Static and dynamic error of a biplanar videoradiography system using marker-based and markerless tracking techniques

Area of Science:

• Biomechanics
• Medical Imaging
• Orthopedics

Background:

• Biplanar videoradiography is increasingly used for in vivo joint function evaluation.
• Current 3D bone motion reconstruction methods include marker-based and markerless tracking.
• Systematic error evaluation is crucial for reliable quantification of in vivo bone motion.

Purpose of the Study:

• To present novel markerless tracking software utilizing GPU technology.
• To describe a method for quantifying the systematic error of biplanar videoradiography systems.
• To evaluate the systematic error of marker-based and markerless tracking at the W.M. Keck XROMM Facility.

Main Methods:

• Developed new markerless tracking software with GPU acceleration.
• Quantified systematic error using independent gold standard instrumentation (linear/rotary stages, angular displacement transducer).
• Tested marker-based tracking with a multi-marker polycarbonate flag and markerless tracking on human cadaveric bones (femur, radius, ulna).

Main Results:

• Marker-based tracking achieved motion accuracy within 0.1 mm and 0.1 degrees.
• Markerless tracking accurately captured rapid bone motions (distal femur, radius, ulna) within 0.15 degrees.
• Both tracking methods showed excellent agreement with gold standard instrumentation under static and dynamic conditions.

Conclusions:

• The novel markerless tracking software effectively quantifies in vivo bone motion.
• Both marker-based and markerless techniques provide reliable and accurate motion capture.
• These validated methods will advance the study of in vivo joint motion during high-impact activities.