Detection of Anatomic Structures in Retinal Images

Pinão, José; Oliveira, Carlos Manta; Mora, André; Dias, João M.S.

doi:10.1007/978-94-007-0726-9_13

Cited by 2 publications

(2 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The image artifacts described before might impact negatively the results of automated diagnostic systems based on retinal image processing [6][7][8][9]. The presence of brighter or Special Sessions Retinal Image Analysis darker areas can be confused with pathological retinal abnormalities and produce false positive detections.…”

Section: Introductionmentioning

confidence: 95%

A template matching technique for artifacts detection in retinal images

Mora¹,

Soares²,

Fonseca³

2013

2013 8th International Symposium on Image and Signal Processing and Analysis (ISPA)

View full text Add to dashboard Cite

The continuous development of automatic retinal diseases diagnosis systems based on image processing has shown their potential for clinical practice. However, the accuracy of these systems is often compromised, mainly due to the intrinsic difficulty in detecting the abnormal structures and also due to deficiencies in the image acquisition which affects image quality. Light flares are one of such deficiencies that usually don't compromise the overall image quality, but can be misclassified by an automatic diagnosis system. In this article a method is proposed for detecting light artifacts (flares) on retinal images. The output from the light artifact detection is a binary image mask that is useful to reject those pixels from being further processed. The proposed method uses a template matching algorithm to detect artifacts similar to the predefined template artifact images. Two main types of light artifacts were identified: light flares and the central artifact. To reduce over-segmentation the light artifact candidates are characterized by their shape and color and are classified by a decision tree. The method was developed using a dataset of 61 images from which 20 were used for the classifier training and the remaining 41 for independent testing. With the test dataset the method obtained an average sensitivity/false detection per image pairs of 0.97/0.12 for the central artifact and 0.73/0.36 for the light flares, what were considered good results regarding the heterogeneity of the dataset which contain low and high quality images.

show abstract

Section: Introductionmentioning

confidence: 95%

A template matching technique for artifacts detection in retinal images

Mora¹,

Soares²,

Fonseca³

2013

2013 8th International Symposium on Image and Signal Processing and Analysis (ISPA)

View full text Add to dashboard Cite

show abstract

“…Como estes artefactos são causados pela reflexão da luz ou pela sujidade ao longo do sistema ótico, podem variar em número e intensidade (Burke, 1996). Apesar de poder parecer inofensiva, a ocorrência de artefactos nas imagens de fundoscopia pode ter um impacto significativo nos resultados dos sistemas de diagnóstico automático (Davis et al, 2009;Mora et al, 2011;Pinão et al, 2013;Yu et al, 2012) podendo as áreas da imagem com alta ou baixa luminosidade ser facilmente confundidas com patologias, aumentando assim o número de falsos positivos.…”

Section: Introductionunclassified

Metodologia para a deteção de artefactos luminosos em imagens de retinografia com aplicação em rastreio oftalmológico

Mora¹,

Fonseca²

2014

RISTI

View full text Add to dashboard Cite

Resumo: O diagnóstico automático de doenças da retina baseado em processamento de imagem vê frequentemente a sua precisão comprometida pela dificuldade intrínseca na deteção de estruturas anormais e por deficiências na aquisição de imagem. Em cenários de rastreio, estas deficiências podem levar a que um número significativo de imagens tenham de ser repetidas, implicando custos e ineficiência do sistema. Neste artigo propõe-se uma metodologia que avalia de forma automática a qualidade das imagens captadas possibilitando ao operador repetir a aquisição caso se justifique. O método proposto identifica diferentes tipos de artefactos com base na sua forma, cor e intensidade. Utilizando um conjunto de 61 imagens foi obtida uma sensibilidade de 97% com 0.12 falsos positivos por imagem na deteção do artefacto central e 73% de sensibilidade com 0.36 de falsos positivos para os reflexos luminosos. Estes resultados podem considerar-se positivos tendo em conta a baixa qualidade e a heterogeneidade das imagens processadas.Palavras-chave: classificação automática; processamento de imagem; imagem da retina; reflexos luminosos; rastreio oftalmológico. Methodology for image artifacts detection in retinal images with application in Ophthalmology ScreeningAbstract: Automatic diagnostic systems for retinal diseases based on image processing have continuously demonstrated its potential for clinical practice. However, their accuracy is often compromised by the inherent difficulty in detecting abnormal structures and aggravated by deficiencies in image acquisition. In screening scenarios, these deficiencies can lead to a significant amount of repeated images, implying costs and system's inefficiency. In this paper we propose a methodology to automatically evaluate the quality of captured images allowing the operator to repeat the acquisition if appropriate. The proposed method identifies different types of artefacts based on color, shape and image intensity. Using a set of 61 images a sensitivity of 97% at a rate of 0.12 false

show abstract

Detection of Anatomic Structures in Retinal Images

Cited by 2 publications

References 39 publications

A template matching technique for artifacts detection in retinal images

A template matching technique for artifacts detection in retinal images

Metodologia para a deteção de artefactos luminosos em imagens de retinografia com aplicação em rastreio oftalmológico

Contact Info

Product

Resources

About