Targeted transfer learning to improve performance in small medical physics datasets
This addresses the problem of limited data in medical physics for researchers and practitioners, though it is incremental as it reviews and compares existing methods.
The paper tackles the challenge of applying deep learning to small medical imaging datasets by evaluating state-of-the-art techniques, finding that transfer learning with images from the same body part can significantly improve performance with as few as 50 training images.
The growing use of Machine Learning has produced significant advances in many fields. For image-based tasks, however, the use of deep learning remains challenging in small datasets. In this article, we review, evaluate and compare the current state-of-the-art techniques in training neural networks to elucidate which techniques work best for small datasets. We further propose a path forward for the improvement of model accuracy in medical imaging applications. We observed best results from one cycle training, discriminative learning rates with gradual freezing and parameter modification after transfer learning. We also established that when datasets are small, transfer learning plays an important role beyond parameter initialization by reusing previously learned features. Surprisingly we observed that there is little advantage in using pre-trained networks in images from another part of the body compared to Imagenet. On the contrary, if images from the same part of the body are available then transfer learning can produce a significant improvement in performance with as little as 50 images in the training data.