
Multiplex fluorescent imaging [Nucleai]
Innovation, in many cases, happens at the overlap of different domains, and that’s where new things emerge. We’re very fortunate to be at the right timing on that,” explained Avi Veidman, CEO of Nucleai.
The convergence of disparate trends — including advances in AI, improvements in imaging technology and the digitization of biological data — is opening up new frontiers in spatial biology. “By combining digitized biology with AI, computer vision, and machine learning, we can uncover entirely new phenomena,” explained Avi Veidman, CEO of Nucleai. “It’s like having a new scientific instrument — first the telescope to see the stars, then the microscope to see cells, and now AI to unlock the deeper secrets within those cells.”
This approach has practical applications in both drug development and patient care. Veidman elaborates on how their technology works in practice: “Our software platform automatically maps biopsy tissue samples of patients,” Veidman said. Usually, those are oncological patients. “We can then correlate these cellular interactions with other layers of information, whether it’s genomic data or historical electronic medical records.”
In addition to its use in research, Nucleai’s technology is finding use in clinical settings. “We’re the first company where our algorithm is part of an active clinical trial and patient enrollment criteria,” Veidman said.
Inside the Nucleai-Proscia alliance

[Nucleai]
The partnership between Nucleai and Proscia comes at a moment when interest is building in spatial biology. Recognized as the 2020 ‘Method of the Year‘ by Nature Methods, spatial biology is transforming our understanding of complex diseases by analyzing the spatial relationships between cells and tissues.
This rapidly evolving field is attracting significant investment, with the top five private companies in the sector seeing a 40% year-over-year increase in capital raised in 2023. This surge underscores the potential of spatial biology in transforming disease diagnosis and treatment across various fields, from oncology and immunology to infectious diseases.
Spatial biology, also known as spatially resolved transcriptomics, allows researchers to study gene expression patterns while preserving the spatial context within tissues. This approach provides a more comprehensive understanding of cellular organization and function compared to traditional bulk sequencing methods. By maintaining spatial information, scientists can uncover how the positioning of cells influences their behavior and interactions, offering crucial insights into tissue architecture and disease progression.
From biopsy data to actionable insights

Avi Veidman
In its approach to spatial biology, Nucleai takes advantage of AI architectures and algorithms, primarily based on convolutional neural networks (CNNs) and other deep learning approaches. These algorithms are trained on datasets annotated by expert pathologists to identify different cell types and regions within pathology slide images. “We have a stack of algorithms that learn from examples trained by pathologists,” Veidman said. “These algorithms identify and locate individual cells and specific areas, such as cancerous regions, within the biopsy sample.”
This AI-guided analysis transforms biopsy data into actionable insights, providing a dynamic understanding of the tumor microenvironment through spatial mapping. Veidman notes that the technology offers potential benefits including improved clinical trial design, enhanced patient recruitment, and personalized treatment selection through companion diagnostics. Veidman emphasized, “We’re helping find the right drug for patients while making clinical trials shorter and more likely to succeed for pharma companies, potentially saving years and billions of dollars in drug development.”
Support for bispecific antibodies and ADCs
Nucleai’s technology is already demonstrating a significant impact on drug development, particularly in areas like bispecific antibodies and antibody-drug conjugates (ADCs). The company says its the first to have their algorithm integrated into the patient enrollment criteria of an active clinical trial.
Nucleai’s software analyzes biopsy stains to identify specific protein expressions on cancer cell membranes, a factor in determining the efficacy of certain oncological drugs. “Let’s say an oncological drug targets the expression of two specific proteins on the membrane of cancer cells,” Veidman said. “Our software can detect the presence or absence of those protein expressions based on biopsy stains.”
By enabling more targeted therapies, Nucleai’s platform has the potential to significantly impact the drug development process. “For pharma companies, we make clinical trials much shorter and increase the probability of success in developing a drug,” Veidman explains. This is particularly crucial in oncology, where trials often struggle to enroll suitable patients and many promising drugs fail to demonstrate efficacy. With Nucleai’s spatial analysis, researchers can identify patients most likely to respond to a specific therapy based on the unique cellular makeup of their tumors. This targeted approach promises to lead to smaller, faster trials with higher success rates, potentially translating to reduced development costs and a quicker path to market.
Veidman is optimistic about the future of this approach, emphasizing its potential in biotechnology. “When you’re coming from a new domain and asking very basic questions — why can’t we do A or B or C – especially when you’re entering that domain at a perfect timing where there’s a perfect storm of AI, machine learning, GPUs, and cloud meeting the digitization of biology, you can really innovate.”
Filed Under: Biospecimens, machine learning and AI