| 足立 旭 | M, 2回目発表 | 計算システムズ生物学 | 金谷 重彦, | 安本 慶一, | 小野 直亮, | MD.ALTAF-UL-AMIN, | 黄 銘 |
|
title: Diversity of the soil microbiome in various land use types
abstract: In soils, various microorganisms compose a complex community called the microbiome. The microbiome in soils is known to play an essential role in determining soil fertility and plants' growth conditions. This study compares various microbiome of soils throughout Japan by a metagenomic approach. Also, this study tries to reveal the ecological relationship between soil microbiomes and the usage of grounds such as for crops, vegetables, and fruits. language of the presentation: Japanese 発表題目: 様々な日本の農耕地における土壌マイクロバイオームの多様性 発表概要: 土壌中には膨大な数の微生物が棲息し、土壌マイクロバイオームを形成している。土壌マイクロバイオームは、土壌の肥沃度や植物の健康に関わる重要な役割を担っていることが知られている。本研究では、様々な農耕地のタイプ（田、畑、果樹園）における土壌マイクロバイオームのメタゲノムデータを解析する。そして、農耕地ごとの土壌マイクロバイオームの違いや特徴を明らかにする。 | |||||||
| 小谷 行樹 | M, 2回目発表 | 計算システムズ生物学 | 金谷 重彦, | 安本 慶一, | 小野 直亮, | MD.ALTAF-UL-AMIN, | 黄 銘 |
|
title: Diagnostic assistance on Chest Radiographs using deep learning
abstract: In Japan, there is a concern about the shortage of radiologists. For this reason, computer aided diagnosis for radiologists using deep learning has been strongly expected. In previous studies, it was reported that deep learning is comparable to radiologists' diagnosis in identifying some diseases. However, it is still difficult to evaluate quantitativity and interpretability of the deep learning models. In this study, we developed a model to identify scoliosis and pleural thickening, which are frequently observed in health checkups, and evaluated consistency with the diagnosis by the professional radiologists. language of the presentation:Japanese | |||||||
| 久家 拓也 | M, 2回目発表 | 計算システムズ生物学 | 金谷 重彦, | 安本 慶一, | 小野 直亮, | MD.ALTAF-UL-AMIN, | 黄 銘 |
|
title: Automatic detection of lung tumor using 3D convolutional neural network
abstract: Lung cancer is the most frequently diagnosed cancer, and computed tomography (CT) is currently an effective method for the early detection of lung cancer. However, each image must be carefully read, which places a heavy burden on the radiologist. PET-CT examination is a technology that combines the functional (glucose metabolism) images of PET and the morphological images of CT, and is known to improve diagnostic accuracy and shorten examination time. Based on these findings, we will develop a diagnostic support system that automatically detects the location of lung cancer by combining CT and PET images. Previous studies have used the LUNA16 dataset, but in this study, we will use the Lung-PET-CT-Dx data to build a system that automatically detects the location of lung cancer. language of the presentation: Japanese | |||||||
| 櫻井 紀利 | M, 2回目発表 | 計算システムズ生物学 | 金谷 重彦, | 安本 慶一, | 小野 直亮, | MD.ALTAF-UL-AMIN, | 黄 銘 |
|
title: Feature extraction from pathological images of pancreas tumor using deep learning
abstract: Pathological images are generally used in many clinical scene to make a definitive diagnosis, e. g. for cancer treatment. In addition, it is important to analyze phenotypic changes in order to understand the characteristics of cancer. However, it is still difficult to quantitatively evaluate and measure phenotypes such as cell types. Therefore, understanding the morphological features of pathological images is useful in clinical scene and research. In recent years, research on image processing using deep learning has progressed rapidly, and new models for feature extraction from medical images have been proposed. In this study, we aim to understand the morphological features in pathological images by extracting features based on unsupervised learning using three different staining methods. language of the presentation: Japanese | |||||||