[New Product] AI Provides Powerful Support for Visual Interpretation of Gram Stain Images

To Members of the Press

March 2026

KD Icons Co., Ltd.

[New Product] AI Provides Powerful Support for Visual Interpretation of Gram Stain Images

"Gram Stain Image AI Analysis System" to be released in April 2026

~ Contributing to resolving the shortage of experts and supporting the education of next-generation laboratory technicians ~

KD Icons Co., Ltd. (Headquarters: 4-6-15-304 Omori-minami, Ota-ku, Tokyo) will launch the "Gram Stain Image AI Analysis System 'ICONS21 AI Gram'," which supports the visual interpretation of Gram stain images in infectious disease microbiology testing, in April 2026. This system can be introduced as a subsystem of a hospital's bacteriology system or as a standalone system.

Basic operations are as follows:

1) Image loading

↓

2) Material (target bacteria) selection

↓

3) Execution

↓

4) Result display

■ Background of Development: Serious Shortage of Interpreters and Educational Burden

Currently, in the field of infectious disease diagnosis, the chronic shortage of "skilled interpreters" responsible for the visual judgment of Gram stain images is a serious issue. Additionally, the increasing burden of training young medical professionals and clinical laboratory technicians has become apparent, creating a need to balance accuracy with efficiency.

Against this backdrop, our company has developed this system to support medical professionals' judgments, utilizing technology and patents (already acquired) accumulated through grant projects from the Tokyo Metropolitan Small and Medium Enterprise Support Center, following joint research with experts.

■ Features of the System

This system presents AI analysis results online not as "diagnostic results," but as "reference information" to support the final judgment of medical professionals, allowing for flexible operation according to each facility's practices and educational policies.

1. Advanced AI Analysis Function: AI analyzes bacterial morphology such as general bacteria, fungi, mycobacteria, leukocyte phagocytosis, and Geckler classification.

2. Presentation of "Reference Information" to Support Medical Judgment: Rather than replacing diagnosis, the AI contributes to quality improvement as a tool to support education.

3. High Compatibility with Existing Systems: Seamless online reporting is possible by linking with our company's bacteriology testing system.

4. Flexible Customization for Each Facility: Comments and display content accompanying the judgment results can be revised according to facility needs.

■ Strong Joint Research Structure

The development of the first version of this system was guided by leading experts in infectious diseases and information engineering:

・Dr. Mitsuo Kaku (Professor, St. Marianna University School of Medicine / Professor Emeritus, Tohoku University)

・Dr. Koichi Hirata (Professor, Kyushu Institute of Technology)

・Dr. Kazuhiro Tateda (Professor, Toho University)

・Ms. Kimiko Matsuoka (Infection Control Certified Clinical Microbiologist, ICMT)

■ Future Outlook

Through this system, our company will contribute to improving the quality of infectious disease microbiology testing and fostering the next generation of laboratory technicians. We will continue to incorporate feedback from the field...