A complementary method for automated detection of microaneurysms in fluorescein angiography fundus images to assess diabetic retinopathy

作者：

Highlights：

• Algorithm is not based on morphological functions in order to find microaneurysms.

• Algorithm can find microaneurysms for evaluating diabetic retinopathy.

• Algorithm is robust to noise because of its linear integral transformation concept.

• Identifying the retinal landmarks improves the classification of microaneurysms.

摘要

•Algorithm is not based on morphological functions in order to find microaneurysms.•Algorithm can find microaneurysms for evaluating diabetic retinopathy.•Algorithm is robust to noise because of its linear integral transformation concept.•Identifying the retinal landmarks improves the classification of microaneurysms.

论文关键词：Computer aided diagnosis,Diabetic retinopathy,Fluorescein angiography,Fundus images,Radon transform,Microaneurysms

论文评审过程：Received 3 October 2011, Revised 28 December 2012, Accepted 13 March 2013, Available online 26 March 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.03.011