The Measurement of Focal Diurnal Variation in the Femoral Articular Cartilage of the Knee A. D. Brett 1 , J. C. Waterton 2 , S. Solloway 1 , J. E. Foster 3 , M. C. Keen 3 , S. Gandy 3 , B. J. Middleton 4 , R. A. Maciewicz 2 , I. Watt 5 , P. A. Dieppe 6 , and C. J. Taylor 1 1 Division of Imaging Science, Manchester University, Manchester, UK 2 Cardiovascular, Metabolism & Musculoskeletal Research Dept AstraZeneca Pharmaceuticals, Cheshire, UK 3 Medical Physics & Bioengineering, United Bristol Healthcare Trust, Bristol, UK 4 Safety of Medicines Dept., AstraZeneca Pharmaceuticals, Cheshire, UK 5 Dept. Radiology, United Bristol Healthcare Trust, Bristol, UK 6 Dept. Rheumatology, University of Bristol, Bristol, UK Abstract. Our objective was to test the hypothesis that focal diurnal changes occur in the femoral articular cartilage of the knee in asymp- tomatic young adults. Six volunteers each were scanned early in the morning, and at the end of a working day spent mainly standing. This protocol was repeated on three successive weeks. Femoral cartilage seg- mentations were obtained using a region-growing algorithm. These seg- mentations then were regridded onto a 500-pixel template, and differ- ences in the resulting thickness maps were assessed. Analysis of variance showed no significant diurnal variation in mean thickness. There were, however, statistically-significant diurnal changes in the thickness maps. Cartilage thickness decreased during the day in three specific locations which suffer the greatest biomechanical force. 1 Introduction Osteoarthritis (OA) is one of the principal causes of disability in elderly people. The disease is characterised by focal structural changes, and eventual loss, of articular cartilage. The knee joint is often most severely affected, and anatomical locations suffering the highest biomechanical force are most likely to exhibit cartilage damage. Recently, it has become possible to measure accurately and precisely the volume of the articular cartilage with MRI [13,4]. Fat-suppressed 3D spoiled gradient-echo MRI has been widely adopted, and provides good contrast with reasonable scanning times. However, because of the shape of the cartilage, man- ual segmentation of the images is tedious, so that for large-scale studies, it is desirable to develop semi-automatic techniques [6]. It now appears possible that MRI assessment of cartilage volume will be accurate and precise enough to mea- sure OA disease progression and therapeutic intervention in small-scale trials, C. Taylor, A. Colchester (Eds.): MICCAI’99, LNCS 1679, pp. 328–338, 1999. c  Springer-Verlag Berlin Heidelberg 1999