Abstract
Background and Motivation
Accurate placement of glenoid components in reverse and total shoulder arthroplasty has been shown to reduce the risk of implant failure (1, 2, 6). Surgical techniques and literature describe methods to determine favorable positions for implant placement (3, 4, 5) but achieving that position surgically remains a challenge. Placement of glenoid components is faced with the challenge of variable glenoid morphology on which conventional instrumentation does not always provide a reliable reference (6, 7, 8). Limited surgical exposure is another challenge since many anatomic landmarks are not visible to the surgeon to use as spacial reference. Anatomic landmarks and angles can be more reliabily selected on CT scans with 3-dimentional reconstruction (9,10) yet few methods allow for the reproducible translation of these plans to surgery. Navigation has produced better accuracy and lower variability than conventional instrumentation (11), yet its regular usage remains limited, especially in the shoulder.
Methods
A patient specific planning and guiding system has been developed for glenoid implant placement of total and reverse shoulder arthoplasty procedures. This method allows for preoperative planning on a patient specific virtual 3D model of the scapula derived from CT images (Figure 1), and guided placement of a pin which which serves as the central axis for determining proper implant position. An initial implant position was presented on the virtual model based on the methods described by the surgical technique of the corresponding procedure. These plans were either approved or adapted to a desired position within the planning software by the surgeons. Using this planned position as input, patient specific surgical guides were created which fit onto the exposed anatomy and guide the drilling of the pin (Figure 1). This method was tested on 14 cadavers, with attention directed to translation of the starting point from the original plan, the ability to reproduce the intended degree of inferior tilt, and the ability to reproduce the glenoid version angle.
Results
The ability to reproduce the surgical plan was found to be highly accurate for the 14 cadaveric specimens. Translational accuracy amongst the 14 cadavers was found to be 1.01 ± 0.53 mm, tilt was 0.46 ± 0.53 degrees, and the accuracy of version was found to be 1.16 degrees ± 1.15 degrees.
Conclusion
Surgical planning on patient specific virtual bone models and the corresponding surface matched drilling guides for glenoid implant positiong provide surgeons with an accurate method to achieve the desired surgical implant position. The measured accuracy compares favorably to both conventional and navigated techniques.
