Coulter counted blood cells as electrical pulses

By the late 1940s, a blood cell count was still what it had been for a century: a lab technician diluted a blood sample, placed a drop in a calibrated glass counting chamber, and counted the individual red and white cells visible under a microscope, one by one, extrapolating from a tiny sample to the whole. The count took real time and skill, and different technicians counting the same slide routinely reported different numbers, since it rested entirely on a trained eye's judgment about what it was looking at.

The available fixes were incremental refinements of the same visual method — better staining reagents, more disciplined counting-chamber technique, more rigorously trained technicians — and none of them addressed the two things actually limiting the test: a human eye can only scan so fast, and no two humans see a blurred, overlapping field of cells identically. Wallace Coulter, an electrical engineer who had spent years servicing hospital X-ray equipment and had watched technicians hunched over microscopes doing this by hand, was working the problem from outside the lab rather than inside it.

Forced cells one at a time through a current-carrying aperture; each cell's poor conductivity made a countable pulse, turning a visual judgment into an electrical count.

Cells conduct electricity less than saline, so each cell passing through the aperture creates a distinct voltage pulse.

A circuit can count pulses far faster and more consistently than a human eye can count cells.

The pulse height correlates with cell volume, enabling sizing as well as counting.

The method eliminates inter-operator variability because the signal is physical, not subjective.

To automate a manual inspection, find a measurable physical property that changes with each item, rather than trying to replicate human vision.

Coulter filed for a patent in 1949 — turned away first by patent attorneys who told him "you cannot patent a hole" — and it was granted on October 20, 1953. In his own technical paper published that year, the device counted in excess of 6,000 cells per second in a 15-second interval, a sample size roughly 100 times larger than a standard microscope count, cutting the statistical error of the result by a factor of about 10. The National Institutes of Health independently evaluated two prototypes that same year and published findings confirming the method's improved accuracy and convenience.