Stereo Investigator®AI
The unbiased stereology you trust — without the weeks at the microscope.Gold-standard cell counting, automated by machine learning and built by the people who set the standard. So your expertise goes where it matters: the science, not the counting.
Product Overview
Unbiased stereology is the gold standard for quantifying cells in tissue — and it’s also one of the most time consuming jobs in the lab. Counting by hand can mean weeks at a microscope, and even then the result depends on who’s doing the counting and how trained or tired they are.
Stereo Investigator AI changes that equation. Built by the same stereology experts who helped define the field, it uses machine learning to replicate the judgments of an expert human observer — recognizing each cell, where it sits, and how big it is. Once you’ve trained it, it identifies cells in 3D image volumes across entire brain regions using the same criteria you would apply yourself, only far faster. The rigor stays intact. The bias controls stay intact. What changes is your time: studies that once took months are finished in a fraction of that — putting big-science throughput within reach of a single lab.
Big Science in Small Labs.
Key Benefits
- Trust every number. Deep-learning algorithms deliver accurate, unbiased, reproducible counts — consistent across runs and across operators.
- Work with the images you already have. Single- and multi-channel fluorescence microscopy, supported out of the box.
- See what's co-labeled. Detect co-localized cellular and sub-cellular markers.
- Get your time back. Substantially faster than manual stereology — complete whole studies in a fraction of the time.
- Better than guesswork. More accurate and more rigorous than ad hoc 3D detection methods that now proliferate.
Stereo Investigator AI pairs state-of-the-art machine learning with the methods stereologists rely on: unbiased counting rules, systematic random sampling (SRS), and clear, learnable workflows for classifying and quantifying cells in 3D. It automates the most widely used unbiased stereological probe for counting cells — the optical fractionator.
Working from the counting frame, grid size, and disector parameters you define, Stereo Investigator AI detects every cell at each counting site. It then applies the counting rules automatically, keeping only the cells that fall within the disector — exactly as you would by hand.
It can be trained to tell different cell types, sub-cellular objects, and non-cell objects apart. Because neuron density varies from region to region, classifiers can be trained on both dense and sparse populations — so entire studies process cleanly, start to finish.
Recommended Hardware Requirements
| 64-bit Windows 11 operating system |
| CPU with 12 cores (16 threads) or more. More cores improve performance when using large data sets (>1 GB). |
| Solid state hard drive(s). Preferably, non-volatile memory express (NVMe) drives. |
| 64 GB of system memory or more. More memory is better for large data sets, especially for image handling and reconstruction in the 3D environment. |
| Graphics card with 8 GB memory or more. Most graphics cards from NVIDIA have been tested with MBF Bioscience software. |
| Minimum Hardware Requirements |
|---|
| 64-bit Windows 11 operating system |
| 8 core processor (16 threads) |
| 32 GB memory |
| Graphics card with 6 GB memory or more: This is the minimum needed for the 3D environment |
| Computer-Hardware Upgrade Priorities |
|---|
| To upgrade your system for better performance with MBF Bioscience software, we suggest that you prioritize computer hardware upgrades as follows: |
|
|
|
Please contact us to consult with our technical specialists for a more precise recommendation specific to your data needs.
Download Stereo Investigator AI brochure here.
Combining machine learning with stereology: The next generation of unbiased 3D stereology for cell counting
The next generation of unbiased 3D stereology for cell counting: Automatic cell counts generated by Stereo Investigator’s artificial intelligence technology are analogous to manual cell counts
Stereo Investigator AI: Applying deep learning methods for automatic cell counts in a new unbiased 3D stereology software package
Key Features
Stereology Parameters
A clear, guided workflow walks you through setting up a study — drawing contours and defining the counting frame, grid sizes, and disector parameters. Powerful where it counts, simple where it should be.
Key Features
Calculate Section Thickness
Stereo Investigator AI was built for real labs, where not every section is perfectly stained and thickness varies. It measures section thickness reliably under those conditions — and testing shows results consistent with manual measurement by expert users.
Key Features
Intelligent 3D Image Segmentation
To remove the bias that comes with standard segmentation and thresholding, we built an intelligent, automated parameter-estimation step. It analyzes individual sub-volumes containing many objects and tunes the key segmentation parameters to best match each object’s structure — so you’re not hand-tuning thresholds image by image.
Key Features
Train the Machine Learning Classifier
You train the classifier by marking the center and diameter of cells in a handful of small sub-regions. In one published study, training drew on roughly 90 sub-regions across the full rostro-caudal extent and took about an hour. That hour of your expertise then carries across the entire study.
Key Features
Apply Counting Rules and Calculate Population Estimates
Objects are detected automatically in 3D at each counting-frame site. The unbiased counting rules are then applied so that only cells falling within the optical disector are counted. From there, the established design-based stereology formulae produce statistically robust population estimates — with coefficient-of-error estimates so you can judge the accuracy of every result.
Who Is Using MBF Products?
MBF products are used across the globe by the most prestigious laboratories.
Cited in Peer Reviewed Scientific Publications
The utility of Stereo Investigator AI is underscored by the number of references it receives in the world’s leading scientific publications. See examples below:
Boeglin, M., E. Leyva-Díaz, et al.
Expression and function of C. elegans UNCP-18, a paralogue of the SM protein UNC-18View Publication
Rentsch, P., T. Egan, et al.
The ratio of M1 to M2 microglia in the striatum determines the severity of L-Dopa-induced dyskinesiasView Publication
Wang, Z., D. Zheng, et al.
Enabling Survival of Transplanted Neural Precursor Cells in the Ischemic BrainView Publication
Villar-Conde, S., V. Astillero-Lopez, et al.
Synaptic involvement of the human amygdala in Parkinson’s diseaseView Publication
Stimpson, C. D., J. B. Smaers, et al.
Evolutionary scaling and cognitive correlates of primate frontal cortex microstructureView Publication
Russ, T., L. Enders, et al.
2,4-Dichlorophenoxyacetic Acid Induces Degeneration of mDA Neurons In VitroView Publication
Olkhova, E. A., C. Bradshaw, et al.
A novel mouse model of mitochondrial disease exhibits juvenile-onset severe neurological impairment due to parvalbumin cell mitochondrial dysfunctionView Publication
Zhang, X., C. Wang, et al.
Analysis of Error Sources in the Lissajous Scanning Trajectory Based on Two-Dimensional MEMS MirrorsView Publication
Lu, J., Behbahani, A.H., Hamburg, L. et al.
"Transforming representations of movement from body- to world-centric space."
Frequently Asked Questions (FAQ)
Does Stereo Investigator AI work with fluorescent images?
Yes — including multichannel fluorescent images.
Can I audit the probe runs?
Yes. You can audit and inspect every result in the dynamic 3D visualization environment to check its accuracy for yourself.
Can I create different classifiers for different cell populations?
Yes. You can train separate classifiers for different cell types or imaging methods. Stereo Investigator AI ships with several classifiers built in — and if you'd rather not train your own, we'll train a classifier for your specific experiment for you.
Will it report estimated cell totals and coefficients of error?
Yes. Results are computed using established stereological formulae, including coefficient-of-error estimates.
Testimonials
"I rarely have encountered a company so committed to support and troubleshooting as MBF."
"MBF Bioscience is extremely responsive to the needs of scientists and is genuinely interested in helping all of us in science do the best job we can."
"I am so happy to be a customer of your company. I always get great help related with your product or not. With the experienced members, you are the best team I've ever met. All of your staff are very kind and helpful. Thank you for your great help and support all the time."
"We’ve been very happy for many years with MBF products and the course of upgrades and improvements. Your service department is outstanding. I have gotten great help from the staff with the software and hardware."
"Our experience with the MBF equipment and especially the MBF people has been outstanding. I cannot speak any higher about their professionalism and attention for our needs."
"MBF provides excellent technical support and helps you to find the best technical tools for your research challenges on morphometry."
Robust Professional Support
Our service sets us apart, with a team that includes Ph.D. neuroscientists, experts in microscopy, stereology, neuron reconstruction, and image processing. We’ve also developed a host of additional support services, including:
- Forums
We have over 25 active forums where open discussions take place.
>> Learn More - On-Site/Training
We’ve conducted over 750 remote software installations.
>> Learn More - Webinars
We’ve created over 55 webinars that demonstrate our products & their uses.
>> Learn More
Request More Information
At MBF, we’ve spent decades learning how researchers and their labs actually work — and we’ve built our tools around what you need, not the other way around. Before you decide anything, we’ll sit down with you and talk through your study, your tissue, and your goals, so the solution you get is the one your lab will actually use. That’s our commitment to you, before, during, and long after the decision. We’re the technology behind your next discovery — and we’d love to talk.
Big Science in Small Labs.
Related Products
Stereo Investigator®
The complete stereology solution. The gold standard for unbiased cell counting.
SLICETM - The New Standard for Light Sheet Imaging
SLICE: a paradigm shift in light-sheet microscopy that combines high performance with unprecedented affordability and a compact device footprint.