Thus far, there have been no reported specific rules for systematically determining the appropriate augmented sample size to optimize model performance when conducting data augmentation. In this paper, we report on the feasibility of synthetic data augmentation using generative adversarial networks (GAN) by proposing an automation pipeline to find the optimal multiple of data augmentation to achieve the best deep learning-based diagnostic performance in a limited dataset. We used Waters’ view radiographs for patients diagnosed with chronic sinusitis to demonstrate the method developed herein. We demonstrate that our approach produces significantly better diagnostic performance parameters than models trained using conventional data augmentation. The deep learning method proposed in this study could be implemented to assist radiologists in improving their diagnosis. Researchers and industry workers could overcome the lack of training data by employing our proposed automation pipeline approach in GAN-based synthetic data augmentation. This is anticipated to provide new means to overcome the shortage of graphic data for algorithm training.

There is growing evidence that tumor cells depend on unique metabolism for their continued growth and survival, and that cancer cells are peculiarly addicted to the rapacious uptake of glucose and glutamine. AMP-activated protein kinase (AMPK) is one of the central regulators of cellular and organismal metabolism in eukaryotes, playing critical roles in regulating cell growth and reprogramming metabolism. Several of the downstream effects of AMPK on metabolic adaptation can be attributed to the AMPK-dependent activation of the p53 tumor suppressor, resulting in its transcriptional activation and initiating a metabolic cell cycle checkpoint. However, in the absence of p53, treatment with the blood glucose-lowering drug metformin delays tumor progression, suggesting that the AMPK-p53 signaling axis responds in diverse manner depending on genetic alteration. Due to the gain-of-function, mutant p53 protein contributes to tumor metastasis and shortening the latency of tumor progression. Thereby activated K-ras/mutant p53 mouse is preferentially used as lung adenocarcinoma preclinical model. Here, we have established mouse tumor cells derived from activated K-ras/mutant p53 mouse to generate syngeneic lung cancer mouse model, allowing in vivo imaging. We asked whether the gain-of-function metastatic phenotype observed in the mouse is responsible for unique metabolic conditions resulting in the formation of lung tumors. We have evaluated tumor metabolic environment of this model and examined AMPK signaling pathways. Citation Format: Se-Young Jo, Hye-Min Moon, ChuHee Lee, Se Jin Jang, Young-Ah Suh. AMP-activated protein kinase (AMPK) signaling in the context of gain-of-function mutant p53 in vivo. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2295. doi:10.1158/1538-7445.AM2015-2295