98
L. Fang and S. Li
resolution of ~4.5 μm per pixel in tissue. In addition, we also performed our experiments on a mouse dataset acquired by a different SDOCT system, (Bioptigen Envisu
R2200), with ~2 μm axial resolution in tissue.
Based on our experiments on training data, we empirically selected the parameters
for the proposed 3D-ASRC algorithm. We chose the patch size in each slice to be
a rectangle of size 6 × 12 pixels (height × width). The number of nearby slices T
was set to 5 (corresponding to ~300 μ azimuthal distance). In retinal imaging, slices
from farther distances may have significant differences and thus adding them might
actually reduce compression efficiency. In the dictionary training stage, the value of
cluster H was chosen to be 10. In each cluster, the size of the trained dictionary was
set to 72 × 500. The parameter b in (4.20) was set to 0.001. For the test datasets in our
experiments, the mean and standard deviation of parameter C for the compression
ratios [10, 15, 20, 25, 30, 35, 40] were [1.00, 1.07, 1.12, 1.15, 1.17, 1.19, 1.21],
and [0.053, 0.054, 0.052, 0.055, 0.055, 0.057, 0.059], respectively. The parameters
for the JPEG 2000 and MPEG-4 were set to the default values in the Matlab [43]
and QuickTime Player Pro 7.0 software [61], respectively. For the K-SVD algorithm,
the patch size was set to 6 × 12 and the trained dictionary was of size 72 × 500. For
the 2D-ASRC method, the number of spatial nearby patches was selected to 9 and
the other parameters were set to the same values as in our 3D-ASRC method. We
adopted the peak signal-to-noise-ratio (PSNR) and feature similarity index measure
(FSIM) [62] to evaluate the performances of the compression methods.
We tested the JPEG 2000, MPEG-4, SPIHT [60], K-SVD [30], SRC-Dif, 2DASRC, 3D-ASRC-WA, and 3D-ASRC methods on seven different compression
ratios ranging from 10 to 40. The corresponding quantitative comparisons (PSNR
and FSIM) of all the test methods at different compression ratios are reported in
Table 4.3. As can be seen in Table 4.2, the proposed 3D-ASRC method consistently
delivered better PSNR and FSIM results than the other methods. Figures 4.17 show
qualitative comparisons of reconstructed results from the tested methods using the
compression ratio of 10. Since boundaries between retinal layers and drusen contain
meaningful anatomic and pathologic information [44], we magnified one boundary
area and one dursen area in each figure. As can be observed in Fig. 4.17, results
from JPEG 2000, K-SVD, and MPEG-4 methods appear very noisy with indistinct
boundaries for many important structural details (see the zoomed boundary areas of
both the dataset 1 and 2). The SPIHT method greatly suppresses noise, but increases
blur and introduces visible artifacts (see the zoomed drusen and boundary areas in
Fig. 4.17). Compared to the above methods, the proposed SRC-Dif, 2D-ASRC, 3DASRC-WA methods deliver comparatively better structural details, but still show
some noise artifacts. By contrast, the proposed 3D-ASRC method achieves noticeably improved noise suppression, and preserves meaningful anatomical structures.
L. Fang and S. Li
resolution of ~4.5 μm per pixel in tissue. In addition, we also performed our experiments on a mouse dataset acquired by a different SDOCT system, (Bioptigen Envisu
R2200), with ~2 μm axial resolution in tissue.
Based on our experiments on training data, we empirically selected the parameters
for the proposed 3D-ASRC algorithm. We chose the patch size in each slice to be
a rectangle of size 6 × 12 pixels (height × width). The number of nearby slices T
was set to 5 (corresponding to ~300 μ azimuthal distance). In retinal imaging, slices
from farther distances may have significant differences and thus adding them might
actually reduce compression efficiency. In the dictionary training stage, the value of
cluster H was chosen to be 10. In each cluster, the size of the trained dictionary was
set to 72 × 500. The parameter b in (4.20) was set to 0.001. For the test datasets in our
experiments, the mean and standard deviation of parameter C for the compression
ratios [10, 15, 20, 25, 30, 35, 40] were [1.00, 1.07, 1.12, 1.15, 1.17, 1.19, 1.21],
and [0.053, 0.054, 0.052, 0.055, 0.055, 0.057, 0.059], respectively. The parameters
for the JPEG 2000 and MPEG-4 were set to the default values in the Matlab [43]
and QuickTime Player Pro 7.0 software [61], respectively. For the K-SVD algorithm,
the patch size was set to 6 × 12 and the trained dictionary was of size 72 × 500. For
the 2D-ASRC method, the number of spatial nearby patches was selected to 9 and
the other parameters were set to the same values as in our 3D-ASRC method. We
adopted the peak signal-to-noise-ratio (PSNR) and feature similarity index measure
(FSIM) [62] to evaluate the performances of the compression methods.
We tested the JPEG 2000, MPEG-4, SPIHT [60], K-SVD [30], SRC-Dif, 2DASRC, 3D-ASRC-WA, and 3D-ASRC methods on seven different compression
ratios ranging from 10 to 40. The corresponding quantitative comparisons (PSNR
and FSIM) of all the test methods at different compression ratios are reported in
Table 4.3. As can be seen in Table 4.2, the proposed 3D-ASRC method consistently
delivered better PSNR and FSIM results than the other methods. Figures 4.17 show
qualitative comparisons of reconstructed results from the tested methods using the
compression ratio of 10. Since boundaries between retinal layers and drusen contain
meaningful anatomic and pathologic information [44], we magnified one boundary
area and one dursen area in each figure. As can be observed in Fig. 4.17, results
from JPEG 2000, K-SVD, and MPEG-4 methods appear very noisy with indistinct
boundaries for many important structural details (see the zoomed boundary areas of
both the dataset 1 and 2). The SPIHT method greatly suppresses noise, but increases
blur and introduces visible artifacts (see the zoomed drusen and boundary areas in
Fig. 4.17). Compared to the above methods, the proposed SRC-Dif, 2D-ASRC, 3DASRC-WA methods deliver comparatively better structural details, but still show
some noise artifacts. By contrast, the proposed 3D-ASRC method achieves noticeably improved noise suppression, and preserves meaningful anatomical structures.