Extracorporeal Shockwave Versus Chitosan-Nanoparticles Phonophoresis On Functional Improvement And Anatomical Changes Detected With Artificial Intelligence-Based Texture Analysis In Knee Osteoarthritis: A Single Blind Randomized Controlled Trial

Abstract

Introduction: Osteoarthritis (OA) is a leading cause of disability worldwide, with knee osteoarthritis (KOA) being the most prevalent form and a major contributor to pain and functional decline. Once considered primarily a cartilage disorder, KOA is now recognized as a disease involving the entire joint, particularly the subchondral bone, which plays a crucial role in disease progression and pain generation. Recent therapeutic strategies have thus shifted focus toward modulating subchondral bone remodeling using non-invasive modalities such as extracorporeal shockwave therapy (ECSWT) and low-intensity pulsed ultrasound (LIPUS). The integration of LIPUS with bioactive agents like Chitosan nanoparticles offers a promising approach for enhancing tissue repair and reducing inflammation. Moreover, artificial intelligence (AI)-based texture analysis provides sensitive and objective quantification of subchondral bone changes, enabling early detection and precise monitoring of therapeutic effects in KOA. This study aimed to compare the effects of ChNPS-phonophoresis and ECSWT on clinical outcomes and subchondral bone microarchitecture in KOA, using AI-based texture analysis. and to assess the sensitivity and objectivity of AI-based texture analysis for detecting early bone changes and monitoring treatment responses. Materials and Methods: A single blind randomized trial was conducted on 120 patients aged between 40-60 years old, with mild to moderate KOA. Patients with previous knee surgery or lower limb fractures, inflammatory or neurological disorders, significant synovitis or acute inflammation were excluded. Patients were divided equally into three groups (n = 40). Group A received ChNPs gel-phonophoresis with exercise, Group B received ECSWT with exercise, and Group C (control) received exercise only. Interventions lasted 8 weeks. Pain and function were assessed using VAS and Arabic WOMAC. Subchondral bone changes were evaluated using IB Lab TX Analyzer™ via Bone Structure Value (BSV) and Fractal Dimension (FD) and recorded at baseline and post-treatment. Results: All groups showed significant post-treatment clinical improvements (P < 0.001), with Group A achieving the most pronounced reductions in VAS and WOMAC scores (P < 0.05). Group A also showed significant anatomical improvement: BSV increased from 0.43 ± 0.23 to 0.62 ± 0.24, and FD increased from 0.67 ± 0.11 to 0.77 ± 0.13 (both P < 0.001), with effect sizes of η² = 0.081 and η² = 0.114, respectively. Conclusion: ChNPs gel-phonophoresis yielded superior outcomes in pain, functional, and bone microarchitecture. AI-based TX enhances precision in monitoring KOA treatment response, supporting structure-targeted, non-invasive approaches.