coronal plane alignment of the knee(CPAK) distribution of 246 cases(477 knees) of OA patients and 107 cases(214 knees) of healthy people were examined on long-leg radiographs retrospectively. Radiological measurements and CPAK classification on different Kellgren- Lawrence(K-L) Grade in patients with unilateral total knee arthroplasty(TKA) were compared. Clinical outcomes of CPAK type I patients performed by mechanical alignment(MA) and restricted kinematic alignment(rKA) during TKA were examined. ML algorithm including K nearest neighbors (KNN), random forest(RF), artificial neural networks (ANN), logistic regression(LR) and gradient boosting (GBDT) were established, the dependent variable was set as whether the constitutional phenotype of an arthritic knee classified as type I was type I.

Study groups and design OA group Data for consecutive OA patients with knee pain who underwent primary TKA by one senior surgeon between August 2021 and July 2023 at a single institution were analyzed retrospectively. A total of 273 patients underwent long-leg radiographs. The following 27 patients were excluded from this study: prior total hip arthroplasty, 2; obvious bony deficiency of the femur or tibia, 10; simultaneous flexion of the knee and rotation of the leg on radiographs, 15. CPAK distribution of the remaining 246 cases(unilateral, 15; bilateral, 231; totally 477 knees) was examined.

healthy group Data for consecutive visitors at outpatient clinic who underwent long-leg radiographs but without any sign of cartilage degeneration or medical history of low extremity between January 2023 and July 2023 at the same institution were analyzed retrospectively. A total of 136 visitors were recruited. The following 18 visitors were excluded from this study: extra-articular deformity of the femur or tibia, 15; simultaneous flexion of the knee and rotation of the leg on radiographs, 13; poor quality image, 1. CPAK distribution of the remaining 107 cases(214 knees) was examined.

Radiological measurements All participants underwent standard digital long leg radiographs. The mechanical axis of the femur was marked from the centre of the femoral head to the centre of the knee. The centre of the head was marked using the concentric-circle method to identify the centre. The centre of the ankle was marked as the point on the talar dome at mid-width. The mHKA angle was the angle subtended by the mechanical axes of the femur and tibia. The LDFA was defined as the lateral angle formed between the femoral mechanical axis and the joint line of the distal femur. The MPTA was defined as the medical angle formed between the tibial mechanical axis and the joint line of the proximal tibia. Joint line convergence angle(JLCA) was the angle formed between joint orientation lines on opposite sides of the same joint[17]. All measurements were carried out by an orthopaedic fellow. A senior author undertook a subgroup analysis of 60 knees in OA group and repeated at one- week intervals to assess for inter- and intra- observer agreement.

CPAK classification of OA and healthy group aHKA was calculated based on the following formula: aHKA=MPTA-LDFA. JLO was calculated by the formula: JLO=MPTA+LDFA. With aHKA and JLO measured, patients can be matched to 9 possible CPAK alignment groups. The mean aHKA and JLO of the two groups were rounded to the nearest whole number for final allocation to a CPAK Class. CPAK limits for the definition of neutral knees was an aHKA of 0 ± 2 degrees. CPAK boundaries for a neutral JLO were defined as 180 ± 3 degrees, varus aHKA less than-2 degrees and a valgus aHKA more than+2 degrees. An apex distal JLO was less than 177°, while an apex proximal JLO was greater than 183°[9]. The OA and healthy group were categorized according to the CPAK classification, and their distribution were investigated.

Inclusion Criteria:

- OA group: Data for consecutive OA patients with knee pain who underwent primary TKA by one senior surgeon between August 2021 and July 2023 at a single institution were analyzed retrospectively.

Healthy group: Data for consecutive visitors at outpatient clinic who underwent long-leg radiographs but without any sign of cartilage degeneration or medical history of low extremity between January 2023 and July 2023 at the same institution were analyzed retrospectively.

Exclusion Criteria:

- OA group: prior total hip arthroplasty; obvious bony deficiency of the femur or tibia, simultaneous flexion of the knee and rotation of the leg on radiographs Healthy group: extra-articular deformity of the femur or tibia; simultaneous flexion of the knee and rotation of the leg on radiographs; poor quality image

OA group: consecutive OA patients with knee pain who underwent primary TKA Healthy group: consecutive visitors at outpatient clinic who underwent long-leg radiographs but without any sign of cartilage degeneration or medical history of low extremity