After attending the talks, exploring the exhibition floor and engaging with the conference attendees, we walked away with some key insights that highlight the direction the industry is heading. Here are four takeaways that stood out to us and captured the major themes of the conference.
Organoids take the prize
One of the hottest topics at the start of 2025 were organoids and 3D imaging. In fact, Taci Pereira of Systemic Bio, a 3D Systems company based in Houston, TX, took home this year’s SLAS Innovation Award. This award recognizes groundbreaking technological advancements that drive innovation in laboratory science and automation. Pereira’s presentation titled “h-VIOS™: A human-relevant drug discovery and development platform using bioprinted human tissues” was awarded a $10,000 prize, sponsored by HighRes Biosolutions.
3D Systems, like numerous other organoid companies represented at SLAS, focuses on developing organ models for research and drug discovery. In particular, the research presented by Pereira demonstrated the application of its platform to evaluate the safety of antibody-drug conjugates (ADCs), enabling early identification of safety concerns. Safety risks are typically identified during clinical trials and are one of the main reasons for drug approval failure. 3D organoids are disrupting this drug safety research space by allowing researchers to evaluate safety risks prior to testing the drug in the clinic and reducing the need for non-human primate studies.
3D imaging pushes the boundaries of phenotypic screening
A research area that goes hand-in-hand with organoid studies is 3D imaging. Organoids, which is a biological model designed to capture the morphological structure and function of real organ tissues, offer many advantages over conventional 2D tissue culture; however, studying them presents a new challenge for researchers who want to investigate phenotypic changes associated with different types of perturbations. 3D imaging is an advanced cell imaging approach, which has been increasingly utilized in organoids research.
Computer vision and machine learning are revolutionizing the analysis of 3D images of organoids by enabling automated, high-throughput and quantitative analysis. With the help of these tools, researchers can extract meaningful biological information from complex 3D tissue structures and significantly accelerate analysis. For example, computer vision and AI can help with feature extraction, such as measuring organoid shape, size, volume, and surface roughness; dynamically tracking morphological changes over time; and performing cellular composition analysis via multi-channel imaging. Altogether, these advanced computational methods expand the range of possibilities of organoid research and improve its efficiency, reproducibility and accuracy.
AI and computer vision are essential not only for 3D tissue analysis but for other types of imaging as well. For example, AI is transforming phenotypic profiling and high-content screening (HCS) by helping researchers extract deeper insights faster. Krzysztof Rataj, Ardigen’s Cheminformatics Expert, showcased the analysis of HCS imaging research with Ryvu Therapeutics in a poster titled “Artificial Intelligence predicts cell proliferation from DAPI images of (hISC)-derived colorectal cancer models.” You can view the poster here.
Negative results transform into valuable data for AI
A theme that ran throughout many of the conference discussions was the fact that negative results, which historically have been discarded as “unpublishable data”, are becoming an important resource in the age of AI. Most researchers do not publicize the results of failed experiments that do not support the hypothesis; however, negative data are critical for model training, helping establish parameters and putting guardrails for AI models.
Specifically, in AI-driven drug discovery, negative results play a crucial role. They help models differentiate between ineffective compounds (true negatives) and misleading candidates (false positives), reducing the risk of biased predictions due to model overfitting to successful drugs. Training with negative data broadens the model’s understanding of chemical space, improves drug candidate selection, and enhances predictions of toxicity and resistance. AI-driven toxicology models, for example, leverage failed clinical trial data to refine drug safety assessments.
Nasim Jamali, Director of Morphological Profiling at Ardigen, discussed this among other topics during a panel discussion on Phenotypic Drug Discovery. Overall, SLAS 2025 showed that AI and ML have already become a mainstay of the industry, transforming all areas of research, from lab automation to drug discovery and more.
Driving discovery through collaboration
The Ardigen team enjoyed all the stimulating discussions and connecting with brilliant minds in lab automation and drug discovery at SLAS 2025. With so many exciting announcements and innovative perspectives, we are proud to contribute to the research advancing biomedical science and discovery. If you want to know more about how to use advanced computational tools, such as AI, machine learning and computer vision, to harness the power of automation and big data, reach out to one of our experts!
