Upgrade to Pro — share decks privately, control downloads, hide ads and more …

論文紹介:Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs / CNN for Diabetic Retinopathy

論文紹介:Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs / CNN for Diabetic Retinopathy

Shotaro Ishihara

February 10, 2018
Tweet

More Decks by Shotaro Ishihara

Other Decks in Technology

Transcript

  1.  Development and Validation of a Deep Learning Algorithm for

    Detection of Diabetic Retinopathy in Retinal Fundus Photographs Varun Gulshan (Google Inc) et al. Journal of the American Medical Association, December 13, 2016 https://jamanetwork.com/journals/jama/fullarticle/2588763 Feb. 9th, 2018
  2.  •  ' "#&%  deep learning ! •

        $ 2 https://research.googleblog.com/2016/11/deep-learning-for-detection-of-diabetic.html (Available: Feb. 9th, 2018)
  3.  • 128175/ ," • 54#*3-7# - • )$. [0,

    1] • Deep convolutional neural networks • $ 22 million • ' 80%20% +% (! ' & 0  3
  4. • B$JA 1. +8(.9 2. 4sensitivity specificity 3. C=;79 4.

    sensitivity and specificityI?) >: A • 53@H • GF0K  • <-?$D* • $2, • '6C=/ %"&E '61  1  • neural network!%#9 • "Hence, this algorithm is not a replacement for a comprehensive eye examination" 5 
  5.   • Tensorflow Tutorial CNN $# • Kaggle 

     • 100"!100" • 80 %  6 Step Training Accuracy   sensitivity = 12/20 = 0.6 specificity = 13/20 = 0.65 https://github.com/upura/deep.learning.diabetic.retinopathy