Altis Labs is a Toronto-based computational imaging company that uses deep learning models trained on a very large real‑world cancer imaging database to help biopharma sponsors run smaller, faster, and more successful clinical trials by predicting patient outcomes from existing scans and generating digital‑twin control arms for trials[3][1].
High‑Level Overview
- Mission: Altis Labs aims to advance precision medicine by operationalizing medical imaging as a predictive source of clinical insight to improve patient outcomes and accelerate therapeutic development[5][3].
- Investment philosophy / key sectors / impact on startup ecosystem: (Not an investment firm — Altis is a portfolio company / product company focused on healthcare AI and computational imaging serving life‑sciences and clinical research markets)[7][1].
- What product it builds: Altis builds Nota, a cloud‑based computational imaging platform that hosts prognostic AI models to predict clinically meaningful outcomes from imaging scans and to manage and analyze imaging data for trials[4][3].
- Who it serves: Primary customers are biopharmaceutical sponsors and clinical development teams (examples cited include AstraZeneca and Bayer), as well as clinicians and cancer research centers partnering on imaging research[1][3].
- What problem it solves: Nota quantifies treatment effect earlier and more accurately from existing imaging, enabling better patient stratification, creation of computational control arms (digital twins), and shortened or de‑risked clinical development timelines[3][4].
- Growth momentum: Altis reports large-scale training data (reported figures range in coverage but have been cited as over 182 million to 220+ million real‑world images) and published peer‑reviewed research with clinical collaborators, and has active commercial engagements with major biopharma partners, indicating accelerating commercial traction[3][2][5].
Origin Story
- Founding year and founders: Altis Labs was founded by Felix Baldauf‑Lenschen (Founder & CEO) with co‑founders including Sally Daub (Co‑Founder & Chair) and early engineering leadership such as Matt (lead engineer) who joined the founding team in 2019; the company is headquartered in Toronto, Canada[5][3].
- How the idea emerged: The team formed around the conviction that medical imaging is an underutilized, information‑rich clinical data source and that deep learning trained on real‑world imaging plus longitudinal clinical data could predict outcomes useful for drug development and clinical care[5][4].
- Early traction / pivotal moments: Altis built what it describes as the world’s largest real‑world cancer imaging database in collaboration with Canadian cancer centers and has co‑published research in venues including the Journal of Clinical Oncology and presented at the American Society of Clinical Oncology; it has also announced collaborations with major pharma (AstraZeneca, Bayer) to pilot digital‑twin approaches in trials[1][3][5].
Core Differentiators
- Data scale and curation: Access to an extremely large, longitudinal cancer imaging database (reported figures: 182M–220M+ images) with linked diagnostics, treatments, and outcomes, which supports robust model training and external validity claims[2][3].
- Specialty product — Nota: A purpose‑built, cloud platform that ingests imaging from past and ongoing trials and runs prognostic models to produce patient‑level outcome predictions and computational control arms[4][3].
- Clinical & regulatory experience: Team includes clinicians and engineers with experience building regulatory‑grade imaging AI products and co‑authoring peer‑reviewed clinical research, supporting credibility for deployment in trials[5][4].
- Biopharma partnerships: Active incorporation of Altis predictions into statistical analyses by large sponsors (AstraZeneca, Bayer) demonstrates enterprise adoption and pragmatic utility in trial decision making[3][1].
- Demonstrated performance: Example results cited include substantial improvements in outcome prediction (e.g., a reported 48% improvement in lung cancer outcome prediction versus standard approaches in validation work)[4].
Role in the Broader Tech Landscape
- Trends they are riding: Altis sits at the intersection of clinical AI, real‑world data (RWD), and digital clinical trial innovation — specifically the use of imaging‑derived biomarkers and synthetic/computational control arms to make trials more efficient[3][4].
- Why timing matters: Rising costs and long timelines in oncology drug development, broader regulatory and industry interest in RWD and synthetic controls, and maturation of deep learning for imaging create strong tailwinds for platforms that can reliably extract outcome signals from routine scans[4][3].
- Market forces in their favor: Pharmaceutical pressure to de‑risk pipelines, greater acceptance of AI tools in clinical workflows, and improvements in cloud infrastructure for large‑scale model training enable faster commercialization of imaging AI for trials[3][4].
- Influence on ecosystem: By operationalizing imaging for prognostic insights and partnering with major cancer centers and pharma, Altis helps normalize use of imaging AI in trial design and could accelerate adoption of digital‑twin and RWD methods across oncology development programs[1][3].
Quick Take & Future Outlook
- What’s next: Expect continued expansion of Nota’s model library and dataset coverage, deeper regulatory and statistical validation of digital‑twin methods, and more integrated deployments with sponsors running adaptive or model‑augmented trials[3][4].
- Trends that will shape their journey: Regulatory guidance on synthetic controls and RWD, reproducibility/validation standards for imaging AI, and biopharma’s appetite to replace or augment conventional controls with computational methods will be decisive[3][4].
- How their influence might evolve: If Altis sustains strong external validation and broader pharma adoption, it could become a standard imaging‑AI partner for oncology trials—reducing sample sizes, shortening timelines, and shifting how imaging is valued in development programs[3][1].
Quick reminder: this profile is drawn from Altis’s published materials, industry coverage, and innovation briefs; specific figures (e.g., number of images) are reported by the company and press releases and have appeared with slightly different totals in different sources (182M vs. 220M+), which I noted above[3][2][4].
