This study was performed to compare the consequences of different zone-divisions in MRI T2 of articular cartilage using an animal model of early osteoarthritis (OA).The progression of osteoarthritis was studied using eighteen mature canines, which was approved by the institutional review committee. Six animals were the non-operated normal controls (Normal) and twelve animals underwent anterior cruciate ligament (ACL) transection on one of their knees. The operated animals were sacrificed 8 week and 12 weeks post-surgery (8wk-OA, 12wk-OA). The non-transected contralateral knees in these animals were also studied (8wk-C, 12wk-C). Within 24 hours of sacrifice, the intact knee joints were first imaged using a Varian 7T/20cm macro-MRI. T2 experiments were carried out using a multi-slice multi-echo (MSME) pulse sequence. The ten echo times were 10, 20, 30, ... 100ms, with TR set at 3 seconds. The FOV was 5cm with a matrix size at 256 x 256. The imaging matrix was reconstructed in post-acquisition to 512 x 512 using Varian’s Fourier reconstruction. Ten interleaved image slices were selected, approx. 2.5mm apart from each other. To determine the sub-tissue zonal division accurately, the joint capsules were opened after macro-MRI. Each medial tibial plateau was cut into five cartilage-bone blocks, which were T2 imaged on a Bruker 7T/9cm µMRI, at 17µm pixel resolution. Both macro-MRI and µMRI images were approximately at the same tissue locations in the knee. The data in this report were entirely from the macro-MRI experiments; the µMRI data, which would be presently in a separate report, only provided the criteria for the sub-tissue zone division ¹ in this report.Fig 1a shows the quantitative T2 maps by macro-MRI, with the medial tibial location highlighted in Fig 1b (the cropped ROI) and the equivalent µMRI T2 map from the similar location in Fig 1c. Two methods of image analysis were used in macro-MRI T2 data analysis: (1) an equal division of the entire cartilage from the surface to the cartilage-bone interface (Fig 2), and (2) an unequal division of the same cartilage based on the knowledge from µMRI zones (Fig 3). With the equal division (Fig 2), significant difference can be detected statistically in Zone I between Normal and both OA groups; however, there were no statistical significance between Normal and Contralateral cartilage, or among any comparison in Zone II and Zone III. With the zonal knowledge from µMRI, the unequal division of the same T2 profiles from macro-MRI enabled the significant differences to be detected statistically in the superficial zone (SZ) between the Normal and OA, and between the Normal and Contralateral cartilage. In addition, significant differences can also be detected statistically in two sub-tissue zones (the transitional zone TZ and the upper radial zone RZ1).T2 in cartilage increases due to the progression of OA. This comparative project studies the intermediate-resolution macro-MRI T2 data, which shows the critical role of the zonal division in the detection of OA by MRI. The use of the equal division of cartilage thickness, which is common in clinical MRI of human cartilage ², can detect cartilage degradation, but only in the surface tissue, and only between the directly lesioned tissue and the healthy tissue. This equal division has failed to detect the smaller differences in T2 between the normal and contralateral cartilage. In contrast, an unequal zonal division can significantly improve the detection of cartilage degradation – in the deeper tissues as well as between the contralateral and normal tissue. This improved detection may become important in the clinical diagnostics of early OA. We are completing a topographical analysis of this data at multiple resolutions, which could provide further insight into the complications of both depth-dependent and topographically distributed OA degradation on the medial tibial plateau.In this study, we utilized the knowledge from the high-resolution µMRI zonal division to divide the T2 profiles at the intermediate resolution macro-MRI both equally and non-equally. We show significant differences between the two image analysis methods in the detection of small changes in T2 due to OA progression. A similar approach ³ could be developed and implemented in the image analysis in clinical MRI of human OA, to improve the detection sensitivity towards the early signs of cartilage degeneration in clinical MRI of human.