New Parametric Imaging Method with Fluorescein Angiograms for Detecting Areas of Capillary Nonperfusion / 대한의료정보학회지
Healthcare Informatics Research
;
: 191-198, 2014.
Artigo
em Inglês
| WPRIM
| ID: wpr-76101
ABSTRACT
OBJECTIVES:
Fluorescein angiography (FAG) is currently the most useful diagnostic modality for examining retinal circulation, and it is frequently used for the evaluation of patients with diabetic retinopathy, occlusive diseases, such as retinal venous and arterial occlusions, and wet macular degeneration. This paper presents a method for objectively evaluating retinal circulation by quantifying circulation-related parameters.METHODS:
This method allows the semiautomatic preprocessing and registering of FAG images. The arterial input function is estimated from the registered set of FAG images using gamma-variate fitting. Then, the parameters can be computed by deconvolution on the basis of truncated singular value decomposition, and they can finally be presented as parametric color images in a combination of three colors, red, green, and blue.RESULTS:
After the estimation of arterial input function, the parameters of relative blood flow and mean transit time were computed using deconvolution analysis based on truncated singular value decomposition.CONCLUSIONS:
The parametric color image is helpful to interpret the status of retinal blood circulation and provides quantitative data on retina ischemia without interobserver variability. This system easily provides the status of retinal blood circulation both qualitatively and quantitatively. It also helps to standardize FAG interpretation and may contribute to network-based telemedicine systems in the future.
Texto completo:
DisponíveL
Índice:
WPRIM (Pacífico Ocidental)
Assunto principal:
Oftalmologia
/
Retina
/
Retinaldeído
/
Engenharia Biomédica
/
Circulação Sanguínea
/
Capilares
/
Angiofluoresceinografia
/
Variações Dependentes do Observador
/
Diagnóstico por Computador
/
Telemedicina
Tipo de estudo:
Estudo diagnóstico
Limite:
Humanos
Idioma:
Inglês
Revista:
Healthcare Informatics Research
Ano de publicação:
2014
Tipo de documento:
Artigo
Similares
MEDLINE
...
LILACS
LIS