Development of a myelin dysfunction assessment method based on single axon conduction velocimetry and machine learning of axon morphology.

2023 Fiscal Year Final Research Report

• Project/Area Number: 22K15344
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 47060:Clinical pharmacy-related
• Research Institution: Tohoku Institute of Technology
• Principal Investigator: Shibata Mikako 東北工業大学, 工学部, 研究員 (40887056)
• Project Period (FY): 2022-04-01 – 2024-03-31
• Keywords: 末梢神経障害 / MEA / 感覚ニューロン / 薬剤性ミエリン障害 / 抗がん剤

Outline of Final Research Achievements

Immunostained images and changes in impedance values were measured after exposure of sensory neurons cultured on MEA to negative compounds and anticancer agents. The results showed that by focusing on changes in impedance values, it was possible to detect peripheral neuropathy at concentrations lower than those at which morphological changes occur. Next, an AI that could detect differences at lower doses was created using machine learning of stained images 24 hours after addition. The created AI judged negative compounds as negative and vincristine as positive in the unlearned data. It also detected toxicity of Suramin, a myelin-disrupting positive compound, from 10 μM. The AI was able to predict drug-induced myelin damage at previously undetectable doses.

Free Research Field

農学

Academic Significance and Societal Importance of the Research Achievements

本研究で開発したMEAを用いたインピーダンス計測による末梢神経障害の検出法および、染色画像の機械学習を用いた薬剤性ミエリン障害の予測法は、いずれも先行研究で報告されている毒性濃度や形態変化が認められる濃度と比較して、低濃度で早期に障害を予測できる評価系であり、創薬開発の探索段階におけるスクリーニングにおいて有効なin vitro試験として期待できる。