Non-invasive Detection of Neuronal Electrical Activity in the Knee Joint Using Quantum Sensing MRI

Abstract Noninvasive assessment of neuronal activity in the knee may improve understanding of osteoarthritis (OA)-related pain and joint degeneration. A 3T quantum-sensing based MRI framework was developed to estimate firing-rate-related signal fluctuations from free induction decay (FID) phase measurements, separating water and fat signals for compartment-specific estimates. The approach was evaluated in agar-gel phantoms, simulations, and six healthy volunteers and six participants with early knee OA, at rest and during repeated ankle flexion. No firing events were detected in the phantom. Simulation performance improved with signal-to-noise ratio, reaching R 2 = 0.99 at SNR = 250. In vivo, water-derived estimates correlated negatively with T 1ρ relaxation times (cartilage, r = -0.86 to -0.58; meniscus, r = -0.78 to -0.73). Fat-derived estimates differentiated OA participants from healthy volunteers (d = 3.96–4.89; p < 0.05), correlated positively with T 1ρ (cartilage r = 0.59–0.70; meniscus r = 0.70–0.71) and showed moderate exploratory associations with WOMAC scores (r = 0.43–0.55). Repeatability was higher during ankle flexion than at rest. These findings support the technical feasibility of compartment-specific FID phase analysis in the knee. Larger studies with physiological-noise control and independent biological validation are required to determine whether these MRI-derived measures specifically reflect peripheral neuronal activity and have clinical utility in OA.