Artificial intelligence (AI) utilizes computational tools to help automate processes that would be difficult for a human to complete. With its adoption and spread into a variety of areas, it is unsurprising that it would also serve as a useful tool for scientists. For flow cytometry users, panel building, reagent selection, instrument selection, and data analysis all play major roles in uncovering unique phenotypes that drive our understanding of immune cells. Each of these processes is complex and requires advanced planning. Yue et al. conducted an in-depth assessment of the capabilities of AI and how they could benefit flow cytometry experiments and workflows, including:

• Instrument optimization/standardization
• Sample control, fixation, and treatment
• Reagent design and creation
• Assay development
• Multicolor panel design
• Standard operating procedures (SOPs)
• Automated gating, gating strategies, and software analysis
• Quality control
• Drawing conclusions and diagnosis

These potential AI improvements in the flow cytometry workflow could then be applied in education fields, where the content and training could be personalized to the individual user. It could also be used to share information, evaluate data, and draw conclusions for a diagnosis, particularly if data sets are matched to labs that have previously conducted similar analyses.

While these advancements sound incredible, AI still needs to be given guidance and reliable datasets (e.g. protocols, patient characteristics, flow cytometry results) to efficiently improve its logic. At the same time, there is an inherent human element to AI development, as there must be oversight to ensure: the data fed to AI is unbiased; ethical and privacy concerns are addressed for patients; and complex AI is monitored for data ‘hallucinations’, where inaccurate information can be presented as factual. This last element would be most dangerous when it comes to clinical diagnoses. 

Despite these challenges, AI presents opportunities to improve nearly every facet of the flow cytometry workflow, from bench to bedside. AI in flow cytometry cannot exist sufficiently without its human counterparts, as training, data sharing, and system oversight will be required to ensure the AI becomes dependable when it comes to data interpretation. To add that human element to your planning, talk with our experts to construct panels, advise on experimental design, or aid in troubleshooting.

To learn more about AI in flow cytometry, read the full article by Yue et al. at Cytometry Part B: Clinical Cytometry.

References:

1. Yue, Alice et al. “AI in flow cytometry: Current applications and future directions.” Cytometry. Part B, Clinical cytometry, 10.1002/cyto.b.22255. 23 Sep. 2025, doi:10.1002/cyto.b.22255