A Minimal Sensor Inertial Measurement Unit System Is Replicable and Capable of Estimating Bilateral Lower-Limb Kinematics in a Stationary Bodyweight Squat and a Countermovement Jump.
AuraLea FainBenjamin R HindleJordan T AndersenBradley C NindlMatthew B BirdJoel Thomas FullerJodie Anne WillsTim Leo Atherton DoylePublished in: Journal of applied biomechanics (2023)
This study aimed to validate a 7-sensor inertial measurement unit system against optical motion capture to estimate bilateral lower-limb kinematics. Hip, knee, and ankle sagittal plane peak angles and range of motion (ROM) were compared during bodyweight squats and countermovement jumps in 18 participants. In the bodyweight squats, left peak hip flexion (intraclass correlation coefficient [ICC] = .51), knee extension (ICC = .68) and ankle plantar flexion (ICC = .55), and hip (ICC = .63) and knee (ICC = .52) ROM had moderate agreement, and right knee ROM had good agreement (ICC = .77). Relatively higher agreement was observed in the countermovement jumps compared to the bodyweight squats, moderate to good agreement in right peak knee flexion (ICC = .73), and right (ICC = .75) and left (ICC = .83) knee ROM. Moderate agreement was observed for right ankle plantar flexion (ICC = .63) and ROM (ICC = .51). Moderate agreement (ICC > .50) was observed in all variables in the left limb except hip extension, knee flexion, and dorsiflexion. In general, there was poor agreement for peak flexion angles, and at least moderate agreement for joint ROM. Future work will aim to optimize methodologies to increase usability and confidence in data interpretation by minimizing variance in system-based differences and may also benefit from expanding planes of movement.
Keyphrases
- total knee arthroplasty
- lower limb
- knee osteoarthritis
- anterior cruciate ligament
- anterior cruciate ligament reconstruction
- high intensity
- total hip arthroplasty
- healthcare
- magnetic resonance imaging
- machine learning
- magnetic resonance
- high resolution
- electronic health record
- high speed
- deep learning
- artificial intelligence