Table 1 Comparison between this paper and other literatures

Moglo [41]

Non-linear 3D finite element model

Meshing, Definition of material properties

 − 5° to 90°

 − 1° to 1°

 − 16° to 0.5°

Avoiding the simultaneous consideration of large rotations about different axes on a single body

Caution should be exercised while extending the findings to the knee joint in living subjects in presence of different loads

Chen [43]

Dual fluoroscopic and MR image technique

Create 3D computer models, image registration

0–50°

 − 2° to 3°

/

Being accurate, non-invasive

The view of the two fluoroscopes is limited

Kozanek [44]

Dual fluoroscopic imaging system

Create 3D computer models, image registration

0–40°

 − 6° to 2.5°

 − 7.5° to 0

Being accurate, non-invasive and no-requiring placement of external devices or markers on the knee

The relatively slow walking speed (0.67 m/s), the view of the two fluoroscopes is limited

Mizuno et al. [45]

The dynamic knee simulator

Standardized radiographic technique

0°–90°

1°–6°

 − 7° to − 4°

Performing stable knee flexion under heavy muscle load

The cadaver model exactly no-representing anatomic Q-angle variations or surgical procedures modifying the Q-angle

Merican and Amis [42]

A Polaris optical tracking system

Thawed, kept moist with normal saline

0–70°

 − 2° to 5°

 − 10° to 4°

Being applied to the clinical on the effects of lateral retinacular releases to correct patellar mal-tracking

Elderly knees

Resultsof this paper

MATLAB

Pre-program and patch program

(0°–90°)*

(− 14° to 7°)*

(− 30° to 2°)*

The patch program developed in MATLAB making reasonable compensation for the undetectable points

Data preprocessing is not as fast as commercial software

(0–45°)**

(− 6° to 7°)**

(− 14° to 0°)**