Singapore Institute of Technology, 10 Dover Drive 138683, Singapore.
Sengkang General Hospital, 110 Sengkang East Way 544886, Singapore.
Radiography (Lond). 2024 Oct;30(6):1578-1587. doi: 10.1016/j.radi.2024.09.067. Epub 2024 Oct 11.
The criteria for determining the acceptability of total knee replacement (TKR) radiographs are not established in current clinical practice. In TKR patients, the implant components replaced the anatomical landmarks, making it more difficult for radiographers to determine the degree of rotation. This study aims to establish an acceptable range of knee rotation for TKR radiographs.
Rejected TKR radiographs (199 AP and 186 lateral) were analysed retrospectively. Radiographers objectively measured rotation on the radiographs. A subset of 46 AP and 46 lateral radiographs were rated by orthopaedic surgeons for rotation and diagnostic value. Inter-rater reliability (IRR) of radiographic measurements and surgeons' ratings were analysed using Bland-Altman and Cohen's kappa, respectively. Spearman's rank-order correlation and Receiver Operator Characteristic analyses were used to determine the correlation and diagnostic performance of the radiographic measurements against the surgeon's ratings.
Strong IRR was observed for the radiographic measurements. Only slight to fair agreement was observed for the surgeons' rotation and diagnostic value ratings of the radiographs. Moderate to strong correlation was observed between the radiographic measurements and the surgeons' ratings. The radiographic measurements provided acceptable to excellent discrimination of acceptable and unacceptable radiographs. The acceptable range of measured rotation for usability was AP: 0-5.29 mm and lateral: 0-6.01 mm.
The proposed measurement methods and the established rotation range could potentially be used by radiographers in clinical practice to determine the acceptability of TKR radiographs. Follow-up studies could investigate uncommon knee implants and seek consensus across different institutions on the acceptable degree of rotation.
The proposed method suggests that accepting radiographs within the threshold (AP: 5.29 mm, lateral: 6.01 mm) reduces repeated examination and radiation exposure and improves imaging efficiency.
目前临床实践中尚未确定全膝关节置换术（TKR）放射影像的可接受标准。在 TKR 患者中，植入物组件替代了解剖学标志，这使得放射技师更难以确定旋转的程度。本研究旨在确定 TKR 放射影像可接受的膝关节旋转范围。
回顾性分析了 199 张前后位（AP）和 186 张侧位的不合格 TKR 放射影像。放射技师客观测量放射影像上的旋转角度。对 46 张 AP 和 46 张侧位放射影像进行了一部分由骨科医生对旋转和诊断价值的评估。使用 Bland-Altman 和 Cohen 的 Kappa 分别分析放射测量和外科医生评分的组内相关系数（IRR）。Spearman 秩相关和受试者工作特征（ROC）分析用于确定放射测量值与外科医生评分之间的相关性和诊断性能。
放射测量的 IRR 很高。只有轻微到公平的协议观察到外科医生的旋转和诊断价值的评分的放射影像。在放射测量值与外科医生评分之间观察到中度到高度的相关性。放射测量值对可接受和不可接受的放射影像具有可接受到极好的区分能力。可用性测量旋转的可接受范围为 AP：0-5.29mm 和侧位：0-6.01mm。
所提出的测量方法和建立的旋转范围有可能被放射技师在临床实践中用于确定 TKR 放射影像的可接受性。后续研究可以研究不同的膝关节植入物，并在不同机构之间就可接受的旋转程度达成共识。
该方法建议，在阈值内（AP：5.29mm，侧位：6.01mm）接受放射影像可以减少重复检查和辐射暴露，提高成像效率。
