Osteoarthritis (OA) is a degenerative joint disease and one of leading causes of chronic disability in the world. It is suggested that in addition to degeneration of the cartilage, changes are also seen in the adjoining subchondral and trabecular bone. Articular cartilage and subchondral bone act in concert with regards to the mechanical loading of the joint and the subchondral mineralized zone is recognized to play an important role in reducing the axial impact forces typically encountered during dynamic joint loading and adapts to the mechanical demands during normal and abnormal joint loading. However, to date there is no quantitative evidence that suggest that both changes in the bone and the cartilage are interlinked. We performed the first-in-human study analyzing cartilage biochemistry and its interactions with the bone function across the entire joint in patients with knee osteoarthritis using state-of-the-art simultaneous time-of-flight (TOF) Positron Emission Tomography/Magnetic Resonance (PET/MR) (GE Healthcare, USA) imaging.Sixteen knee osteoarthritis patients (Age: 57±8.6 years, BMI: 25.0±3.8 kg/m², 69 % male, KL = 0-3) received intravenous injections of ¹⁸F-NaF (9.1±0.9 mCi) at the onset of the PET/MR scan. Dynamic PET and quantitative MR images were acquired for 60-minutes followed by a 30-minute break and a static PET/MR scan was acquired for 10 minutes, resulting in a 90-minute post injection scan. An 8-channel medium size flex coil was used (GE healthcare) due to minimal attenuation for the PET signal. The MRI protocol included 3D FSE (CUBE), a quantitative combined T_1ρ/T₂ as well as MRA opt for fusion with PET images. PET was acquired in list mode for 60-minutes and dynamic multi-frames were reconstructed using an iterative algorithm. Irreversible two-tissue kinetic model was used to derive Slope K_i, influx rate constant using the dynamic data. Maximum Standardized Uptake Values [SUV_max normalized by body weight] were examined in bone regions located adjacent to the cartilage using the static data (90-minute post injection scan). An atlas-based registration technique was used to obtain voxel-based T_1ρ maps and statistical parameter mapping was used to investigate the relationship (Pearson correlation) between bone and cartilage in the entire knee joint after adjusting for age, gender, BMI and KOOS scores. Bone remodelling (i.e., uptake of ¹⁸F-NaF) and cartilage biochemistry (T_1ρ) at each voxel were examined across the knee joint and correlations or connectivity between the two metrics were highly correlated not just in adjacent bone and cartilage, but in non-adjacent cartilage compartments (Table 1). The Slope K_i of NaF in medial tibia (MT) is highly correlated with T_1ρ in 10.1 % of the voxels (R=0.68, p = 0.027) (Fig. 1A) and with 14.1 % of the voxels in patella (P) (R = 0.64, p = 0.022), and smaller areas of correlations are found in other cartilage compartments. SUV_max in lateral tibia (LT) is highly correlated with T_1ρ in 20.6 % of the voxels in LT (R = 0.74, p = 0.007) (Fig. 1B), in 27.5 % of the voxels in lateral femural condyle (LFC) (R = 0.77, p = 0.007), and in 10.3 % of the voxels in P (R =0.7, p 0.018), and smaller correlations are found in other cartilage compartments. SUV_max in P is negatively correlated with T_1ρ in 3.7% of the voxels in P (R = -0.54, p = 0.035) (Fig. 1C), furthermore a positive correlation is seen in a more focal region in the bone layer of the cartilage of P (Fig 1C). SUV_max in P is also highly correlated with T_1ρ in 16.4 % of the voxels in the cartilage of TrF (R = -0.68, p = 0.013) (Table 1). A direct relationship was seen between early degenerative changes using MRI and bone remodelling and bone blood flow using PET imaging. Not only did the adjoining bone to cartilage show interactions, but also bone regions further away from the cartilage, this we name the “Bone-Connectome”. A trend for higher uptake was seen in patients with pain compared to patients without pain (p < 0.1), but only when no lesions were observed on the MR (data not shown). Correlations of standardized uptake values obtained from static PET acquisitions and kinetic uptake rate obtained from dynamic acquisition was seen to correlate. Considering the knee joint as a whole we successfully demonstrate the relationship between early cartilage biochemical degeneration, bone remodelling in adjoining and non-adjoining regions, this provides important information for understanding the pathophysiology of OA, as well as elucidating the natural history of the disease and in assessing therapeutic targets in the treatment of OA.