Deep learning can improve non-invasive alopecia diagnosis using hair images.
Transfer learning with three neural network architectures accurately classifies hair diseases.
The method creates realistic, anonymous acne face images for research, achieving 97.6% accuracy in classification.
November 2022 in “bioRxiv (Cold Spring Harbor Laboratory)” Using deep learning to predict gene expression from images could help assess colorectal cancer metastasis.
October 2023 in “Sinkron” The system can accurately classify hair diseases with 94.5% accuracy using a CNN.