Patellar malalignment is a potential risk factor of patellofemoral pain (PFP) [[1], [2], [3], [4]] and many studies have investigated the association between patellar malalignment and PFP [[5], [6], [7], [8]]. Excessive lateral alignment of the patella, a typical patellar malalignment, can lead to an increase in contact pressure in the patellofemoral joint [9,10], resulting in PFP and early degeneration of the patellofemoral joint [10]. Patellar tilt and shift are proxies for the lateral patellar alignment, and both are potentially related to PFP [5,6,[11], [12], [13], [14]]. Thus, a quantitative evaluation of the patellar tilt and shift is necessary to evaluate knee condition [12,[15], [16], [17]].

PPatellar alignment is generally evaluated using X-ray, computed tomography, and magnetic resonance imaging (MRI) [5,18,19]. Although these instruments can accurately evaluate patellar alignment, they are disadvantageous because of factors such as radiation exposure and taking long time. In contrast, ultrasound (US) is a non-invasive instrument and can quickly obtain images. However, the method for evaluating patellar alignment via US has been challenging.

No study has reported the US method for the patellar tilt evaluation. The lateral patellar angle (LPA), which has been evaluated using the MRI, is used as the indicator of patellar tilt [20]. LPA is calculated from the patellar tilt with respect to the anterior condyle of the femur. Because US images at the anterior knee allow us to simultaneously capture both the patella and the anterior condyle of the femur, the US method equivalent to MRI measured LPA may be developed with two US images at the anterior knee.

A US method has been developed to determine patellar shift [21,22]. The distance between the lateral edge of the patella and anterior lateral condyle of the femur in a single US image (US-lateral distance), obtained at the anterior-lateral knee, was defined as the indicator of patellar shift. Herrington et al. [23] reported that the US-lateral distance shows high intraobserver reliability and a moderate correlation with lateral patella displacement (LPD1), as an indicator of patellar shift, measured via MRI at 20° knee flexion. Several previous studies investigated the association between patellar shift and soft tissue using this method (i.e. US-lateral distance) [[24], [25], [26]]. However, it has been unclear as to whether the method is valid at other knee angles (e.g., at 0° knee flexion). As per a recent systematic review, patellar alignment that is most related to PFP was patellar shift at 0° knee flexion [11]. Therefore, it is necessary to investigate whether the method can be applied at 0° knee flexion. Additionally, lateral patella displacement (LPD2) and bisect offset (BO), in addition to LPD1, are commonly used for evaluating the patellar shift [27,28]. However, it remains unclear whether US-lateral distance can correlate with these indicators (LPD2 and BO). Hence, it is crucial to explore the relations of LPD2 and BO with US-lateral distance and to establish the US method to evaluate the accurate patellar shift at 0° knee flexion.

This study aimed to investigate the reliability and validity of evaluating patellar tilt and shift using US images. We hypothesized that measuring the patellar tilt and shift with US images show correlations with the MRI measurements. In addition, we also hypothesized that the US measurement methods which most corelated to MRI indices of patellar shift are different for 0° and 20° knee flexions.