Comparison of various neural network-based models for retinal lesion analysis

Gil, Eddie M.; Keppler, Mark A.; Yakovlev, Vladislav V.; Boretsky, Adam; Bixler, Joel N.

doi:10.1117/12.2507908

Cited by 1 publication

(1 citation statement)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each starting point was trained for 200 additional epochs. In this stage, the models were trained using a Jaccard based loss function as seen in () 35

{Loss}=1-\,\frac{\sum {\rm{\min }}\left({y}_{{true}}^{i},{\,y}_{{pred}}^{i}\right)}{\sum {\rm{\max }}\left({y}_{{true}}^{i},{\,y}_{{pred}}^{i}\right)}

…”

Section: Methodsmentioning

confidence: 99%

Segmentation of laser induced retinal lesions using deep learning (December 2021)

Gil

Keppler

Boretsky³

et al. 2022

Lasers Surg Med

Self Cite

View full text Add to dashboard Cite

Objective Detection of retinal laser lesions is necessary in both the evaluation of the extent of damage from high power laser sources, and in validating treatments involving the placement of laser lesions. However, such lesions are difficult to detect using Color Fundus cameras alone. Deep learning‐based segmentation can remedy this, by highlighting potential lesions in the image. Methods A unique database of images collected at the Air Force Research Laboratory over the past 30 years was used to train deep learning models for classifying images with lesions and for subsequent segmentation. We investigate whether transferring weights from models that learned classification would improve performance of the segmentation models. We use Pearson's correlation coefficient between the initial and final training phases to reveal how the networks are transferring features. Results The segmentation models are able to effectively segment a broad range of lesions and imaging conditions. Conclusion Deep learning‐based segmentation of lesions can effectively highlight laser lesions, making this a useful tool for aiding clinicians.

show abstract

“…Each starting point was trained for 200 additional epochs. In this stage, the models were trained using a Jaccard based loss function as seen in () 35