Limits...
Designing and testing scene enhancement algorithms for patients with retina degenerative disorders.

Al-Atabany WI, Memon MA, Downes SM, Degenaar PA - Biomed Eng Online (2010)

Bottom Line: To test our techniques, we have therefore built a retinal image processing model and compared the results to data from patient testing.Results from our model are compared to actual data obtained from testing the performance of these algorithms on 27 patients with an average visual acuity of 0.63 and an average contrast sensitivity of 1.22.Image Cartoonization was most beneficial for spatial feature detection such as face detection.

View Article: PubMed Central - HTML - PubMed

Affiliation: Institute of Biomedical Engineering, Imperial College, South Kensington, London, UK. walid.atbany06@imperial.ac.uk

ABSTRACT

Background: Retina degenerative disorders represent the primary cause of blindness in UK and in the developed world. In particular, Age Related Macular Degeneration (AMD) and Retina Pigmentosa (RP) diseases are of interest to this study. We have therefore created new image processing algorithms for enhancing the visual scenes for them.

Methods: In this paper we present three novel image enhancement techniques aimed at enhancing the remaining visual information for patients suffering from retina dystrophies. Currently, the only effective way to test novel technology for visual enhancement is to undergo testing on large numbers of patients. To test our techniques, we have therefore built a retinal image processing model and compared the results to data from patient testing. In particular we focus on the ability of our image processing techniques to achieve improved face detection and enhanced edge perception.

Results: Results from our model are compared to actual data obtained from testing the performance of these algorithms on 27 patients with an average visual acuity of 0.63 and an average contrast sensitivity of 1.22. Results show that Tinted Reduced Outlined Nature (TRON) and Edge Overlaying algorithms are most beneficial for dynamic scenes such as motion detection. Image Cartoonization was most beneficial for spatial feature detection such as face detection. Patient's stated that they would most like to see Cartoonized images for use in daily life.

Conclusions: Results obtained from our retinal model and from patients show that there is potential for these image processing techniques to improve visual function amongst the visually impaired community. In addition our methodology using face detection and efficiency of perceived edges in determining potential benefit derived from different image enhancement algorithms could also prove to be useful in quantitatively assessing algorithms in future studies.

Show MeSH

Related in: MedlinePlus

Efficacy of enhancement algorithms on the simulated pelli Robson's images at different object contrasts. (a-c) The efficacy of each algorithm compared to the original image on an image with object of 2.26, 1.66 and 0.75 contrasts, respectively, based on the Pelli Robson's method with respect to the eccentricity from the macula.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2914026&req=5

Figure 9: Efficacy of enhancement algorithms on the simulated pelli Robson's images at different object contrasts. (a-c) The efficacy of each algorithm compared to the original image on an image with object of 2.26, 1.66 and 0.75 contrasts, respectively, based on the Pelli Robson's method with respect to the eccentricity from the macula.

Mentions: In order to relate our simulator with the patient results, we tested our algorithms on the Pelli Robson's contrast sensitivity method [43] using our model for validation. Based on this method, we developed 16 (800 × 800) images with a white background and a gray box (of 44% the diameter of macula) with contrast ranging from 0 to 2.26. Each image is repeated 14 times to simulate the effect of eccentricity from the center of the macula. The eccentricity step was 30 pixels (equivalent to 0.26 of the macula's diameter). Degenerated versions of these images were developed from the simulator with a virtual scotoma of the same macula's size added to the center. Figure 8 shows a sample of image with contrast of 1.8 and eccentricity of 12.74 mm from the centre of the macula. Processed versions of these images have been generated using our three algorithms. The percentages of extracted edges have been calculated for the degenerated images before and after enhancement. Figure 9(a-c), shows the percentages of extracted edges for the unprocessed and processed image at contrasts of 2.26, 1.66 and 0.75 respectively. We can see that there is not much difference between the processed and unprocessed image of high contrast. The efficacy of the processed image over the unprocessed ones increases while decreasing the object's contrast as shown in Figure 9(b-c). This observation is clearly shown in Figure 10, which shows the percentage of extracted edges at eccentricity of 5.46 mm with respect to the patient CS (which opposite to the object's contrast here). To illustrate more what Figure 10 shows, the percentage of extracted edges at CS of 0.75 will be increased from 4.5 to 16.5. There is not much difference between the unprocessed and processed image neither at very low nor very high object contrasts. This is because the image processing algorithms we have used have difficulty in detecting very low contrast features. For high contrast objects, the enhancement algorithms will not add more detail to the object recognition process. However, the effort needed to recognize the processed image over the unprocessed is still decreased.


Designing and testing scene enhancement algorithms for patients with retina degenerative disorders.

Al-Atabany WI, Memon MA, Downes SM, Degenaar PA - Biomed Eng Online (2010)

Efficacy of enhancement algorithms on the simulated pelli Robson's images at different object contrasts. (a-c) The efficacy of each algorithm compared to the original image on an image with object of 2.26, 1.66 and 0.75 contrasts, respectively, based on the Pelli Robson's method with respect to the eccentricity from the macula.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2914026&req=5

Figure 9: Efficacy of enhancement algorithms on the simulated pelli Robson's images at different object contrasts. (a-c) The efficacy of each algorithm compared to the original image on an image with object of 2.26, 1.66 and 0.75 contrasts, respectively, based on the Pelli Robson's method with respect to the eccentricity from the macula.
Mentions: In order to relate our simulator with the patient results, we tested our algorithms on the Pelli Robson's contrast sensitivity method [43] using our model for validation. Based on this method, we developed 16 (800 × 800) images with a white background and a gray box (of 44% the diameter of macula) with contrast ranging from 0 to 2.26. Each image is repeated 14 times to simulate the effect of eccentricity from the center of the macula. The eccentricity step was 30 pixels (equivalent to 0.26 of the macula's diameter). Degenerated versions of these images were developed from the simulator with a virtual scotoma of the same macula's size added to the center. Figure 8 shows a sample of image with contrast of 1.8 and eccentricity of 12.74 mm from the centre of the macula. Processed versions of these images have been generated using our three algorithms. The percentages of extracted edges have been calculated for the degenerated images before and after enhancement. Figure 9(a-c), shows the percentages of extracted edges for the unprocessed and processed image at contrasts of 2.26, 1.66 and 0.75 respectively. We can see that there is not much difference between the processed and unprocessed image of high contrast. The efficacy of the processed image over the unprocessed ones increases while decreasing the object's contrast as shown in Figure 9(b-c). This observation is clearly shown in Figure 10, which shows the percentage of extracted edges at eccentricity of 5.46 mm with respect to the patient CS (which opposite to the object's contrast here). To illustrate more what Figure 10 shows, the percentage of extracted edges at CS of 0.75 will be increased from 4.5 to 16.5. There is not much difference between the unprocessed and processed image neither at very low nor very high object contrasts. This is because the image processing algorithms we have used have difficulty in detecting very low contrast features. For high contrast objects, the enhancement algorithms will not add more detail to the object recognition process. However, the effort needed to recognize the processed image over the unprocessed is still decreased.

Bottom Line: To test our techniques, we have therefore built a retinal image processing model and compared the results to data from patient testing.Results from our model are compared to actual data obtained from testing the performance of these algorithms on 27 patients with an average visual acuity of 0.63 and an average contrast sensitivity of 1.22.Image Cartoonization was most beneficial for spatial feature detection such as face detection.

View Article: PubMed Central - HTML - PubMed

Affiliation: Institute of Biomedical Engineering, Imperial College, South Kensington, London, UK. walid.atbany06@imperial.ac.uk

ABSTRACT

Background: Retina degenerative disorders represent the primary cause of blindness in UK and in the developed world. In particular, Age Related Macular Degeneration (AMD) and Retina Pigmentosa (RP) diseases are of interest to this study. We have therefore created new image processing algorithms for enhancing the visual scenes for them.

Methods: In this paper we present three novel image enhancement techniques aimed at enhancing the remaining visual information for patients suffering from retina dystrophies. Currently, the only effective way to test novel technology for visual enhancement is to undergo testing on large numbers of patients. To test our techniques, we have therefore built a retinal image processing model and compared the results to data from patient testing. In particular we focus on the ability of our image processing techniques to achieve improved face detection and enhanced edge perception.

Results: Results from our model are compared to actual data obtained from testing the performance of these algorithms on 27 patients with an average visual acuity of 0.63 and an average contrast sensitivity of 1.22. Results show that Tinted Reduced Outlined Nature (TRON) and Edge Overlaying algorithms are most beneficial for dynamic scenes such as motion detection. Image Cartoonization was most beneficial for spatial feature detection such as face detection. Patient's stated that they would most like to see Cartoonized images for use in daily life.

Conclusions: Results obtained from our retinal model and from patients show that there is potential for these image processing techniques to improve visual function amongst the visually impaired community. In addition our methodology using face detection and efficiency of perceived edges in determining potential benefit derived from different image enhancement algorithms could also prove to be useful in quantitatively assessing algorithms in future studies.

Show MeSH
Related in: MedlinePlus