Image Analytics Techniques for Diabetic Retinopathy Detection
Resumo
Diabetic Retinopathy is the leading cause of blindness in working- age adults. This work aims at enhancing lesion detection, reinforcing referral decisions, and integrating our solutions with low-cost retinal imaging devices. For lesion detection, we proposed a novel coding technique robust to any kind oflesion. For referral decisions, we designed a robust method that does not rely upon lesion detection, proposed an effective data-driven model that significantly improves the performance, designed an accountable model that produces a re- liable response and enables pixel-based importance comprehension, and create local descriptors that are encoded into a rich mid-level representation. Our work has invaluable impacts both in biomedical and technical contexts.
Referências
Pires, R., Avila, S., Jelinek, H. F., Wainer, J., Valle, E., and Rocha, A. (2017). Beyond lesion-based diabetic retinopathy: a direct approach for referral. IEEE Journal of Biomedical and Health Informatics, 21(1):193–200.
Pires, R., Avila, S., Wainer, J., Valle, E., Abr`amoff, M. D., and Rocha, A. (2019a). A data-driven approach to referable diabetic retinopathy detection. Artificial Intelligence in Medicine, 96:93–106.
Pires, R., Carvalho, T., Spurling, G., Goldenstein, S., Wainer, J., Luckie, A., Jelinek, H. F., and Rocha, A. (2015). Automated multi-lesion detection for referable diabetic retinopathy in indigenous health care. PLoS ONE, 10(6):e0127664.
Pires, R., Ferreira, A., Avila, S., Wainer, J., and Rocha, A. (2019b). An accountable saliency-oriented data-driven approach to diabetic retinopathy detection. In Elsevier, editor, Photo Acoustic and Optical Coherence Tomography Imaging: An Application in Ophthalmology, chapter 10. Elsevier.
Pires, R., Jelinek, H. F., Wainer, J., Valle, E., and Rocha, A. (2014b). Advancing bag- of-visual-words representations for lesion classification in retinal images. PLoS ONE, 9(6):e96814.
