for Early Detection of Bovine Mastitis Using AI-Powered Image Analysis and Automation
— Accelerating Implementation Through SBIR Phase 2 —
— Enhancing Productivity and Profitability in Sustainable Dairy Farming —
CarbGeM Inc. (Headquarters: Shibuya-ku, Tokyo; CEO: Masakazu Nakajima; hereinafter “CarbGeM”) and Hiroshima University have jointly initiated the development of an innovative diagnostic solution for bovine mastitis, the most common bacterial infection affecting dairy cows. This initiative aims to shorten the time required for mastitis diagnosis from over two days to approximately 30 minutes, while also standardizing interpretations that previously varied depending on the examiner's level of expertise.
This project, initially selected in FY2024 for Phase 1 (Proof of Concept) of the "SBIR Program for Startups" funded by the Japanese government, has been highly evaluated and will proceed to Phase 2 (Practical Development) in FY2025 with expanded research and development activities.
▶ Learn more about the SBIR Program : (リンク »)
▶ See Phase 1 subjects (リンク ») <Japanese website>
Background
Mastitis in dairy cattle significantly disrupts stable milk production through reduced yields and discarded milk. As bacterial infection is a primary cause, antimicrobial treatments such as antibiotics are commonly used. However, the rise of antimicrobial resistance (AMR) has heightened the need for more evidence-based use of antibiotics, even in livestock settings.
Hiroshima University has previously developed a technique for applying Gram staining to milk samples, enabling the identification of pathogens within approximately 30 minutes after test initiation (References 1 and 2). The collaboration in this project integrates the university’s technology with CarbGeM’s proprietary AI image analysis and laboratory automation technologies—originally developed for Software as a Medical Device (SaMD)—to achieve faster and more standardized mastitis diagnostics.
Role of CarbGeM in the Project
CarbGeM is responsible for the development of the following technologies to support early mastitis diagnosis:
・ Establishment of automated Gram staining protocols for bovine milk samples
・ Development of a CNN-based classification model for identifying bacterial morphology (e.g., cocci/rods) and Gram staining results (positive/negative)
・ Integration of AI analysis outputs into diagnostic decision support for veterinarians, including assistance with antimicrobial prescriptions
By shifting diagnostic procedures from reliance on technicians’ skills to a standardized, automated system, the project aims to promote further improvements in both the speed and accuracy of mastitis testing in dairy operations.
Contribution to Sustainable Dairy and AMR Mitigation
The technology is expected to reduce the misuse of antibiotics, curb antimicrobial resistance, and ultimately serve as a cornerstone for both improved productivity and environmental sustainability in dairy farming. The solution also supports the “One Health” approach by addressing antibiotic resistance in both animal and human healthcare systems.
Future Outlook
CarbGeM and Hiroshima University aim to establish simple, fast, and field-applicable technologies for early detection of mastitis pathogens. By realizing practical implementation of this diagnostic system, the project seeks to minimize economic losses and enhance profitability and sustainability in the dairy sector.
References
1： Rapid determination of pathogens in mastitic milk of dairy cows using Gram staining
Naoki Suzuki, Sohei Kaneko, Naoki Isobe
Journal of Veterinary Medical Science (2022)
DOI: 10.1292/jvms.21-0631
2： Determining causal pathogens and inflammatory state of mastitis in dairy cows via Gram staining of precipitates in milk
Naoki Suzuki, Naoki Isobe
Frontiers in Veterinary Science (2025)
DOI: 10.3389/fvets.2024.1492564
About CarbGeM Inc.
CarbGeM integrates biology with digital technologies, fostering innovation through medical-engineering collaboration and convergence of software and hardware. Its proprietary AI image analysis technologies are designed to support infection diagnostics and antimicrobial stewardship. Through open innovation with leading institutions, CarbGeM contributes to addressing global AMR challenges.
Headquarters: 5-13 Jinnan 1-chome, Shibuya-ku, Tokyo, Japan
CEO: Masakazu Nakajima
Website: (リンク »)
This project, initially selected in FY2024 for Phase 1 (Proof of Concept) of the "SBIR Program for Startups" funded by the Japanese government, has been highly evaluated and will proceed to Phase 2 (Practical Development) in FY2025 with expanded research and development activities.
▶ Learn more about the SBIR Program : (リンク »)
▶ See Phase 1 subjects (リンク ») <Japanese website>
Background
Mastitis in dairy cattle significantly disrupts stable milk production through reduced yields and discarded milk. As bacterial infection is a primary cause, antimicrobial treatments such as antibiotics are commonly used. However, the rise of antimicrobial resistance (AMR) has heightened the need for more evidence-based use of antibiotics, even in livestock settings.
Hiroshima University has previously developed a technique for applying Gram staining to milk samples, enabling the identification of pathogens within approximately 30 minutes after test initiation (References 1 and 2). The collaboration in this project integrates the university’s technology with CarbGeM’s proprietary AI image analysis and laboratory automation technologies—originally developed for Software as a Medical Device (SaMD)—to achieve faster and more standardized mastitis diagnostics.
Role of CarbGeM in the Project
CarbGeM is responsible for the development of the following technologies to support early mastitis diagnosis:
・ Establishment of automated Gram staining protocols for bovine milk samples
・ Development of a CNN-based classification model for identifying bacterial morphology (e.g., cocci/rods) and Gram staining results (positive/negative)
・ Integration of AI analysis outputs into diagnostic decision support for veterinarians, including assistance with antimicrobial prescriptions
By shifting diagnostic procedures from reliance on technicians’ skills to a standardized, automated system, the project aims to promote further improvements in both the speed and accuracy of mastitis testing in dairy operations.
Contribution to Sustainable Dairy and AMR Mitigation
The technology is expected to reduce the misuse of antibiotics, curb antimicrobial resistance, and ultimately serve as a cornerstone for both improved productivity and environmental sustainability in dairy farming. The solution also supports the “One Health” approach by addressing antibiotic resistance in both animal and human healthcare systems.
Future Outlook
CarbGeM and Hiroshima University aim to establish simple, fast, and field-applicable technologies for early detection of mastitis pathogens. By realizing practical implementation of this diagnostic system, the project seeks to minimize economic losses and enhance profitability and sustainability in the dairy sector.
References
1： Rapid determination of pathogens in mastitic milk of dairy cows using Gram staining
Naoki Suzuki, Sohei Kaneko, Naoki Isobe
Journal of Veterinary Medical Science (2022)
DOI: 10.1292/jvms.21-0631
2： Determining causal pathogens and inflammatory state of mastitis in dairy cows via Gram staining of precipitates in milk
Naoki Suzuki, Naoki Isobe
Frontiers in Veterinary Science (2025)
DOI: 10.3389/fvets.2024.1492564
About CarbGeM Inc.
CarbGeM integrates biology with digital technologies, fostering innovation through medical-engineering collaboration and convergence of software and hardware. Its proprietary AI image analysis technologies are designed to support infection diagnostics and antimicrobial stewardship. Through open innovation with leading institutions, CarbGeM contributes to addressing global AMR challenges.
Headquarters: 5-13 Jinnan 1-chome, Shibuya-ku, Tokyo, Japan
CEO: Masakazu Nakajima
Website: (リンク »)
本プレスリリースは発表元企業よりご投稿いただいた情報を掲載しております。
お問い合わせにつきましては発表元企業までお願いいたします。
お問い合わせにつきましては発表元企業までお願いいたします。
