The History of Flow Cytometry: From Inkjet Printers to Cancer Diagnostics
Published 2026-06-14
Imagine standing on the shoulder of a 12-lane superhighway. Every car is screaming past at 100 mph, and your job is to identify the make, model, year, color, and number of passengers in every single one — and then teleport just the red sports cars into a separate parking lot.
That's flow cytometry. Tens of thousands of cells per second, each one measured, classified, and (if you want) sorted into its own tube. The wild part? The machine that does this was inspired by an inkjet printer, partially funded by nuclear fallout research, and built around a physics trick a guy noticed in 1742.
Want the visual, interactive version of this story? Explore the Flow Cytometry History Tool →
1. The Pre-History: Lenses, Light, and Droplets (1700s–1934)
In the early 1700s, Antonie van Leeuwenhoek ground tiny single-lens microscopes capable of 300x magnification and became the first human to stare a bacterium in the face. Cool — but painfully slow. You can't count a million cells one eyepiece at a time.
Two physics observations would eventually rescue him:
- 1742 — Mikhail Lomonosov noticed that particles suspended in fluid scatter light in predictable patterns. Today we call this Forward Scatter (FSC, ~ cell size) and Side Scatter (SSC, ~ internal complexity) — the two parameters every cytometer measures before it measures anything else.
- 1833 — Felix Savart showed that a continuous fluid stream, when vibrated at the right frequency, breaks into perfectly uniform droplets. Hold onto that — it becomes the heart of cell sorting 130 years later.
Then in 1934, bacteriologist Andrew Moldovan published the first real attempt to automate cell counting: push a cell suspension through a tiny capillary past a photoelectric sensor. Brilliant idea. Awful execution — the capillary clogged constantly.
2. The Coulter Principle: Counting Cells with Electricity (1953)
Wallace H. Coulter sidestepped optics entirely. His insight: cell membranes are lipid bilayers, which are excellent electrical insulators.
Put cells in salt water (an electrolyte), pull them through a microscopic orifice with an electrical current running across it, and every time a cell squeezes through it briefly displaces conductive fluid equal to its own volume. The current dips. The size of the dip tells you the size of the cell.
That's the Coulter Principle, U.S. Patent 2,656,508. It's still the engine inside almost every automated CBC analyzer your hospital runs today — red cell counts, platelet counts, MCV, RDW. All of it.
For the lab director: modern Multisizer-style instruments digitize pulse height, pulse width, and pulse area, which lets the algorithm reject coincidence events (two cells through the orifice at once) instead of silently inflating your count.
3. Hydrodynamic Focusing: The Trick That Makes Everything Work (1953)
Same year as Coulter's patent, P.J. Crosland-Taylor published a paper in Nature solving Moldovan's clogging problem with pure fluid dynamics.
The trick: don't make the tube narrow. Make the stream narrow.
- Inject your sample slowly into the center of a wide chamber.
- Surround it with a fast-moving sheath fluid.
- Keep everything in laminar flow (low Reynolds number, no turbulence).
The sheath physically pinches the sample stream down to a 10–20 µm core — narrower than a single white blood cell. Cells now march through the laser interrogation point in a perfect single file, each one hit by identical light.
For the cytometer operator: turning up sample pressure doesn't speed cells up — it widens the core stream. Too wide and you get coincidence events again. That's why high-purity sorts are run on low sample pressure.
4. Mack Fulwyler, Los Alamos, and the Inkjet Printer (1965)
Here's where it gets weird.
Mack Fulwyler was an engineer at Los Alamos National Lab studying the biological effects of nuclear fallout. When atmospheric weapons testing was banned in 1963, his project evaporated and he pivoted to working on the Coulter Counter.
A pathologist nearby, Clarence Lushbaugh, claimed he was seeing a weird subpopulation of red blood cells on the Coulter trace. Fulwyler thought it was an artifact. To prove it, he needed to physically pull those specific cells out of the stream and look at them under a microscope.
Then he read a paper from Richard Sweet at Stanford about a new continuous inkjet printer that vibrated a fluid stream into droplets, charged specific droplets, and deflected them onto paper with high-voltage plates.
Fulwyler's stroke of genius: do exactly that, but with cells instead of ink.
In 1965, he published in Science the first automated cell sorter. He sorted human red cells from mouse red cells. The entire modern field of cell sorting traces back to a printer.
5. The Fluorescence Revolution: FACS Is Born (1968–1974)
Fulwyler could only sort by size. To sort by identity — "give me only the CD4+ T cells" — you needed fluorescence.
- 1968 — Wolfgang Göhde (University of Münster) built the ICP 11, the first fluorescence-based flow cytometer, commercialized by Partec.
- 1967–1969 — Leonard and Leonore Herzenberg at Stanford visited Fulwyler, took the design home, swapped the impedance trigger for a fluorescence trigger, replaced flickery arc lamps with stable lasers, and built the Fluorescence-Activated Cell Sorter (FACS).
- 1974 — Stanford partnered with Becton Dickinson to release the BD FACS-1, the first commercial fluorescent sorter. BD trademarked "FACS" in 1985, which is why your immunology professor pedantically says "flow cytometer" instead.
Fun fact: Len Herzenberg had famously poor vision and partially built FACS because he physically couldn't keep counting fluorescent cells under a microscope by hand.
6. Inside a Modern Sorter: 12,000 Volts and Microsecond Timing
The physics hasn't really changed since the 60s — just the precision.
- A piezoelectric crystal vibrates the flow cell nozzle at 30,000–100,000 Hz, breaking the stream into uniform droplets at a stable "break-off point."
- The instrument calculates the drop delay — the exact number of microseconds between a cell crossing the laser and that same cell being trapped inside a free-flying droplet.
- If the cell matches your sort gate, a brief electrical charge is applied to the entire stream at the moment that droplet snaps off. The droplet keeps the charge. The stream is instantly grounded.
- The charged droplet falls between two deflection plates carrying ~12,000 volts and is yanked sideways into a collection tube. Uncharged droplets fall straight into waste.
Get the drop delay wrong by one droplet and your "99% pure" sort is garbage. This is why sort QC with fluorescent beads happens before every experiment.
7. Compensation: The Linear Algebra Nobody Warns You About
Fluorophores don't emit at a single wavelength — they emit across a spectrum. FITC's tail spills into the PE detector. PE's tail spills into PerCP. Etc.
Bagwell and Adams formalized this in the late 1980s with compensation matrices — basically a system of linear equations that subtracts the predictable spill from each detector.
For the resident: "compensation" is just matrix algebra. Single-stained controls let the software solve for the spillover coefficients, then it inverts the matrix and applies it to every event. Over-compensation creates the classic "negative population pulled below zero" artifact — a giveaway that someone set comp manually instead of using proper controls.
8. The Clinical Payoff: Leukemia, Lymphoma, and Stem Cells
This is where flow cytometry stopped being a physics toy and started saving lives.
- Leukemia/lymphoma immunophenotyping — multi-color panels (CD45, CD19, CD20, CD5, CD10, kappa/lambda, etc.) distinguish reactive lymphocytes from clonal B-cell neoplasms in minutes instead of days. TRBC1 clonality assessment now does the same for T-cell processes.
- CD4 counts in HIV — the original killer app for clinical flow.
- The ISHAGE protocol — the global gold-standard sequential Boolean gating strategy for enumerating viable CD45-dim / CD34+ hematopoietic stem cells before every bone marrow transplant. Without this number, the transplant doesn't happen.
9. The Future: Spectral, Mass, and Acoustic Cytometry
Classical flow cytometry uses one detector per fluorochrome behind a bandpass filter. Modern instruments throw that away.
- Spectral cytometry (Cytek Aurora, Sony ID7000) replaces individual PMTs with dispersive prisms or gratings that capture the entire emission spectrum of every cell. Spectral unmixing then separates 30–40+ fluorophores with overlapping spectra that would be impossible to compensate the old way.
- Mass cytometry / CyTOF tags antibodies with rare-earth lanthanide metal isotopes instead of fluorophores. Cells get vaporized in a 5,500 K argon plasma torch and the resulting ion cloud is sorted by time-of-flight mass spectrometry. Result: 50+ parameters per cell with zero spectral overlap. The catch — your cells are destroyed, so no sorting.
- Microfluidics and acoustic focusing — lab-on-a-chip devices use standing acoustic waves to align cells without sheath fluid, shrinking entire sort cores onto a credit-card-sized chip.
Pair all of that with t-SNE, UMAP, and AI-driven clustering, and the cytometer of 2030 will look less like a fluidics machine and more like a single-cell data-science platform.
Wrap-Up
In about 60 years, flow cytometry went from "engineer at a nuclear lab borrows an inkjet printer" to the cornerstone of leukemia diagnosis, transplant medicine, HIV monitoring, and modern immunology research.
Every CBC, every leukemia workup, every stem cell transplant in the world today rests on a stack of ideas that includes droplet physics from 1833, electrical insulation of lipid bilayers, laminar flow equations, and a 12,000-volt deflection field that traces straight back to a printer.
Ready to see it animated? Walk through the full timeline — Coulter, hydrodynamic focusing, FACS, spectral, CyTOF — in our interactive Flow Cytometry History tool.