Musculoskeletal diseases have a great impact on the quality of life for the individual. Diseases such as osteoporosis (OP), osteoarthritis (OA), and rheumatoid arthritis (RA) often cause pain, stiffness, and loss of mobility, thereby preventing an active life style as well as reducing the work ability. The high prevalence of these diseases combined with the reduced work ability implies a large socioeconomic impact. It is estimated that the total economic burden of arthritis is between 1% and 2.5% of the gross national product of Western countries ( ).
There are many factors related to the onset of these degenerative joint diseases such as trauma, level of exercise, weight, and genetics. However, in general, they are age-related and slowly progressing. Common for these diseases is also that—although there may be preventive treatments with some effect (such as hormone replacement treatment for OP) ( )—there are no effective treatments beyond symptom control. This is in spite of extensive research into both disease etiology and intervention strategies. Important limiting factor for the development of new, effective treatments are the accuracy, precision, and sensitivity of the methods used to monitor the disease progression during clinical studies. Both during development and during validation trials, it is necessary to be able to quantify the effect of a potential treatment. For this, precise biomarkers that accurately quantify central aspects of the state of the disease are needed.
Development of biomarkers for musculoskeletal diseases is challenging due to the relatively slow progression of the disease and because of the mixture of several systemic and biomechanical processes that introduces great variation in the progression of the diseases. A number of relatively old, semiquantitative, golden standard biomarkers exist such as the Kellgren and Lawrence score ( ) for OA based mainly on osteophytes and joint space narrowing or the vertebral fracture risk scoring based on vertebrae height measurements ( ).
These traditional scoring systems are based on expert, medical experience and provide very sensible quantification methods. However, they also suffer from a lack of sensitivity caused by the use of few manual measurements combined with the semiquantitative scores. Since the development of these scores, there have been great advances in both imaging technology and in image analysis methodology. Thereby, it is now possibly to include more anatomical structures true to the actual anatomy, such as the ability to visualize cartilage directly in three dimensions by magnetic resonance imaging (MRI) compared with the indirect joint space width measurements from two-dimensional radiographs. And through automated computer-based analysis it is possible to design more comprehensive scoring systems that are not limited by the time that the medical expert can spend on the individual patient.
In recent years, new computer-based methods for quantitative automated musculoskeletal analysis are surfacing. Some with focus on advances in imaging technology, such as cartilage visualization from very-high-field 7T MRI ( ) or joint kinematics from dynamic MRI ( ). Others with focus on computer-aided automation of existing biomarkers—such as hand joint space measurements ( ). And finally, methods focusing on image analysis methodology—such as automatic cartilage segmentation ( ) or automatic vertebral fracture analysis ( ). However, in spite of the recent advances, the collaboration between clinical researchers and technologically oriented researchers within medical image analysis is still relatively limited (for instance compared with the integration between clinical research and molecular biology research) ( ).
The MICCAI Joint Disease Workshop (see workshop logo in Fig 1 ) was organized with the desire to further exchange of expertise, ideas, and results between the clinically oriented and the technically oriented researchers. The single-day workshop was gifted with many very interesting submissions and with lively discussions during the presentation from the invited speaker (Felix Eckstein from the Paracelsus Medical University in Salzburg), as well as during the seven oral and nine poster presentations.
Open full size image
Figure 1
The workshop logo was composed from a sketch inspired by the little mermaid made for MICCAI 2006. The original little mermaid is a sculpture from 1913 by Edvard Eriksen. Surely, sitting on a rock in the relatively cold Copenhagen harbor water, this aging lady must be familiar with joint stiffness in the morning.
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
Get Radiology Tree app to read full this article<
References
1. Reginster J.Y.: The prevalence and burden of arthritis. Rheumatology 2002; 41: pp. 3-6.
2. Ettinger B., Black D.M., Mitlak B.H., et. al.: Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: results from a 3-year randomized clinical trial. JAMA 1999; 282: pp. 637-645.
3. Kellgren J.H., Lawrence J.S.: Radiological assessment of osteo-arthrosis. Ann Rheum Dis 1957; 16: pp. 494-501.
4. Genant H.K., Wu C.Y., van Kuijk C., et. al.: Vertebral fracture assessment using a semiquantitative technique. J Bone Miner Res 1993; 8: pp. 1137-1148.
5. Pakin S.K., Cavalcanti C., La Rocca R., et. al.: Ultra-high-field mri of knee joint at 7.0T: preliminary experience. Acad Radiol 2006; 13: pp. 1135-1142.
6. Gilles B., Perrin R., Magnenat-Thalmann N., et. al.: Bone motion analysis from dynamic MRI: acquisition and tracking. Acad Radiol 2005; 12: pp. 1285-1292.
7. Pfeil A., Böttcher H., Seidl B.E., et. al.: Computer-aided joint space analysis (CAJSA) of the proximal-interphalangeal joint-normative age-related and gender specific data. Acad Radiol 2007; 14: pp. 594-602.
8. Folkesson J., Dam E.B., Olsen O.F., et. al.: Segmenting articular cartilage automatically using a voxel classification approach. IEEE Trans Med Imaging 2007; 26: pp. 106-115.
9. de Bruijne M., Lund M.T., Tanko L.B., et. al.: Quantitative vertebral morphometry using neighbor-conditional shape models. MICCAI Med Image Comput Comput Assist Interv 2006; 9: pp. 1-8.
10. Schaller S., Henriksen K., Hoegh-Andersen P., et. al.: In vitro, ex vivo, and in vivo methodological approaches for studying therapeutic targets of osteoporosis and degenerative joint diseases: how biomarkers can assist?. Assay Drug Dev Technol 2005; 3: pp. 553-580.
11. Iglesias J.E., de Bruijne M.: Semi-automatic segmentation of vertebrae in lateral X-rays using a conditional shape model. Acad Radiol 2007; 14: pp. 1156-1165.
12. Roberts M., Cootes T.W., Pacheco E.M., et. al.: Quantitative vertebral fracture detection on DXA images using shape and appearance models. Acad Radiol 2007; 14: pp. 1166-1178.
13. Langs G., Peloschek P., Bischof H., et. al.: Model based erosion spotting and visualization in rheumatoid arthritis. Acad Radiol 2007; 14: pp. 1179-1188.
14. Kubassova O.A., Boyle R.D., Radjenovic A.: Quantitative analysis of dynamic contrast-enhanced MRI datasets of the metacarpophalangeal joints. Acad Radiol 2007; 14: pp. 1189-1200.
15. Fripp J., Bourgeat P., Crozier S., et. al.: Segmentation of the bones in MRIs of the knee using phase, magnitude and shape information. Acad Radiol 2007; 14: pp. 1201-1208.
16. Qazi A.A., Dam E.B., Nielsen M., et. al.: Osteoarthritic cartilage is more homogeneous than healthy cartilage: Identification of a superior ROI co-localized with a major risk factor for osteoarthritis. Acad Radiol 2007; 14: pp. 1209-1220.
17. Folkesson J., Dam E.B., Olsen O.F., et. al.: Accuracy evaluation of automatic quantification of the articular cartilage surface curvature from MRI. Acad Radiol 2007; 14: pp. 1221-1228.
18. Ward D.W., Hamarneh G., Ashry R., et. al.: 3D shape analysis of the supraspinatus muscle: a clinical study of the relationship between shape and pathology. Acad Radiol 2007; 14: pp. 1229-1244.