Author Affiliations
1 School of Mechanical Engineering and Automation, Shanghai University, Shanghai 200444, China2 Laboratory of Information Optics and Opto-Electronic Technology, Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800, Chinashow less
Fig. 1. Flow chart of proposed measurement method of corneal thickness
Fig. 2. Noise and artifacts in original cornea B scan image
Fig. 3. A-scan averaged intensity curve
Fig. 4. Average intensity curves of adjacent rows
Fig. 5. Effect of corneal B-scan OCT images after pre-treatment. (a) Original corneal B-scan image; (b) corneal B-scan image with adaptive median-filtering; (c) corneal B-scan image with horizontal artifact removal; (d) corneal B-scan image with suppressed central artifact
Fig. 6. Boundary conditions. (a) Outer boundary condition; (b) hole boundary condition
Fig. 7. Upper and lower edge fitting results of human corneal OCT B-scan image. (a) High quality original corneal B-scan image; (b) high quality edge fitting corneal B-scan image; (c) original corneal B-scan image with noise and artifacts; (d) edge fitting corneal B-scan image with noise and artifacts
Fig. 8. Fitting effects of upper and lower edges of human corneal B-scan images obtained by different pre-processing denoising algorithms. (a) No preprocessing denoising. (b) mean filter preprocessing; (c) median filter preprocessing; (d) adaptive median filter preprocessing
Fig. 9. Results of upper and lower edge fitting of OCT corneal images. (a) Edge detection and random sampling consistency method; (b) proposed method
Algorithm type | Ta±σ /μm | Tc±σ /μm |
---|
High quality corneal B-scan image | 561.6±1.2 | 571.6±2.9 | Corneal B-scan image with noise and artifacts | 562.1±2.3 | 572.3±3.8 |
|
Table 1. Average thickness, central thickness and corresponding standard deviations along Y-axis of corneal images with different qualities
Preprocessing algorithm | Ta±σ /μm | Tc±σ /μm |
---|
No preprocessing denoising | 567.2±5.1 | 575.3±5.4 | Mean filter preprocessing | 563.1±4.4 | 573.9±4.8 | Median filter preprocessing | 564.6±3.8 | 573.2±4.5 | Adaptive median filter preprocessing | 562.1±2.3 | 572.3±3.8 | Manual measurement | 561.4±1.4 | 571.2±2.4 |
|
Table 2. Average thickness, central thickness and corresponding standard deviations along Y-axis after treatments with different pretreatment algorithms
Algorithm type | Proposed algorithm (manual measurement) | Random sampling consisitency (RANSC) algorithm (manual measurement) |
---|
Ta±σ /μm | 1.0±0.3 | 2.3±1.8 | Tc±σ /μm | 1.2±0.6 | 5.2±3.9 |
|
Table 3. Average thickness deviation, corneal center thickness deviation and corresponding standard deviations along Y-axis of corneal by different methods