Enhancing Meniscal Allograft Tissue Size-Matching Accuracy - A Novel Processing Protocol from the Queensland Tissue Bank

Objectives: Meniscal allograft transplant (MAT) is a recognized salvage procedure for the meniscal deficient knee. Accurate sizing of allograft tissue is inherently linked to clinical outcomes and must be matched to the recipient’s native anatomy to replicate suitable knee kinematics. Currently the Queensland Tissue Bank (QTB) is the only Australian processing facility offering donor meniscal allograft tissue. The current QTB protocol involves manual measurement of the meniscus, antero-posterior (AP) length and medio-lateral (ML) width, at the time of processing. This is then matched against surgeon request based on 2D-imaging and templating which may result in measurement errors leading to size mis-match. Advancements in 3D technology enable acquisition, computational modelling, and storage of available allograft tissue in a virtual library. 3D surface scanning and cross-sectional imaging (CT and MRI) has consistently demonstrated more accurate measurements of meniscal allografts during the matching process. The aim of this study is to evaluate the accuracy and efficiency of a novel processing protocol for meniscal allograft tissue from the QTB. Secondary aims include development and validation of a virtual library and an AI-based auto-select algorithm that allows real-time identification of the most suitable allograft tissue. The development of a novel processing protocol for meniscal allograft tissue using 3D Technology can improve the accuracy of size-matching for MAT procedures. A Virtual library and an Auto-Select Algorithm using artificial intelligence (AI) can identify best matched available allograft for patient specific requests. Methods: Fifty (50) deidentified research-grade menisci were processed using manual, 3D surface scanning and CT acquisition. The scanned data was processed into high-quality 3D models and stored in the QTB virtual library. An Auto-Select Algorithm was developed using artificial intelligence to identify the closest allograft matches available in the library. The selected allografts can then be assessed for clinical suitability and referenced against patient-specific imaging. Results: Previous analysis indicates strong correlation between manual versus 3D surface scanned and CT measurements. An Auto-Select Algorithm has been developed to facilitate real-time identification of the best-matched allografts. This innovative approach shows promising potential to improve the accuracy and efficiency of meniscal allograft matching and serves to establish a new processing protocol. Conclusion: Advances in allograft processing allows for more accurate and efficient sizing of meniscal allograft tissue which is fundamental for successful patient outcomes. The development of an auto-select algorithm allows real-time analysis of available allograft. This newly proposed protocol could be expanded to include other allograft tissues.