Original title: Smart membrane: high content in situ monitoring barrier on chip with artificial neural network
Bo Tang, Victor Krajka, Mengxi Liu, Wei Zhao +6 · August 2026
Scientists test human tissues in tiny devices called organ-on-chip systems, but tracking cell health usually requires moving them to a microscope. Standard electrical tests only give basic numbers, missing how cells change shape and connect over time. To solve this, researchers designed a smart, ultrathin porous membrane with built-in sensors. Cells grow directly on this membrane inside the chip. The device sends tiny electrical signals through the cells to track changes in how electricity flows, a method called impedance sensing. By pairing these signals with artificial neural networks, the system automatically identified four growth phases: sticking, spreading, forming a full sheet, and building tight barriers. It recognized these phases with 95% confidence. The system also detected whether chemical drugs caused temporary or permanent damage to the cell barrier. This tool allows automated, non-invasive cell monitoring without taking chips out to use microscopes.
Imagine trying to tell if a fruit is ripe without touching it. You might tap it and listen, but you can't be sure without cutting it open. Here, the chip sends tiny electrical signals through the cell layer and listens to the response, and a neural network learns the patterns that correspond to different stages of ripeness, so you never have to cut the fruit open.
Researchers studying tissue barriers in organ-on-chip systems no longer need to repeatedly remove samples for microscopy. This chip continuously monitors barrier formation and breakdown in real time, automatically detecting stages with high confidence. It could speed up drug testing and reduce animal experiments by providing immediate, non-invasive data.
The work is a proof of concept demonstrated with HUVECs and a small training dataset. Generalization to other cell types or conditions is not yet shown, and the chip requires custom fabrication. Long-term stability and scalability are not addressed.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Smart membrane: high content in situ monitoring barrier on chip with artificial neural network — Bo Tang, Victor Krajka, Mengxi Liu et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.