Classify in vivo confocal microscopy corneal images by tortuosity is complicated by the presence of variable numbers of fibres of different tortuosity level. Instead of designing a function combining manually selected features into a single coefficient, as done in the literature, we propose a supervised approach which selects automatically the most relevant combination of shape features from a pre-defined dictionary. To our best knowledge, we are the first to consider features at different spatial scales and show experimentally their relevance in tortuosity modelling. Our results, obtained with a set of 100 images and 20 fold cross-validation, suggest that multinomial logistic ordinal regression, trained on consensus ground truth from 3 experts, yields an accuracy indistinguishable, overall, from that of experts when compared against each other.
|Title of host publication||Proceedings of the Ophthalmic Medical Image Analysis First International Workshop, OMIA 2014, Held in Conjunction with MICCAI 2014|
|Editors||Xinjian Chen, Mona K. Garvin, Jimmy J. Liu|
|Publisher||Iowa Research Online|
|Number of pages||8|
|Publication status||Published - 2014|
|Event||Ophthalmic Medical Image Analysis First International Workshop - Boston, United States|
Duration: 14 Sep 2014 → 14 Sep 2014
|Conference||Ophthalmic Medical Image Analysis First International Workshop|
|Abbreviated title||OMIA 2014|
|Period||14/09/14 → 14/09/14|
Annunziata, R., Kheirkhah, A., Aggarwal, S., Cavalcanti, B. M., Hamrah, P., & Trucco, E. (2014). Tortuosity classification of corneal nerves images using a multiple-scale-multiple-window approach. In X. Chen, M. K. Garvin, & J. J. Liu (Eds.), Proceedings of the Ophthalmic Medical Image Analysis First International Workshop, OMIA 2014, Held in Conjunction with MICCAI 2014 (pp. 113-120). Iowa Research Online.