講演抄録/キーワード |
講演名 |
2016-01-20 13:05
培養細胞の自動品質判定アルゴリズムの開発 ○佐藤和紀(山形大)・佐々木寛人・加藤竜司(名大)・湯浅哲也・姜 時友(山形大) MI2015-130 |
抄録 |
(和) |
再生医療現場においては,培養細胞の品質判定・品質管理がその安全性確保のためにも重要である.しかし現在は作業者の主観に依るところが大きく,精度と再現性,人的・時間的コストなどが問題になっている.本研究では,ヒト皮膚繊維芽細胞の培養過程における位相差顕微鏡画像から,それらに特異的な画像処理を施すことによって,非侵襲的かつ自動的な細胞品質判定を可能とするアルゴリズムを開発・構築するとともにその有効性を定量的に検証する. |
(英) |
With recent progress of cell and tissue therapy in regenerative medicine, novel classification method of cell quality has been expected toward safety clinical application. We have developed the automatic image-based algorithm for the normal human dermal fibroblasts (NDHF) cells. The quality assessment of confluent NHDFs has inevitable difficulty due to their high density and strong heterogeneity of orientation distribution. It is also difficult to capture the morphological outlines of individual confluent cells. In order to quantify the population property of cell orientations, we performed the feature extraction based on the scale-invariance feature transform (SIFT) after some pre-processes. Through the learning by kernel support vector machine (kSVM) for the feature values, we tried to discriminate the remaining lifespan of cells and manipulation error during culturing. As a result, we demonstrated that our classification algorithm showed high accuracy and robustness for the assessments of cell quality without any invasive treatment. These results suggest that present study could provide reproducible and automatic classification algorithm of cell quality for future clinical application in tissue engineering fields. |
キーワード |
(和) |
再生医療 / 位相差顕微鏡 / ヒト皮膚線維芽細胞 / 細胞画像 / Scale-Invariant Feature Transform (SIFT) / サポートベクターマシン (SVM) / パターン認識 / |
(英) |
Regenerative medicine / Phase-contrast microscope / Normal human dermal fibroblasts (NHDF) / Cellular image / Scale-Invariant Feature Transform (SIFT) / Support Vector Machine (SVM) / pattern recognition / |
文献情報 |
信学技報, vol. 115, no. 401, MI2015-130, pp. 277-282, 2016年1月. |
資料番号 |
MI2015-130 |
発行日 |
2016-01-12 (MI) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
MI2015-130 |