Research Use Only. Not for use in diagnostic procedures.
PredictLung™
AI-Based Biomarker Research for Lung Cancer
Detecting oncogenic driver mutations and biomarkers in non-small cell lung cancer (NSCLC) from H&E-stained tissue images.
The Growing Need for Rapid Biomarker Profiling
Lung cancer is one of the most prevalent malignancies and the leading cause of cancer-related mortality worldwide.
Comprehensive genomic profiling can take several weeks, and some patients start treatment before results are available.
Non-small cell lung cancer (NSCLC) has multiple molecular subtypes, each with distinct treatment pathways.
The number of targeted therapies for NSCLC subtypes continues to grow, and this trend is expected to persist.
Biomarkers in the NCCN Guidelines® for NSCLC | ||
|---|---|---|
EGFR | NTRK1/2/3 | ALK |
KRAS | ERBB2 (HER2) | ROS1 |
BRAF | RET | METex14 / cMET (protein) |
Identifying actionable alterations helps guide first-line targeted therapy. Faster, tissue-sparing approaches to biomarker assessment are an active area of research.3,4
Barriers to Comprehensive Genomic Profiling today

Long turnaround
Often takes 3-5 weeks, in which case results are not available
for early
treatment decisions.

Tissue-intensive
Consumes extra tumor tissue,
which is often scarce, especially in lung cancer. Leaves less material
for further diagnostic testing

Costly
Expensive, limiting access
for many patients
and
many
healthcare systems
PredictLung™: AI-Based Biomarker Research from H&E Images
PredictLung™ is being developed to analyze digitized H&E slides, which are already created for every lung biopsy, and generate biomarker-related probability outputs for research use.
How does it work? At the time of diagnosis, every patient undergoes a biopsy, which is then processed and stained with Hematoxylin and Eosin (H&E) so pathologists can evaluate the cancer, its aggressiveness, and isolate the tumor region for downstream genomic testing. PredictLung™ applies BioAI’s AI models to the digital H&E image to estimate the likelihood of specific genetic alterations. Because it uses an existing image, it does not consume additional tissue. PredictLung™ is under validation and is not intended for diagnosis, patient management or treatment selection.
Where PredictLung™ Fits
Pharma Biomarker Services
Custom H&E-based AI biomarker models and studies for pharma and research partners.
Clinical Trials
Exploring H&E- and IHC-based biomarker pre-screening to support clinical trial screening and enrollment.
Clinical Workflow
Upstream use in clinical laboratory workflows to refine and target definitive blood or tissue biomarker testing.
From H&E Slide to Biomarker Insights
Research Benefits
- Fast analysis
of existing digital slides
- No additional tumor tissue required
image taken from standard tissue slide after biopsy
- Scalable
across large retrospective cohorts
BioAI’s approach may be extended to other tumor types in research settings, such as breast, colorectal and prostate cancer.
PredictLung™ is intended for Research Use Only and is not intended for clinical diagnosis, patient management, treatment selection or use in diagnostic procedures. Any future clinical or regulated use would require further validation and applicable regulatory clearance, approval or laboratory validation.
References
- State of Lung Cancer 2023 report. American Lung Association. https://www.lung.org/getmedia/186786b6-18c3-46a9-a7e7-810f3ce4deda/SOLC-2023-Print-Report.pdf
- Impact of Clinical Practice Gaps on the Implementation of Personalized Medicine in Advanced Non-Small-Cell Lung Cancer. JCO Precis Oncol. 2022 Oct 6: https://pubmed.ncbi.nlm.nih.gov/36315914/
- Oncogenic alterations in advanced NSCLC: a molecular super-highway
https://doi.org/10.1186/s40364-024-00566-0 - Future perspective for the application of predictive biomarker testing in advanced stage non-small cell lung cancer https://doi.org/10.1016/j.lanepe.2024.100839
- NCCN Clinical Practice Guidelines in NSCLC Oncology https://jnccn.org/view/journals/jnccn/22/4/article-p249.xml
- An Ensemble AI Model for RET Alteration Detection Using H&E Images as a Putative Screening Tool for More Efficient Genomic Alteration Detection https://doi.org/10.1089/aipo.2024.0015