecular bone structure parameters". EADMFR 2012.
- Moreno R., Borga M., Smedby Ö. "Evaluation of the plate-rod model assumption of trabecular bone". ISBI 2012, pp. 470-473. DOI:10.1109/ISBI.2012.6235586
- Moreno R., Borga M., Smedby Ö. "Estimation of trabecular thickness in gray-scale images through granulometric analysis". SPIE-MI 2012, Vol. 8314, 831451, DOI:10.1117/12.906734
- Klintström E., Moreno R., Brismar T.B., Smedby Ö. "Estimating trabecular bone structure parameters with computed tomography". 3rd Norwegian PhD Conference in Medical Imaging, 2011.
- Moreno R., Smedby Ö., Borga M. "Soft Classification of trabeculae in trabecular bone". ISBI 2011, pp. 1641-1644. DOI:10.1109/ISBI.2011.5872718
- Moreno R., Smedby Ö., Borga M. "On the efficiency of the mean intercept length tensor". SSBA 2011.
- Petersson J., Brismar T.B., Smedby Ö. "Analysis of skeletal microstructure with clinical multislice CT". MICCAI 2006, Springer Lecture Notes in Computer Science 4191:880-87. DOI: 10.1007/11866763_108
In view of the great effects on public health of osteoporosis, there is a need for a reliable and valid method for quantitative assessment of the disease in vivo, taking into account not only the total amount of calcium, but also the internal structure of trabecular bone. Therefore, the aims of this project are: • to develop methods for quantitative assessment in vivo of trabecular bone structure using computed tomography and 3D image processing • to validate these methods by comparing them to established reference methods in bone specimens • to make preliminary tests of the methods in patients
DATA ACQUISITION - SPECIMENS:
In a collaboration with research labs at UCSF, USA and Scanco Medical, Switzerland, 15 donated bone specimens of radius are examined using clinical MDCT, clinical dental CT (cone-beam CT) and micro-CT. The micro-CT images, which constitute the gold standard, have been acquired using dedicated research equipment in Switzerland. The ability of the techniques to visualize the microarchitecture of the trabecular bone is studied. In a material consisting of human lumbar vertebrae, harvested in 1981, bone densitometry using DXA and QCT has been acquired. The trabecular bone structure of the defatted vertebrae will be evaluated using MDCT and cone-beam CT.
DATA ACQUISITION - IN VIVO:
Image data will be acquired in patients with osteoporosis of varying degree, with diabetes and with hyperparathyroidism. Both traditional spiral CT, dual-energy CT and cone-beam CT will be used.
After segmentation of the trabeculae and morphological thinning to detect their midlines, we calculate measures of trabecular number (Tb.N), mean trabecular separation (Tb.Sp), mean trabecular thickness (Tb.Th), bone volume relative to total volume (BV/TV) and the number of nodules (Tb.Nd) and termini (Tb.Tm). To better describe the anisotropy of trabeculae, new analysis methods based on a tensor description will be developed.
The methods developed within this project may have importance for understanding the pathophysiology of osteoporosis, for evaluating the effect of given treatment, and possibly as a screening method.