Generalization error analysis for deep convolutional neural network with transfer learning in breast cancer diagnosis.

08:00 EDT 24th March 2020 | BioPortfolio

Summary of "Generalization error analysis for deep convolutional neural network with transfer learning in breast cancer diagnosis."

Deep convolutional neural network (DCNN), now popularly called artificial intelligence (AI), has shown the potential to improve over previous computer-assisted tools in medical imaging developed in the past decades. A DCNN has millions of free parameters that need to be trained, but the training sample set is limited in size for most medical imaging tasks so that transfer learning is typically used. Automatic data mining may be an efficient way to enlarge the collected data set but the data can be noisy such as incorrect labels or even a wrong type of images. In this work we studied the generalization error of DCNN with transfer learning in medical imaging for the task of classifying malignant and benign masses on mammograms. With a finite available data set, we simulated a training set containing corrupted data or noisy labels. The balance between learning and memorization of the DCNN was manipulated by varying the proportion of corrupted data in the training set. The generalization error of DCNN was analyzed by the area under the receiver operating characteristic curve for the training and test sets and the weight changes after transfer learning. The study demonstrates that the transfer learning strategy of DCNN for such tasks needs to be designed properly, taking into consideration the constraints of the available training set having limited size and quality for the classification task at hand, to minimize memorization and improve generalizability.


Journal Details

This article was published in the following journal.

Name: Physics in medicine and biology
ISSN: 1361-6560


DeepDyve research library

PubMed Articles [22723 Associated PubMed Articles listed on BioPortfolio]

An analysis of training and generalization errors in shallow and deep networks.

This paper is motivated by an open problem around deep networks, namely, the apparent absence of over-fitting despite large over-parametrization which allows perfect fitting of the training data. In t...

Theory of deep convolutional neural networks: Downsampling.

Establishing a solid theoretical foundation for structured deep neural networks is greatly desired due to the successful applications of deep learning in various practical domains. This paper aims at ...

Deep convolutional neural network-based detection of meniscus tears: comparison with radiologists and surgery as standard of reference.

To clinically validate a fully automated deep convolutional neural network (DCNN) for detection of surgically proven meniscus tears.

Inverse design of plasmonic metasurfaces by convolutional neural network.

Artificial neural networks have shown effectiveness in the inverse design of nanophotonic structures; however, the numerical accuracy and algorithm efficiency are not analyzed adequately in previous r...

Orthogonal Deep Neural Networks.

In this paper, we introduce the algorithms of Orthogonal Deep Neural Networks (OrthDNNs) to connect with recent interest of spectrally regularized deep learning methods. OrthDNNs are theoretically mot...

Clinical Trials [5806 Associated Clinical Trials listed on BioPortfolio]

Diagnostic Efficacy of CNN in Differentiation of Visual Field

Glaucoma is currently the leading cause of irreversible blindness in the world. The multi-center study is designed to evaluate the efficacy of the convolutional neural network based algori...

A New Method of Muscle Strength Testing Using a Quantitative Ultrasonic Technique and a Convolutional Neural Network

In addition to muscle thickness and average echo intensity, this study aimed to use quantitative ultrasonic technology to increase the number of related parameters of power Doppler ultraso...

Short-term Postural Training for Older Adults

Generalization refers to skill transfer under various working spaces following motor practice. The extent of generalization effect links causal to in-depth recognition of error properties ...

Renal Cancer Detection Using Convolutional Neural Networks

We aim to experiment and implement various deep learning architectures in order to achieve human-level accuracy in Computer-aided diagnosis (CAD) systems. In particular, we are interested ...

Identification of Interscalene Brachial Plexus on Ultrasonography Using a Deep Neural Network

The purpose of the study is to develop and validate an algorithm based on deep neural networks (DNNs) to identify interscalene brachial plexus on ultrasonography automatically.

Medical and Biotech [MESH] Definitions

Meta-analysis of randomized trials in which estimates of comparative treatment effects are visualized and interpreted from a network of interventions that may or may not have been evaluated directly against each other. Common considerations in network meta-analysis include conceptual and statistical heterogeneity and incoherence.

An early embryonic developmental process of CHORDATES that is characterized by morphogenic movements of ECTODERM resulting in the formation of the NEURAL PLATE; the NEURAL CREST; and the NEURAL TUBE. Improper closure of the NEURAL GROOVE results in congenital NEURAL TUBE DEFECTS.

A computer architecture, implementable in either hardware or software, modeled after biological neural networks. Like the biological system in which the processing capability is a result of the interconnection strengths between arrays of nonlinear processing nodes, computerized neural networks, often called perceptrons or multilayer connectionist models, consist of neuron-like units. A homogeneous group of units makes up a layer. These networks are good at pattern recognition. They are adaptive, performing tasks by example, and thus are better for decision-making than are linear learning machines or cluster analysis. They do not require explicit programming.

The two longitudinal ridges along the PRIMITIVE STREAK appearing near the end of GASTRULATION during development of nervous system (NEURULATION). The ridges are formed by folding of NEURAL PLATE. Between the ridges is a neural groove which deepens as the fold become elevated. When the folds meet at midline, the groove becomes a closed tube, the NEURAL TUBE.

The introduction of error due to systematic differences in the characteristics between those selected and those not selected for a given study. In sampling bias, error is the result of failure to ensure that all members of the reference population have a known chance of selection in the sample.

Quick Search

DeepDyve research library

Relevant Topic

  Bladder Cancer Brain Cancer Breast Cancer Cancer Cervical Cancer Colorectal Head & Neck Cancers Hodgkin Lymphoma Leukemia Lung Cancer Melanoma Myeloma Ovarian Cancer Pancreatic Cancer ...

Searches Linking to this Article