TMJ Pericapsular Guided Injection with Visual-Inertial Odometry (Augmented/Mixed Reality): a Novel Pilot Clinical Approach for Joint Osteoarthrosis Drug Delivery.

Introduction: Temporomandibular joint (TMJ) osteoarthrosis often requires pericapsular injections for symptom management. Image-guided techniques have demonstrated superior outcomes compared to traditional manual palpation. Augmented reality (AR) integrates digital images into the clinical scenario in real time and proved to be an efficient surgical guide. This pilot study evaluated the outcomes of AR-assisted versus manual palpation for polynucleotides (PN) pericapsular injections in subjects with TMJ osteoarthrosis.

Materials And Methods: 20 condyles (10 patients) with bilateral TMJ osteoarthrosis were randomly assigned to either the test (N=10 condyles - 5 patients; 53 years old SD 13; 4 female - 6 male) or the control group (N=10 condyles - 5 patients; 58 years old SD 15; 5 female - 5 male): AR-guided or standard injection site localization following manual palpation. AR-guided injections employed custom visual-inertial odometry (VIO) algorithm, using the condyle's segmented images for precise injection site estimation. Mouth opening, lateral movements, and pain intensity were assessed at baseline (T0), 2 weeks (T1), 6 weeks (T2), 12 weeks (T3) post-injection. Statistical analysis included non-parametric Mann-Whitney test between groups for all the outcomes assessed.

Results: AR-guided and manual palpation groups showed improvements in VAS scores and mandibular kinematics over time. The AR-guided group demonstrated significant pain reduction and increased lateral movement. There was no statistically significant difference between the two groups in terms of overall efficacy. The AR-guided procedure took approximately 120 seconds longer compared to manual palpation (mean time 343.8 sec and 211.2 sec).

Conclusion: The new technique showed clinical efficacy and improvement trends in some clinical parameters such as VAS and laterality, with a reasonable increase in timing compared to the standard practice. This first application of AR can be considered for guided TMJ injections, but further research is needed.

Clinical Significance: AR-assisted TMJ injections offered improvements trends in VAS scores and lateral movements, making it a promising digitally guided approach for daily practice without the need of direct palpation.