https://www.selleckchem.com/pr....oducts/irak4-in-4.ht
In this work, we propose to use an artificial neural network to classify limited data of clinical multispectral and autofluorescence images of skin lesions. Although the amount of data is limited, the deep convolutional neural network classification of skin lesions using a multi-modal image set is studied and proposed for the first time. The unique dataset consists of spectral reflectance images acquired under 526 nm, 663 nm, 964 nm, and autofluorescence images under 405 nm LED excitation. The augmentation algorithm was applied for m
Like
Comment
Share
