Artificial Neural Network enabling Clinically Meaningful Biological Image Data Generation
| Autor: | Soonkyum Kim, YaeJun Baik, SeungBeum Suh, Junhyoung Ha, Woosub Lee, Dohee Lee |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
0303 health sciences
Artificial neural network Computer science business.industry Test data generation ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Cancer Pattern recognition 010501 environmental sciences Real image medicine.disease 01 natural sciences Image (mathematics) 03 medical and health sciences medicine Neural Networks Computer Artificial intelligence business 030304 developmental biology 0105 earth and related environmental sciences |
| Zdroj: | EMBC |
| DOI: | 10.1109/embc44109.2020.9176621 |
| Popis: | Biological experiments for developing efficient cancer therapeutics require significant resources of time and costs particularly in acquiring biological image data. Thanks to recent advances in AI technologies, there have been active researches in generating realistic images by adapting artificial neural networks. Along the same lines, this paper proposes a learning-based method to generate images inheriting biological characteristics. Through a statistical comparison of tumor penetration metrics between generated images and real images, we have shown that forged micrograph images contain vital characteristics to analyze tumor penetration performance of infecting agents, which opens up the promising possibilities for utilizing proposed methods for generating clinically meaningful image data. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|