Apr 5, 2018

Evolutionary pruning of transfer learned deep convolutional neural network for breast cancer diagnosis in digital breast tomosynthesis

Physics in Medicine and Biology
Ravi K SamalaKenny Cha

Abstract

Deep learning models are highly parameterized, resulting in difficulty in inference and transfer learning for image recognition tasks. In this work, we propose a layered pathway evolution method to compress a deep convolutional neural network (DCNN) for classification of masses in digital breast tomosynthesis (DBT). The objective is to prune the number of tunable parameters while preserving the classification accuracy. In the first stage transfer learning, 19 632 augmented regions-of-interest (ROIs) from 2454 mass lesions on mammograms were used to train a pre-trained DCNN on ImageNet. In the second stage transfer learning, the DCNN was used as a feature extractor followed by feature selection and random forest classification. The pathway evolution was performed using genetic algorithm in an iterative approach with tournament selection driven by count-preserving crossover and mutation. The second stage was trained with 9120 DBT ROIs from 228 mass lesions using leave-one-case-out cross-validation. The DCNN was reduced by 87% in the number of neurons, 34% in the number of parameters, and 95% in the number of multiply-and-add operations required in the convolutional layers. The test AUC on 89 mass lesions from 94 independent DBT c...Continue Reading

Mentioned in this Paper

Biochemical Pathway
Biological Neural Networks
Genetic Programming
Anatomical Layer
Classification
Digital Breast Tomosynthesis
Neurons
Cross Validation
Dialectical Behavior Therapy
Breast Diseases

Related Feeds

Cell Adhesion Molecules in the Brain

Cell adhesion molecules found on cell surface help cells bind with other cells or the extracellular matrix to maintain structure and function. Here is the latest research on their role in the brain.