Direct arthroscopic cartilage assessment remains the gold standard. It is recommended by the International Cartilage Repair Society (ICRS) to systematically assess cartilage status during arthroscopy but this examination is highly subjective, poorly reproducible, time-consuming and lacks precision. US has shown good potential for cartilage evaluation but is limited in extra-articular conditions. It is also difficult to manually maintain a perfect perpendicularity between the ultrasound beam and the curved surface of the cartilage. Therefore, we have developed a navigated intra-articular US probe (NIAUS). The NIAUS probe could contribute to a more exhaustive and direct intra-articular evaluation of cartilage integrity. Navigation enables control of the US echo pulse perpendicularity and its localisation relative to the joint. Our objectives were (1) to evaluate automatic cartilage thickness measurement with the NIAUS probe in comparison to high definition MRI on cartilage samples, (2) to generate a real-time 3D map of the thickness parameter on samples, and (3) to demonstrate the feasibility of a full NIAUS probe cartilage scan on a specimen distal femur in arthroscopic conditions. The NIAUS probe is a 4.5mm probe consisting of a 64 element linear array transducer with a central frequency of 13 MHz and a motorised head. The NIAUS probe is navigated. The rotating US head position is controlled by navigation in order to enable constant perpendicular acquisition of cartilage. The NIAUS probe thickness measurement (1) was evaluated on bone and cartilage samples of 9 tibial plateaus. The cartilage thickness was measured via automatic segmentation. Each sample was also scanned in a high resolution MRI (4,7 Tesla) and cartilage thickness was semi-automatically extracted for comparison. During NIAUS scan, (2) a visual 3D map was generated. Finally (3), we scanned two distal femurs with the NIAUS probe in arthroscopic navigated conditions on one specimen and a 3D map of the distal femur thickness was generated in real time. NIAUS thickness measurement (1) absolute error compared to MRI for 9 plateaus ranged from 0.15mm to 0.32mm in median, p25=0.07 and 0.18, p75=0.28 and 0.5 respectively. 3D maps of the sample cartilage thickness (2) were generated in real time during the NIAUS scan. The cadaveric procedure (3) was conducted without incident via the two anterior portals and a 3D map of the distal femurs cartilage thickness was generated. A precise US arthroscopic grading and scoring of cartilage during surgery could help for better standardisation, prediction of results and making “live” decisions. Our
