Transfer representation learning for medical image analysis

By Chuen-Kai Shie, C.-H. Chuang, Chun-Nan Chou, Meng-Hsi Wu, Edward Y. Chang. 2015

136 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Otorhinolaryngology, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

Using Rule-Based Labels for Weak Supervised Learning: A ChemNet for Transferable Chemical Property Prediction (2017, background): “quot;image space" on which the network was trained on, seemingly unrelated outcomes can be achieved. For example, a model pre-trained on ImageNet can also be ﬁne-tuned to classify medical images [17]. While medical applications are seemingly unrelated to conventional image recognition tasks, bo”
An Information-Theoretic Approach to Transferability in Task Transfer Learning (2019, methodology): “For example, deep convolutional neural networks trained for the ImageNet classiﬁcation task have been used as the source network in transfer learning for target tasks with fewer labeled data [2], such as medical image analysis [3] and structural damage recognition in buildings [4].”
Artificial Intelligence in XPRIZE DeepQ Tricorder (2017, methodology): “In the second scheme, we used the two-layer classifier fusion structure proposed in [22].”
ChemNet: A Transferable and Generalizable Deep Neural Network for Small-Molecule Property Prediction (2017, background): “For example, a model pre-trained on ImageNet can also be fine-tuned to classify medical images [17].”
An Ensemble of Fine-Tuned Convolutional Neural Networks for Medical Image Classification (2017, methodology): “It has been successfully utilised in various modality or disease specific studies [29]–[31].”
Deep Learning Techniques for Medical Image Segmentation: Achievements and Challenges (2019, background): “Transfer learning is defined as the capability of a system to recognize and employ the knowledge learned in a previous source domain to a novel task [68].”
Automatic Detection and Classification of Colorectal Polyps by Transferring Low-Level CNN Features From Nonmedical Domain (2017, background): “Nevertheless, if the number of images of the target task is small, direct transfer of CNN features from the source task without fine-tuning is also useful [21], [23]–[25].”