Automated fluorescence instruments do the same thing thousands of times a day. A plate imager or high-content screening system moves to a field, sets focus, switches the excitation and emission bands, takes an exposure, and moves on. A slide scanner does the same across a whole tissue section. High-content screening combines the throughput of automated methods with quantitative cell imaging, and it is now part of drug discovery from primary screens through toxicity profiling (Zanella, Lorens and Link, Trends in Biotechnology, 2010).
In these instruments the camera is one step in a repeating loop, and every millisecond it adds is paid back across every field, channel and plate. This post covers how to choose a cooled camera for that loop, using Finger Lakes Instrumentation (FLI) Revelation and Aurora cameras as examples. For the wider OEM relationship, see our post on choosing an OEM camera partner for life-science instruments.
What makes automated imaging different
A research microscope camera is judged on single images. An automated instrument's camera is judged on the whole run: how many fields per hour, how repeatable the measurement is from well to well and from instrument to instrument, and how cleanly it fits the control software.
That shifts the questions. Sensor format sets how many fields you need. Readout mode decides how you pulse the light. Trigger lines decide how tightly the camera, light source, stage and filters stay in step.
Field per tile sets throughput
The field at the sample is the sensor size divided by the total magnification. Take 20x as an example. The 11.3 × 11.3 mm sensor in the RV533 covers about 0.57 × 0.57 mm, or about 0.32 mm² per tile. The 23.5 × 15.7 mm sensor in the RV571 covers about 1.18 × 0.79 mm, or about 0.92 mm², close to three times the area. To cover the same region, the larger sensor needs roughly a third as many stage moves, autofocus steps and filter cycles.
A larger sensor only helps if the optics can fill it. Check the usable image circle at your camera port. The RV533's 16 mm diagonal fits comfortably inside most microscope fields, while the RV571's 28.2 mm diagonal, or the 43.2 mm diagonal of the AR455, needs optics designed for a large field to avoid dark corners and soft edges.
Sampling at the sample
Pixel size at the sample is the pixel size divided by the magnification. All three of these sensors use 3.76 µm pixels, so at 20x each pixel covers about 0.19 µm of the sample. That is fine enough for Nyquist sampling with typical 20x objectives at visible wavelengths. Our post on pixel size and sampling walks through the math.
If an assay needs counts rather than fine detail, you can bin to trade resolution for signal and smaller files. On these Sony sensors, binning is done digitally rather than on the sensor, so plan your noise budget with that in mind.
Rolling shutter, global reset and strobed light
These cameras use a rolling shutter, with a rolling-with-global-reset mode. In plain rolling mode, each row starts and ends its exposure a little later than the one before. With constant illumination that is fine, but if the light switches on or off partway through a frame, some rows see more light than others.
Automated instruments usually pulse LED or laser excitation to limit photobleaching. Global-reset mode pairs well with that: all rows start integrating together, and if the light pulse fits inside the window before readout begins, every row gets the same dose. In practice, trigger the camera first, then fire the light, and keep the sample dark while the frame reads out. Our post on synchronizing cooled cameras covers trigger delay and timing in more depth.
Keeping the camera, light, stage and filters in step
External trigger in and out are standard on Aurora and Revelation cameras. A hardware trigger lets the stage controller or a timing board start each exposure, and the camera's output tells the rest of the instrument when the exposure is under way, without waiting on software round-trips.
Then build a time budget for one field: stage move and settle, autofocus, then for each channel a filter change, an exposure and the readout. Filter changes are often the surprise. A general-purpose stepper wheel such as FLI's CenterLine takes about 1 s per move and needs a host computer to drive it. The FLI High Speed Filter Wheel moves to the adjacent filter in 23 ms on the 6-position model and 30 ms on the 10-position model, and it can step through a stored sequence on a trigger input. With four channels, that difference alone is a few seconds per field. Our post on fast filter switching works through the arithmetic.
Sensitivity and cooling for weak signals
Many assays are dim: low-expression markers, thin samples, short exposures to protect live cells. Back-illuminated sensors help by collecting more of the light that arrives, and cooling keeps dark signal low enough that it does not set the noise floor. Aurora and Revelation cameras cool to about 50 °C below ambient on air, with optional liquid cooling for sealed enclosures. See our posts on air vs liquid cooling and long exposures with cooled CMOS for the detail.
Aurora cameras are not offered with a mechanical shutter. For dark frames, use a blocked filter-wheel position or switch the excitation off; both suit an automated calibration routine.
Data rates and the host
A 16-bit full frame from the RV533 is about 18 MB, and from the RV571 about 52 MB. Four channels across 100 fields is 400 frames: about 7 GB from the RV533 or about 21 GB from the RV571, per plate. Multiply by plates per day and the storage, network and analysis pipeline matter as much as the camera.
The camera's on-board memory is a working frame buffer, not long-term storage, so plan for continuous transfer to the host. One PC can run multiple cameras, which helps when an instrument images several channels at once.
Fitting the camera into the instrument
Revelation puts the same Sony sensors as Aurora into a thin, fin-less body that is shorter front to back, suits tight instrument geometries, and can be liquid-cooled or mounted to your housing as a heatsink. Every Revelation camera includes our software development kit and support for integration into your OEM design, so exposure, cooling, readout mode and triggering can be controlled from your own application. Drivers and manuals are on our support page.
Unit-to-unit consistency matters when you ship many instruments. Data-sheet values are typical, and a production test summary exists for each camera by serial number. Our post on acceptance testing shows what to check on arrival.
Questions to settle before you choose
- What magnification and image circle does your optical path deliver at the camera?
- How many fields, channels and plates or slides per run?
- Will excitation be continuous or pulsed, and what controls the timing?
- What filter-change time fits your per-field budget?
- How dim is your weakest assay, and how long can exposures be?
- How much data can your host store and process per day?
- How much space, airflow or coolant is available inside the enclosure?
Talk to our sales team
FLI cameras are Made in USA, and our sensor chambers are assembled in our 1,000+ sq ft ISO Class 7 clean room. See how on our in-house manufacturing page.
Send our sales team your optical layout, assay list and throughput target, and we'll help you choose a camera and filter wheel for your instrument. Request a quote or email sales@flicamera.com.
