News

In cancer diagnosis, pathology is regarded as the gold standard. For decades, pathologists have reached their conclusions by examining one glass slide after another under a microscope. In recent years, however, a new practice has spread rapidly: scanning an entire slide into a high-resolution digital image, then reading, consulting and archiving it on a computer screen. This is pathology slide scanning — whole-slide imaging (WSI) — and it has become one of the hottest topics in pathology laboratories over the past two years.

1. How a glass slide becomes digital

The principle is not complicated. Using an optical system at low magnification, the scanner photographs an entire slide frame by frame while the stage moves automatically in the X-Y direction and auto-focuses along Z so that every frame stays sharp. Software then stitches thousands of small images seamlessly into a single full-field, high-resolution digital slide. It is like taking a panoramic photo of the whole slide that can be magnified almost endlessly: fine detail keeps emerging, and the viewer can zoom and pan just as if operating a real microscope. A single digital slide often runs to several gigabytes, which is why it places real demands on storage and computing power.

2. Why it has taken off in the past two years

First, policy. In April 2026, China's National Healthcare Security Administration included digital pathology slides in its pricing guidelines for pathology medical services, giving hospitals a basis to charge for digital slide services and AI-assisted diagnosis. When the billing is clear, investment becomes far easier.

Second, AI in pathology has truly arrived. A prominent example is the pathology model developed by Huawei with Ruijin Hospital, trained on millions of digital slides and able to assist in diagnosing common cancers within seconds; several AI-assisted products have received medical device registration. Because AI must be fed digital slides, hospitals that want AI first need to digitise their slides.

Third, domestic substitution is accelerating. Whole-slide scanning systems are on the list of categories given priority for domestic procurement, and Chinese manufacturers have matured, lowering the barrier to purchase.

Fourth, data volumes are exploding. Hundreds of millions of new pathology slides are produced worldwide every year, and China's move to digitise pathology departments is entering a phase of concentrated expansion.

3. What to look for in a scanner

Throughput — how many slides can be loaded at once. Up to 20 is low, 21 to 100 is medium, and over 100 is high throughput. Hospitals pursuing department-wide digitisation tend to choose medium or high throughput and scan continuously overnight.

Resolution — commonly around 0.25 microns per pixel; smaller numbers mean sharper images but larger files.

Speed — recent ultra-fast models cut a single-slide scan to around ten seconds, directly affecting a department's turnaround.

Focus and colour consistency — whether auto-focus is reliable and whether colour is consistent across machines decides whether images can be used for diagnosis and AI analysis.

Software and ecosystem — how good the viewer is, whether it connects to the hospital information system, and whether it supports remote consultation and AI analysis are increasingly decisive in selection.

4. What it is used for

The core use is clinical diagnosis, especially remote consultation and multidisciplinary collaboration: once slides become digital images, experts no longer need to ship physical slides and can review a case hundreds of miles away within minutes. Next is pathology teaching, where virtual microscopy lets dozens of students view the same classic case at once, without the specimens fading or breaking through repeated use. Then come research and drug development, biobanking, and supplying training data for AI models.

5. Where it is heading

The industry is moving from selling equipment to integrated hardware, software and service. Cloud is the clear direction: centralised storage and sharing across sites turn pathology from individual labour into team collaboration. Meanwhile AI is moving upstream — new scanners are beginning to embed image pre-processing and quality checks, completing preliminary analysis during scanning, so that "scan, then compute" becomes "compute while scanning".

Conclusion

The point of pathology slide scanning is not merely to move glass slides onto a computer. It is to move pathology diagnosis from a single person at a single microscope toward a form that is shareable, computable and collaborative. For hospital pathology departments and medical schools, the question is shifting from whether to build such a system to when to build it.