ChexNet is a 121 layer convolutional neural network developed at Stanford University which reads chest X-ray images to diagnose 14 different chest diseases including Pneumonia.In this study 112,120 chest X-ray images of 30,805 patients were analyzed to develop the algorithm.CheXNet achieved an F1 score of 0.435 (95% CI 0.387, 0.481), higher than the radiologist average of 0.387 (95% CI 0.330, 0.442).
Read this paper using the link below:
https://arxiv.org/pdf/1711.05225.pdf
What is an F-1 score?
F1 score is harmonic mean of precision and recall used frequently in machine learning to measure accuracy of a test.
F1 = 2 x (Precision x Recall)/(Precision + Recall)
Best F1 score is 1 and worst is 0.