Plasma‑Based Multi‑Omics AI for Early Lung Cancer Detection
BAIRI (百利研究院) — Biotech AI Research Institute
One blood test. AI‑powered precision. Results in days, not months. Our mission: transform lung cancer and tuberculosis diagnosis through affordable, non‑invasive, AI‑driven blood tests that save lives and reduce healthcare costs.
Contact Us
The Problem
150 Million People at Risk
Lung cancer is the deadliest cancer in China — and the window for life-saving intervention is being missed on a massive scale.
28.5%
Cancer Deaths
Share of all cancer deaths in China attributable to lung cancer
64.6%
Late-Stage Diagnosis
Patients diagnosed at Stage III or IV, where 5-year survival drops below 20%
95%
Benign Nodules
Of CT-detected pulmonary nodules are benign — yet indistinguishable without biopsy
<70%
Biomarker Sensitivity
Traditional protein biomarkers fall below this threshold, leaving critical gaps
CT scans detect pulmonary nodules in 56% of screened individuals, yet the overwhelming majority are benign. Clinicians face an impossible choice: perform invasive biopsy or wait months for follow-up imaging. NGS-based liquid biopsies cost RMB 5,000–10,000 and are slow. The world urgently needs an affordable, accurate, non-invasive alternative.
One Blood Test. AI‑Powered. Over 90% Accuracy.
The BAIRI IPM (Integrated‑Plasma Metabolomics) Test begins with a simple blood draw — no biopsy, no radiation, no prolonged waiting. Our proprietary PPdPM technology simultaneously measures three molecular layers: plasma proteins, amino acids, and bile acids.
1
Blood Draw
Simple, non-invasive plasma sample collected at point of care
2
PPdPM Analysis
Simultaneous proteomics, amino acid, and bile acid profiling
3
AI Integration
Machine learning integrates multi-omics layers with clinical data
4
Risk Score
Clear malignancy score: low, intermediate, or high — for decisive clinical action
For a physician facing an inconclusive CT scan, the BAIRI test acts as a clear, evidence-based tie-breaker — delivering actionable results within days, not months.
Core Technology
AI Makes the Difference
Why Single-Marker Tests Fail
Traditional diagnostic tests evaluate one biomarker at a time. Lung cancer is biologically complex — no single protein or metabolite carries sufficient discriminatory power. Conventional methods plateau below 70% sensitivity, leaving too many patients undetected or misclassified.
The biological signal for malignancy is distributed across hundreds of molecular features. Capturing it requires a fundamentally different approach.
How BAIRI's AI Works
Our machine learning model ingests hundreds of features from plasma proteomics, metabolomics, and structured clinical data simultaneously — identifying patterns invisible to any human clinician or single-marker assay.
With over 3,000 samples today and 10,000+ planned, the model continuously improves. This creates a compounding data-driven moat that competitors cannot easily replicate. Each new sample strengthens the algorithm's precision and generalizability across patient populations.
Clinical Validation
Validated by Science
Based on retrospective studies with over 3,000 samples, supported by prospective trials and rigorous blind testing across multiple institutions.
92%
NSCLC Sensitivity
Screening sensitivity for non-small cell lung cancer
90%
Screening Specificity
Confirmed across retrospective and prospective cohorts
82%
Nodule Differentiation
Benign vs. malignant pulmonary nodule classification accuracy
Our research has been published in peer-reviewed journals including Respiratory Research (2026), Biomarker Research (2023), Briefings in Bioinformatics (2021), and Proteomics (2017). In TB differentiation, BAIRI achieves double 90% — both sensitivity and specificity exceeding 90% simultaneously.
Real‑World Validation in Henan, China
In 2025, BAIRI completed a prospective cohort study across Zhengzhou and Nanyang, Henan Province (February–July 2025), demonstrating strong real-world performance across multiple patient cohorts.
TB Group
91.7% accuracy
12 active tuberculosis patients correctly classified
Healthy Group
100% accuracy
11 healthy individuals — zero false positives
Non-TB Pulmonary
93.5% accuracy
31 patients with non-TB pulmonary diseases correctly differentiated
Our TB database now exceeds 4,000 validated samples. A 20,000+ multi-center trial is planned to further solidify clinical evidence and support regulatory submissions in China and internationally.
Product Portfolio
Three Products. One Platform.
Lung Nodule Differentiation Test
Differentiates benign from malignant pulmonary nodules with >90% accuracy. Affordable for large-scale population screening. Designed as a CT tie-breaker for indeterminate findings.
Tuberculosis Diagnostic Test
Distinguishes active TB from healthy individuals and other lung diseases. Achieves double 90% — both sensitivity and specificity exceed 90% — addressing a critical unmet need in high-burden regions.
Longevity Medicine & Wellness Tests
NAD+ detection and enhancement, sleep management via food-drug homology, and metabolic anti-aging panels. Available now with no regulatory clearance required — sold directly to clinics, wellness centers, and consumers.
A Platform, Not Just a Test
The PPdPM platform's multi-omics architecture is inherently scalable. The same proteomics, metabolomics, and AI infrastructure that powers lung nodule differentiation can be extended across oncology, chronic disease, drug development, and TCM modernization.
The potential total addressable market for the full platform exceeds RMB 100 billion. The longevity product line — already generating revenue — is the first proof point of successful platform extension beyond core oncology diagnostics.
Competitive Moat
Patented. Published. Proprietary.
Intellectual Property
Three granted Chinese invention patents protect the core PPdPM methodology and AI integration framework:
  • ZL 202210865892.6 — PPdPM methodology core
  • ZL 202111060620.0 — AI integration framework
  • ZL 202110352760.9 — Diagnostic application
These patents create a durable IP barrier around BAIRI's foundational technology, covering both the biological assay and the computational architecture.
World's Largest PPdPM Database
BAIRI holds the world's largest plasma multi-omics database for lung nodule differentiation — 3,000+ samples today, expanding to 10,000+. For tuberculosis: 4,000+ samples validated with a 20,000+ multi-center trial planned.
Peer-reviewed publications in Respiratory Research, Biomarker Research, Briefings in Bioinformatics, and Proteomics provide independent scientific validation — building trust with clinicians, regulators, and partners.
Hospital‑Ready AI Appliance
BAIRI's set-top box is a dedicated hardware/software appliance purpose-built for hospital deployment. It brings AI inference directly to the point of care — eliminating cloud dependency and protecting patient data sovereignty.
Direct Instrument Integration
Captures data directly from mass spectrometers at the hospital lab — no manual data transfer, no errors.
On-Device AI Inference
Runs the full IPM algorithm locally. Self-computing, with no external cloud dependency required for operation.
Data Secure by Design
No patient data leaves the premises. Hospital-dedicated architecture ensures compliance with China's data security and privacy regulations.
Clinical Report Output
Generates structured clinical reports ready for physician review, integrating with existing LIS/HIS hospital information systems.
Market Opportunity
Serving China. Expanding Globally.
Mainland China
150 million existing lung nodule patients. 20 million new nodules detected annually. TAM exceeds RMB 15 billion/year (USD 2.1 billion) for lung nodule differentiation alone — a market with virtually no accurate, affordable blood-based solution today.
Hong Kong
6,111 new lung cancer cases diagnosed in 2023. High private healthcare penetration and early-adopter clinician culture make Hong Kong an ideal launch market for premium diagnostics ahead of broader China rollout.
Global Market
The global lung cancer diagnostics market was USD 1.54 billion in 2024, projected to reach USD 2.31 billion by 2030. The multi-omics diagnostics global market is USD 2.5–3.3 billion, growing at 15–17% CAGR. Total global opportunity for AI-driven multi-omics lung nodule differentiation exceeds USD 10 billion annually.
No Direct Competitor
BAIRI Stands Alone
No other company combines plasma proteomics, metabolomics, and AI for lung nodule differentiation at affordable, scalable pricing. The competitive landscape is fragmented and incomplete.
Regulatory Path
Clear Path to NMPA Class III Approval
The total regulatory timeline spans 18–30 months, with a mid-point estimate of 24 months. Part A involves a multi-center prospective trial enrolling 1,200 patients across qualified clinical sites in China. Part B encompasses NMPA Innovation Designation (enabling priority review), full technical documentation review, and final device approval. Contingency plans are in place to maintain momentum throughout the process.
NMPA Innovation Designation
Priority review designation reduces overall review timeline and signals regulatory alignment with BAIRI's novel AI-driven diagnostic category.
Parallel Hong Kong Strategy
Adrian Watson leads MDACS regulatory compliance in Hong Kong, enabling earlier market entry and real-world evidence generation ahead of full China approval.
Leadership
World-Class Team
BAIRI's leadership brings together expertise in investment banking, molecular genetics, artificial intelligence, clinical operations, and regulatory strategy — purpose-built to commercialize a breakthrough diagnostic platform.
Julian So — CEO
Legally qualified in Singapore, Hong Kong, and the UK. Former investment banker at UBS, RBS, and Morgan Stanley. CEO of multiple tech companies; Chairman and Executive Director of public and listed companies in the US and Hong Kong.
Prof. Shulin Zhang — CSO
Inventor of the PPdPM platform. Based at Shanghai Jiao Tong University School of Medicine and Fudan University. PhD in Genetics (Sichuan University), NIH visiting scholar. 20+ national grants, 20+ patents, successful 2022 technology transfer.
Darry Huang — CIO
Leads all AI and machine learning. Responsible for algorithm development, model validation, clinical deployment, and full LIS/HIS integration across hospital systems.
Sun Zhanqiang — Regulatory & Commercial
Master of Biology, MBA, Senior Engineer. Founder of multiple biotech companies. Xiamen Double Hundred Talent. Former Clinical Base Director, Fudan University. HKD 200M+ annual sales track record.
Feng Jian
Business Development & Sales — 8+ years at Sigma Aldrich China, 6 patents, Sales Excellence Award recipient
Wang Liang
Senior R&D Researcher — PhD, SJTU School of Medicine. Leads multi-omics research and assay development
Adrian Watson — COO, Hong Kong
Oversees Hong Kong subsidiary operations, MDACS regulatory compliance, and business development with private clinics and insurers across the region.
Li Ning — COO, China
Leads China subsidiary daily operations, coordinates NMPA submissions, and oversees diagnostic service partnerships with hospitals and health systems.
Partner With BAIRI
BAIRI (百利研究院) — Biotech AI Research Institute is headquartered at BAIRI.at. We welcome clinical collaborators, healthcare system partners, diagnostic distributors, and strategic investors who share our vision of making early cancer detection accessible to everyone.
One blood test. AI-powered precision. Results in days, not months. BAIRI is transforming how the world detects lung cancer — starting with the 150 million patients who need answers today.