研究目的
To determine the accuracy and repeatability of pelvic tilt using Vuforia AR technology for preoperative examination in hip arthroplasty.
研究成果
The results show that pelvic tilt can be measured with good accuracy and repeatability using Vuforia AR technology, with deviations less than 3° in all measurements. The object target provided more accurate results than the image target, especially with the smartphone camera. This method has potential for preoperative examinations to reduce hip dislocation rates by enabling individualized acetabular cup placement. Future work could involve porting the application to digital eyewear for easier use by a single operator.
研究不足
The accuracy decreases with increasing distance between camera and target, and smaller targets may lose tracking faster. The study was conducted on a model, not human subjects, which may not fully replicate clinical conditions. The use of specific cameras and devices may limit generalizability.
1：Experimental Design and Method Selection:
The study used Vuforia AR technology to track an ultrasound probe for measuring pelvic tilt. Two tracking targets were employed: a 2D image target attached to the probe and an object target created by coating the probe with a pattern using water transfer printing. The pelvic tilt was calculated based on three anatomical landmarks (SIAS-R, SIAS-L, SP) using the anterior pelvic plane (APP).
2：Sample Selection and Data Sources:
A hip model embedded in ultrasound-friendly silicon was used as the sample. Measurements were performed with this model to simulate patient conditions.
3：List of Experimental Equipment and Materials:
Ultrasound device MicrUs (Telemed Ultrasound Medical Systems), linear ultrasound probe coated with water transfer printing, Vuforia SDK and Object Scanner, self-generated Vuforia Image Target patterns, pelvis model in silicon, Logitech C920 HD Webcam, Samsung Galaxy S7 Smartphone, Unity3D application with custom code for data handling.
4：Experimental Procedures and Operational Workflow:
The ultrasound probe was tracked using a webcam or smartphone camera. Landmarks were detected by placing a virtual pointer on the ultrasound image via touch input. For each measurement, SIAS-R, SIAS-L, and SP were recorded in sequence, referenced to an image target on the hip model. The APP normal vector was calculated relative to a reference marker. Distance between camera and probe was maintained between 20-40 cm. A total of 100 measurements per tracking target per camera were conducted.
5：Data Analysis Methods:
Data were analyzed for mean pelvic tilt angles and standard deviations. Accuracy was evaluated based on deviations in axes and rotations, with statistical analysis of repeatability and comparison between tracking targets and cameras.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
