Researchers at the Smithsonian National Museum of Natural History have developed a method to digitally image entire microscope slides, making specimens available online. The work was published in PLoS One.
Collection Details
The Denver Pollen Collection, housed at the Smithsonian National Museum of Natural History, contains approximately 70,000 slides of fossilized pollen. The collection holds an estimated 4.3 billion microfossils, exceeding previous estimates for the combined collections of the world's 73 largest natural history museums. This vast repository represents over 60 years of work by the U.S. Geological Survey.
The collection holds an estimated 4.3 billion microfossils, exceeding previous estimates for the combined collections of the world's 73 largest natural history museums.
Digitization Process
The new digitization method leverages high-resolution slide scanners capable of capturing both 2D and 3D images in seconds to minutes.
Crucially, this process preserves information from slides that deteriorate over time through yellowing and cracking. These digital images enable AI-based detection of specimens and provide remote access for verification, protecting fragile originals from handling.
Scientific Applications
Researchers have already used the collection to glean critical insights into past climates. Key findings include:
- Evidence of palm trees north of the Arctic Circle approximately 50–56 million years ago.
- Data indicating Alaska once had a temperate climate suitable for palm trees.
- Measurement of vegetation recovery after the asteroid impact approximately 66 million years ago.
Digital images enable AI-based detection and remote access for verification, protecting fragile originals from handling.
Implementation Requirements
Full digitization of the Denver Pollen Collection would require approximately five years of continuous work. The project would also need a massive 3.5 petabytes of storage capacity to house the resulting high-resolution images.
Access and Benefits
The open-access databases allow students and researchers worldwide to study specimens without handling fragile physical slides. The digitization project achieves three primary goals:
- It preserves deteriorating slides against future physical degradation.
- It enables remote access for experts and students globally.
- It allows AI models to automatically detect, locate, and classify specimens.