CarbGeM Inc.

Driving Innovation in Life Sciences with AI & Digital Transformation (DX)

CarbGeM Inc. is a pioneering biotechnology venture integrating AI, digital transformation (DX), and biology to address key challenges across the life sciences industry. Our mission is to standardize, automate, and enhance the efficiency of medical diagnostics and research through our cloud-based platform, CarbConnect®.

🔬 Our Platform: CarbConnect®

CarbConnect is a cloud-based AI analysis platform designed for hospitals, research institutions, and laboratories. It facilitates real-time image analysis, remote collaboration, and secure data management for hematology and microbiology.

🧠 AI Models & Applications

We develop specialized AI models for medical image analysis, available via the CarbConnect platform. Many of our tools are designated for Research Use Only (RUO).

🩸 Hematology

• BM Smear AI 1.0: A deep learning tool utilizing Faster R-CNN to automate the detection and classification of bone marrow cells (e.g., erythroblasts, neutrophils, blasts) from images captured by CellaVision DC-1.
  • Metrics: Achieves an F1 score of 0.755.
  • Collaboration: Developed with Kyoto University Hospital.

🧫 Microbiology

• BiTTE® lite / iE: AI-powered microbial analysis for classifying Gram-stained microorganisms into 7 categories (GNR, GNC, GPR, GPC, Yeast, etc.).
• Nugent Score AI 1.0: Automated evaluation of Nugent scores from Gram-stained vaginal smears to assist in bacterial vaginosis research.
  • Performance: Deep learning models demonstrated up to 94% accuracy in predicting Nugent scores.
• ZOI Series (Zone of Inhibition):
  • ZOI 1.0: Automates the measurement of inhibition zones on Petri dishes.
  • ZOI Pro: Advanced AI for high-throughput labs with CLSI M100-Ed34 SIR interpretation.
  • ZOI VET: Veterinary-specific solution complying with CLSI VET01S ED7 standards for animal health.
• Colony Counter: Automated counting of bacterial colonies on agar plates for QC and food safety.

📚 Research & Whitepapers

We actively publish research on the application of AI in resource-limited settings and diagnostic workflows:

• Revolutionizing Antimicrobial Susceptibility Testing: How AI and Automation Improve Zone of Inhibition Analysis
• Generative AI & LLMs: Transforming Microbial Image Analysis
• AI-Powered Gram Stain Image Analysis for Microbial Classification

🤝 Partnerships

Our solutions are developed in collaboration with leading institutions, including:

• Kobe University
• National Center for Global Health and Medicine (NCGM)
• Kyoto University Hospital
• Kameda Medical Center

Disclaimer: Some AI applications described are for Research Use Only (RUO) and not intended for clinical diagnostics.