VITAscout: AI-Assisted Sample Scouting for Single-Cell QC

VITAscout

The QC gap in single-cell workflows

Library preparation and sequencing have become increasingly standardized, but pre-sequencing microscopy QC often remains manual, operator-dependent, and difficult to scale. Counting cells or nuclei and judging sample state can vary widely across users, especially for challenging inputs like FFPE-derived nuclei and microbial samples, where fragile structures, small cell size, and background noise complicate interpretation.

From “looks OK” to quantitative QC

• Accurate cell/nucleus enumeration.
• Objective assessment of sample state, including aggregation and morphological integrity.
• Quantitative reporting (e.concentration and aggregation rate) to support consistent go/no-go decisions before downstream steps.

Key Capabilities Designed for Single-Cell QC

VITAscout is engineered as a coordinated hardware, algorithm, and workflow system to enable reliable and reproducible use in routine lab settings:

• High-resolution multicolor fluorescence imaging
  (optimized for DAPI and PI) to visualize of nuclei and bacterial cells, including nosier samples.
• Rapid, automated scanning and analysis
  (up to 15 fields of view in ~1 minute) for timely QC results without slowing workflows.
• AI/ML–based recognition
  to distinguish cells or nuclei from debris and background fragments, improving consistency across users.
• Quantitative outputs
  for core QC metrics such as concentration and aggregation status.
• Broad compatibility
  across eukaryotic and prokaryotic samples, including difficult microbiome inputs.
• User-friendly
  operation with minimal hands-on time and training.
• Compact, portable footprint
  for flexible deployment in space-limited environments.

Together, these capabilities enable consistent, quantitative quality assessment across a wide range of single-cell sample types. The following video illustrates how these features are applied in practice to support sample qualification in both eukaryotic and prokaryotic single cell workflows.

Application Highlights

1. QC of Eukaryotic Single-Nucleus Samples from FFPE Tissue

Archived FFPE tissues can be especially challenging for single-nucleus QC due to nuclear damage, heterogeneous morphology, and background fragments. Below is an example from an FFPE lung cancer sample (Figure 1).

Figure 1. Automated QC results for nuclei isolated from FFPE lung cancer tissue.

The original fluorescence image (nuclei in blue) is paired with an annotated output where green boxes mark singlet nuclei and yellow boxes mark aggregates. The lower panel summarizes the quantitative QC outputs generated by VITAscout.

In this example, VITAscout reported:

• Nuclear concentration:1,918.90 nuclei/μL
• Total nuclei (100uL): 191,890 nuclei
• Aggregation rate:8%.

Using the stated qualification criteria (≥50,000 nuclei and aggregation rate < 20%), the sample met minimum requirements to proceed, providing a clear and objective basis for downstream decisions.

2. QC of Prokaryotic Single-Cell Samples (Single Species and Microbiome)

Microbial single-cell analysis requires accurate discrimination between small bacterial cells and background debris, a task that is particularly challenging under conventional microscopy. Trained on large-scale microbial single-cell datasets generated using the VITA platform, VITAscout enables stable identification and quantification of bacterial cells across both single-species and microbiome samples.

In a fecal microbiome sample stained with PI (Figure 2), VITAscout reported:

• Bacterial concentration:2,720.34 cells/μL.
• Ttotal cells (100uL): 2,720,340
• Aggregation rate:2%.

Using the stated qualification criteria (≥50,000 cells and aggregation rate < 20%), the sample met minimum requirements to proceed.

Figure 2. Automated QC results for bacterial cells isolated from a fecal sample. 

Original PI fluorescence image (cells in red) is paired with annotated ouput where green boxes mark singlets and yellow boxes mark aggregates. The lower panel summarizes the quantitative outputs.

3. Quantitative Microbial Cell Counting Beyond Transcriptomics

Beyond single-cell transcriptomic workflows, VITAscout can also be applied to quantitative microbial cell counting for microbiology research workflows (e.g. antibiotic response profiling and functional microbiology studies).

In a cultured E. coli example (Figure 3), VITAscout reported:

• Bacterial concentration:2,122.73 cells/μL

This provides a reproducible baseline for normalizing downstream assays and comparing bacterial populations across conditions.

Figure 3. Automated QC results for bacterial cells from a cultured Escherichia coli sample. 

Original PI fluorescence image (cells in red) is paired with annotated output (singlets vs aggregates) The lower panel summarizes the quantitative QC outputs.

From Application to Standardized QC

Across these examples, VITAscout converts microscopy-derived observations into consistent quantitative metrics across diverse single-cell sample typeshelping laboratories standardize sample qualification, reducing inter-operator variability, and improve reproducibility, especially for complex or non-standard samples.

As a core component of the VITA single-cell transcriptome sequencing platform, VITAscout establishes a framework for AI-assisted, sample-centric monitoring by integrating imaging and intelligent analysis into a unified workflow, supporting more confident decisions before single-cell profiling. Driven by real-world research challenges, M20 Genomics remains committed to the development of original technologies that address unmet needs in single-cell research, continuously translating methodological innovation into practical and reproducible tools for the research community.

For research use only. Not intended for diagnostic or clinical applications.