In this study, a microfluidic culture device and related evaluation methods were developed using deep learning to construct a rapid assessment platform for peripheral neuropathy caused by typical anticancer drugs. Primary rodent dorsal root ganglia were cultured in a microfluidic culture device that separated the cell body and neurites, and morphological changes in the neuritis were analyzed using immunofluorescence imaging. Successful culture of separated neurites in the microfluidic device for more than 1 month indicated that this test process, including culture, drug stimulation, and fluorescence observation, results in a viable outcome. In addition, cultured samples were treated with several anticancer drugs known to cause peripheral neurotoxicity (i.e., vincristine, oxaliplatin, and paclitaxel), and morphological changes in the neuritis were analyzed using deep learning for image analysis. After training, artificial intelligence (AI) could identify morphological changes in the neurites caused by each compound and precisely predict toxicity, even at low concentrations. For the test compounds, AI could also precisely detect neurotoxicity based on neurite images, even at low concentrations. Our results suggest that this microfluidic culture system is useful for in vitro toxicity assessment.