Realizing Cellino's vision of revolutionizing regenerative medicine

Manual intervention in cell manufacturing - a biologist examining cells by eye and removing unacceptable ones by pipette - introduces operator variability and makes it hard to deliver stem-cell-based therapies affordably at scale. Cellino needed an interactive web application that could present cell manufacturing data like high-resolution microscopy images, give investors and the advisory board a tangible sense of the idea, and accelerate the MVP launch of its research-grade stem cell manufacturing platform on a time-boxed schedule.

We partnered with Cellino to integrate the external lab instrument controller systems and front-end action items with a constantly improving AI engine, and to build an Automation Information Management System (AIMS) to represent all data generated from the manufacturing processes of many cell lines. This system would be a listener and viewer for the colossal amounts of data captured, analyzed, and sent over by their instrumentation as well as the Artificial Intelligence (AI)/Machine Learning (ML) core. AIMS helped biologists monitor cell manufacturing processes in the following ways:

• Data capture and management: Multitudes of data points, including n-dimensional high-resolution, multi-scaled images of cell clusters get sent to AIMS to be converted to human readable form. AIMS ensures a traceable transition, and seamlessly integrates them into existing workflows to identify and nurture the best stem cells.

• Rendering images in the image viewer: Our teams worked with Cellino to build a highly capable image viewer for microscopic instrumentation and other inference images: high-resolution, multi-channel images layered with AI-derived data such as labels, points, shapes, and surfaces. Using web-optimized tiling, the AIMS viewer renders this massive data in milliseconds and enables seamless zooming to very high-resolution layers. Users can examine related metadata, including protocols, derived features, and administrative data, in each layer's attribute panel.

• Play around with brightfield and fluorescence images: Biologists can utilize the image viewer to interact with sub-micron resolution images with various modalities at a lightning speed. They can adjust brightness, colormaps, opacity, contrast, select channels, adjust layer blending modes, annotate, draw, and do much more with the images for observation and inspection of cells.

• Collaboration: Our Slack integration enables seamless communication and collaboration across lab technicians, biologists, and engineering teams, as many cell lines are run through the manufacturing platform.

• Intuitive UI: The user-friendly interface of AIMS presents all this data, from manufacturing run information to plate status, in a way that's easy to consume, and even easier to analyze with connections to BI tools. Users then quickly make informed decisions about the quality of each cell cluster and identify next steps of experimentation.

By building AIMS as a fast, web-optimized viewer for massive, multi-channel microscopy data, Zemoso gave Cellino's biologists a way to examine and annotate cell clusters in milliseconds - turning a research-grade prototype into a tool that could support both scientific decision-making and investor conversations.