Validity and Reliability of an Artificial Intelligence-Based Posture Estimation Software for Measuring Cervical and...

Area of Science:

• Biomechanical analysis
• Medical imaging technology
• Artificial intelligence in healthcare

Background:

• Accurate postural assessment is crucial for musculoskeletal disorder management.
• Radiography for posture analysis faces limitations due to radiation, cost, and accessibility.
• AI enables marker-free, 2D photographic posture analysis.

Purpose of the Study:

• To evaluate the validity and reliability of MORA Vu, an AI-based posture estimation software.
• To compare AI-derived measurements with radiographic parameters for cervical and lower-limb alignment.
• To assess MORA Vu as a potential noninvasive screening tool for postural evaluation.

Main Methods:

• Prospective pilot study with 72 participants.
• Measurement of forward head posture (FHP) and digital hip-knee-ankle (DHKA) angles using MORA Vu.
• Comparison with radiographic parameters and assessment of interrater reliability (ICC).

Main Results:

• Strong correlation between FHP and craniovertebral angle (r = -0.712) and C2-7 sagittal vertical axis (r = 0.704).
• High correlation between DHKA angle and radiographic hip-knee-ankle angle (r = 0.754).
• Excellent interrater reliability for FHP (ICC: 0.84) and DHKA (ICC: 0.90).

Conclusions:

• MORA Vu demonstrates strong validity and high reliability for postural assessment.
• The AI software shows potential as a noninvasive, radiation-free screening tool.
• MORA Vu may offer a viable alternative for routine postural evaluation.