In this paper the morphological, molecular (biochemical) and mechanical features of osteoarthritis and various modalities that can be used to detect knee osteoarthritis (OA) at an early stage are discussed. Based on the facts and other supporting evidences, it is hypothesised that early knee OA detection can be improved by combining the assessment of cartilage thickness and water content of articular cartilage. The main objective of this research work is to develop a non-invasive analysis and diagnostic tool with high sensitivity and specificity using quantitative magnetic resonance (qMR) imaging that can be used to detect knee osteoarthritis below stage 1 with reference to the International Cartilage Repair Society (ICRS) grading system. Research has shown that there is a relationship between water content of cartilage with quantitative value of magnetic resonance (qMR) signals due to the presence of hydrogen in human knee cartilage. It is clear that the current methods (imaging modality, feature selection and classification) of detecting any single molecular changes in the articular cartilage on their own are not able to provide sufficient distinguishing capability of early OA. This research project points toward a multi-feature detection and classification for a successful diagnostic tool. © 2011 IEEE.