Flow Sensor for Cell Culture Perfusion

How a perfusion flow rate gets checked today

You set a pressure on the controller driving your perfusion, then trust that the cells get the flow you planned. In most setups nothing between the controller and the chip measures what gets through. Over a culture that runs for days, the flow you planned and the flow the cells get drift apart. Serum protein and cell debris accumulate in the narrowest parts of the chip and make it harder for the medium to pass. Warming to 37 °C works the other way. The medium thins out, and the same pressure moves more of it. And as the reservoir empties during the run, it pushes a little less. The controller goes on showing the pressure you entered, with no sign that the shear stress on your monolayer has shifted.

Without a flow measurement, you also have no record of the shear stress or the medium renewal your cells had. Neither you nor another lab can then repeat the run as it was. Several methods can check a perfusion flow, and they differ in what they measure and when.

No measurement at all. You take the setpoint on your pump as the record of what the culture received. It costs nothing, and it is right for as long as nothing in the line changes. Once the chip starts to clog or the medium warms, the flow reaching your cells no longer matches that setpoint. Nothing in the run warns you, so you usually find out from the cells themselves. A monolayer detaches or a barrier starts to leak, or an organoid matures more slowly than the last batch did.

A balance under the outlet. If you want the true flow rate, you can collect what comes out of the chip over a set time and weigh it. A balance traces back to a mass standard, which makes this the method sensors are calibrated against. But you only get the answer at the end of the collection, from the far end of the line, so nothing tells you what happens while the cells are under flow. With an open vessel, part of what you collect can also evaporate before it is weighed.

Thermal sensing. A thermal sensor heats a small element in the line and reads the flow continuously from how fast the passing medium carries that heat away. But the element is in direct contact with the medium, so over a long culture it collects a layer of serum protein and cells. This layer changes how the heat leaves the element, and the reading can drift without the display showing it. You then have to plan regular cleaning, even in a run you meant to automate.

Optical sensing, outside the fluid. An optical sensor reads the flow through the tubing wall, so it never touches the medium and raises no sterility concern. Groups building microphysiological platforms have picked it for long cultures, where a sensor in contact with the medium would foul. In return, it needs a clear view through a set length of tubing, and its calibration depends on that tubing as much as on the medium.

Differential pressure, with the flow path in a cartridge. Here the sensor reads the flow from the pressure difference across a known resistance inside a cartridge, without putting any heat into the medium. A pressure-driven line needs such a resistance, usually a length of narrow tubing, to work in a low flow range. The one in the cartridge already plays that role, which leaves you no tubing to size or add. The sensor measures pressure and flow together, so it raises an alert when the line begins to block and flags a reading that drifts. Like a thermal element, this resistance is in contact with the medium and will foul in the end. But it is built into a cartridge that you snap out and replace when the sensor signals that it is time.

The shear your cells were under, recorded hour by hour

The shear stress on your monolayer depends on the flow rate, but few perfusion setups keep track of that rate during the culture. With the sensor in the line, it is measured throughout the run, so you know the shear your cells were under at any point. Those values can go into your paper, and another lab can reproduce the same conditions on its own cells and chip.

A blockage caught before the culture pays for it

A partial blockage forms slowly in the chip or the tubing and can cost a perfused culture days of work, while the pump keeps running as if nothing had changed. Fresh medium reaches the cells more slowly, so they get less nutrient and oxygen, waste accumulates and the shear stress drops. A monolayer can detach or lose its barrier, and an organoid can change phenotype, while the pump display still looks normal. Measuring flow and pressure together, the sensor detects the blockage as soon as it starts and raises an alert.

A culture comparable from its first day to its last

Any sensor that stays in a biological medium fouls over time, and its reading can drift without any sign on the display. Your data then show a steady flow while the real flow was changing, and measurements from the start and the end of the culture are no longer comparable. Here the part that fouls is the cartridge. When the sensor flags a drifting reading, you snap the cartridge out and replace it without tools, while the culture keeps running.

Nothing to size before a fragile model goes under flow

Working in a low flow range with a pressure source usually means adding narrow tubing to the line and recalculating its length and diameter for every new flow rate. The resistance in the cartridge already plays that role, so you connect the reservoir, the sensor and your chip, and start the culture. The line is simpler to set up as a result, with fewer surfaces touching the medium and fewer places for bubbles to get trapped.

Will the shear you designed hold without a loop to tune?

Holding a flow rate steady usually means setting up a feedback loop and finding the right parameters on a live culture. A dedicated algorithm written for this sensor runs that loop with our OB1 pressure controller and needs no tuning. You enter a flow rate, and the controller adjusts the pressure to hold it, even as the chip clogs or the medium warms. The shear stress on your cells stays at the value you planned.