Track topics on Twitter Track topics that are important to you
Deep convolutional neural networks offer state-of-the-art performance for medical image analysis. However, their architectures are manually designed for particular problems. On the one hand, a manual designing process requires many trials to tune a large number of hyperparameters and is thus quite a time-consuming task. On the other hand, the fittest hyperparameters that can adapt to source data properties (e.g., sparsity, noisy features) are not able to be quickly identified for target data properties. For instance, the realistic noise in medical images is usually mixed and complicated, and sometimes unknown, leading to challenges in applying existing methods directly and creating effective denoising neural networks easily. In this paper, we present a Genetic Algorithm (GA)-based network evolution approach to search for the fittest genes to optimize network structures automatically. We expedite the evolutionary process through an experience-based greedy exploration strategy and transfer learning. Our evolutionary algorithm procedure has flexibility, which allows taking advantage of current state-of-the-art modules (e.g., residual blocks) to search for promising neural networks. We evaluate our framework on a classic medical image analysis task: denoising. The experimental results on computed tomography perfusion (CTP) image denoising demonstrate the capability of the method to select the fittest genes for building high-performance networks, named EvoNets. Our results outperform state-of-the-art methods consistently at various noise levels.
This article was published in the following journal.
Name: Medical image analysis
Evolutionary algorithms have been demonstrated to be very competitive in the community detection for complex networks. They, however, show poor scalability to large-scale networks due to the exponenti...
Deep learning algorithms produces state-of-the-art results for different machine learning and computer vision tasks. To perform well on a given task, these algorithms require large dataset for trainin...
We present an approach for fully automatic urinary bladder segmentation in CT images with artificial neural networks in this study. Automatic medical image analysis has become an invaluable tool in th...
This paper presents an interplay between deep learning and game theory. It models basic deep learning tasks as strategic games. Then, distributionally robust games and their relationship with deep gen...
Continuous action policy search is currently the focus of intensive research, driven both by the recent success of deep reinforcement learning algorithms and the emergence of competitors based on evol...
The brain networks controlling movement are complex, involving multiple areas of the brain. Some neurological diseases, like Parkinson's disease, cause abnormalities in the brain networks....
The goal of the study is to address the unmet need of TRD patients by identifying brain networks critical for treating depression and to use next generation precision DBS with steering cap...
This research programme aims to translate expertise in warfarin anticoagulation from resource-rich to low and middle income countries (LMIC)settings with the goal of improving patient heal...
This study evaluates a second review of ultrasound images of breast lesions using an interactive "deep learning" (or artificial intelligence) program developed by Samsung Medical Imaging, ...
To map and identify the major genetic determinants of hypertension and to study possible interactions between genetic and non-genetic factors in defined populations. HyperGEN initially co...
In evolutionary theory, overlapping geographic distribution of diverging species. In sympatric GENETIC SPECIATION, genetic diversion occurs without geographic separation.
The use of COMPUTER COMMUNICATION NETWORKS to store and transmit medical PRESCRIPTIONS.
Body of knowledge related to the use of organisms, cells or cell-derived constituents for the purpose of developing products which are technically, scientifically and clinically useful. Alteration of biologic function at the molecular level (i.e., GENETIC ENGINEERING) is a central focus; laboratory methods used include TRANSFECTION and CLONING technologies, sequence and structure analysis algorithms, computer databases, and gene and protein structure function analysis and prediction.
Production of new arrangements of DNA by various mechanisms such as assortment and segregation, CROSSING OVER; GENE CONVERSION; GENETIC TRANSFORMATION; GENETIC CONJUGATION; GENETIC TRANSDUCTION; or mixed infection of viruses.
The discipline studying genetic composition of populations and effects of factors such as GENETIC SELECTION, population size, MUTATION, migration, and GENETIC DRIFT on the frequencies of various GENOTYPES and PHENOTYPES using a variety of GENETIC TECHNIQUES.
Standard antiretroviral therapy (ART) consists of the combination of at least three antiretroviral (ARV) drugs to maximally suppress the HIV virus and stop the progression of HIV disease. Huge reductions have been seen in rates of death and suffering whe...