.png){kind=icon}
.png){kind=icon}
Return
Homotopy-based image enhancement techniques for medical imaging applications
.png){kind=icon}
PRE
AI
S
M
DOI：10.1007/s13198-026-03195-y.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
A broad spectrum of prevalent conditions associated with imaging and cardiovascular disorders can be effectively identified through advanced image analysis. However, challenges such as low visual saliency, uneven lighting, and obscured images complicate the detection and assessment of low-quality images. To address these challenges and improve visual saliency while assessing image quality in the contexts of compression, transmission, and reconstruction, we introduce a novel enhancement technique termed the linear homotopic parametric value approach. This method has yielded promising outcomes when applied to the structured analysis of publicly accessible image datasets, including MESSIDOR and STARE. Notably, our approach achieved a peak signal-to-noise ratio of 49.9214 and a structured similarity index of 0.9999, outperforming previous studies. These enhancements result in images that are significantly clearer and of higher qualitative value. Our algebraic topology-based enhancement model has been rigorously evaluated across seven distinct quality metrics, consistently achieving superior results and surpassing existing state-of-the-art methodologies, as demonstrated in a comprehensive comparative analysis.
Keywords:
Convolution
Homotopy
Image enhancement
LHP
Retinal images
Journal
I
IF:
1.4
Papers:
291
Citations:
3.0K