In authors’ hospital, anteroposterior and lateral radiograph knee-joint images acquired between June 1, 2020 and June 25, 2020 were included in this study. Cases were excluded where there was noticeable knee joint distortion or knee joint prosthesis replacement.
All cases were randomly divided into two groups. One group would be placed in the conventional position for lateral knee joint radiography, and the other group was placed in an optimized position for lateral knee joint radiography. The conventional position was that the femur and the tibiofibula was at an angle of approximately 120 , the standard middle sagittal plane of the knee joint was parallel to the detector, and the centerline of X-ray was incident vertically through the center of the patellofemoral joint. The optimized position, resulted from the retrospective analysis of the knee-joint CT images, was described in more detail in the following relevant section ‘the optimal radiography angle for the standard lateral knee joint’. All the cases in the two groups applied the same anteroposterior position for radiography. All the knee-joint X-ray images were acquired by a Ruike DR 3500, the source to image-receptor distance (SID) was 120 cm, and the exposure parameter was 60 kV at 8 mA.
Determination of the standard lateral radiography position for knee joint
One hundred cases of knee joint CT images initially acquired in our hospital from June 2018 to May 2019 were retrospectively analyzed. They were divided into mate and female groups, and the optimal inward rotation and outward tilt angles of the left and right knee joints were analyzed.
All CT images for retrospective analysis were acquired by spiral scanning using a Siemens SOMATOM Perverse 128 row of VCT machine. The scan parameter was 130 kV at 200 mA, the pitch was 0.9:1, and the field of view (FOV) was 14 to 15 cm. The standard inspection position was used in all cases. In this position, the long axis of the examined knee joint was parallel to the long axis of the inspection bed, the patient’s feet were faced upward, and the midline of the foot (formed by the third metatarsal bone and the middle point of the posterior edge of the calcaneus) was perpendicular to the surface of the inspection bed. The original thickness of the images was 5 mm, and the reconstruction matrix was 512×512. The original data was reconstructed to a 0.6 mm image thickness. The MIP technique was the most commonly used recombination method.
The specific measurement process is shown in Figure 1. The knee-joint image for each case was first processed using the MIP technique on the syngo MMWP VE40C image post-processing workstation. The workstation produced images similar to the anteroposterior and lateral knee-joint X-ray images, respectively (Figure 1a). Then, the lateral image was adjusted to the required angle while keeping the tibia at the front and the fibula at the back. The proximal knee was tilted at 60° to the dorsal side to allow the distal femur and the medial and lateral femoral condyles to align with the knee in the conventional lateral knee-joint X-ray image [8, 11]; the angle between the distal femur and the vertical line of the ground in the down direction was 120° (Figure 1b). The anatomical positional relationship of the femur’s medial and lateral condyles was observed on the anteroposterior and lateral images, using it as a reference image.
According to the anatomical positional relationship of the upper/lower directions and the front/rear directions, the reference image was adjusted to a specific inward rotation (angle A) and an outward tilt (angle B) to achieve a true standard lateral knee-joint image of the perfectly imposed posterior condyles of the distal femoral epiphysis (Figure 1c). During the adjustment process, angles A and B were observed and analyzed by a radiologist with five years’ experience and a radiographer with more than five years’ experience. The final angle values were confirmed by the clinicians’ consensus and were allocated to group A and group B, respectively.
Clinical practice and comparative analysis
The acquired adjustment direction and angles for the standard lateral knee joints were applied in clinical lateral knee-joint radiography. The outward tilt angle was determined by the anatomical positional relationship of the upper end of the femur’s medial and lateral posterior condyles on the anteroposterior images. The inward rotation angle was determined by touch to define the anatomical positional relationship between the lower end of the femur’s medial and lateral posterior condyles. The captured lateral knee-joint X-ray images by optimized position were compared with ones captured by the conventional lateral knee-joint radiography position. The image-quality scores of the two groups were rated by the radiologist and the radiographer mentioned above. The final scores were confirmed by their accord and were allocated to group O (optimized group) and group T (conventional group), respectively. The scoring standard was (1) five points for a perfect superimposition of the posterior condyles of the distal femoral epiphysis and a clear presentation of patellofemoral clearance, (2) three points for the perfect superimposition of the posterior condyles of the distal femoral epiphysis or the clear presentation of patellofemoral clearance, (3) and one point for the poor superimposition of the posterior condyles of the distal femoral epiphysis and an unclear presentation of patellofemoral clearance.
IBM SPSS 22.0 statistics software was used for the statistical analysis. The normal distribution data was represented by the mean ± standard deviation (X ± S), and the abnormal distribution data was represented by median and quartile values. When comparing image-quality scores of the knee-joint x-ray images in the O and T groups, and when comparing the optimal deflection angles for the projection of the knee joint between males and females or between left and right, independent sample t-tests were used if the data were normal and if the variances were homogeneous. The Mann–Whitney U test was applied if the data were not normally distributed or showed variance.
The study was approved by the local ethics committee (2020-082-2)