Back

An in vitro microfluidic culture device for peripheral neurotoxicity prediction at low concentration based on deep learning

Han, X.; Matsuda, N.; Matsuda, K.; Yamanaka, M.; Suzuki, I.

2022-11-10 pharmacology and toxicology
10.1101/2022.11.09.515778 bioRxiv
Show abstract

In this study, a microfluidic culture device was developed including related evaluation methods using deep learning for the purpose of constructing a rapid assessment platform for peripheral neuropathy caused by compounds. Primary rodent dorsal root ganglion was cultured in the microfluidic culture device that could separate the cell body and neurites, and the neurites morphological changes were analyzed by immunofluorescence images. Separated neurites successful culture in the microfluidic device for more than 1 month indicated that a series of test processes from culture to drug stimulation and fluorescence observation is possible. Additionally, cultured samples were treated with several anticancer drugs known to cause peripheral neurotoxicity (i.e., vincristine, oxaliplatin, and paclitaxel) and analyzed the neurites morphological changes by deep learning for image analysis. After training, artificial intelligence (AI) could identify neurite morphological changes caused by each compound and precisely predicted the toxicity positivity, even at low concentrations. For testing compounds, AI could also precisely detect toxicity negative and positive based on neurite images, even at low concentrations. Therefore, this microfluidic culture system is supposed to be useful for in vitro toxicity assessment.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.