Classification of red cell dynamics with convolutional and recurrent neural networks: a sickle cell disease case study

HIGHLIGHTS

SUMMARY

For the sequences having a number of images smaller than K, the processing differs depending on the methodology applied (CNN or recurrent CNN, see below): for the CNN method, the input size is fixed so the short sequences will be padded to get exactly K images (see section "Approach A: Fixed-size convolutional neural_networks). The authors thus have now three classes: the unreliable sequences, the sequences of tank-treading RBC, hereafter called tank-treading sequences, and the sequence of flipping RBC, hereafter called flipping sequences. Convolutional Neural_Networks (CNN) have shown good results in blood . . .

 



If you want to have access to all the content you need to log in!

Thanks :)

Username or E-mail:

If you don't have an account, you can create one here.