#Japan #Tachikawa #SEGMAT #AI Segmentation #BioNet Lab

Revolutionizing Image Analysis: The New SEGMAT

The world of image analysis is about to be transformed with the significant overhaul of SEGMAT, an advanced image analysis support system developed by BioNet Lab Inc. This major update will be officially available starting June 6, 2025. By implementing a three-module structure comprising SEGMAT Insight, SEGMAT Engine, and SEGMAT Viewer, users can experience a seamless workflow from annotation to learning, inference, and visualization.

Understanding the New System

Module Breakdown

• - SEGMAT Insight: This module is designed for annotation assistance featuring intuitive GUI tools such as 2D/3D brushes, erasers, fill tools, as well as Smart Region Extraction and Edge-aware painting techniques. Users can easily execute precise edits by combining region selection with layer functionalities.

• - SEGMAT Engine: The core of the system, this module simplifies the training and inference processes. It facilitates the training of high-accuracy models even with a limited number of labeled images, thus accommodating various use cases through flexible parameter settings.

• - SEGMAT Viewer: This module enhances result visibility. It supports stacked and split comparative displays, allowing users to save slices and copy them to the clipboard easily.

Advanced Features

The newly updated SEGMAT system boasts several additional features that significantly enhance its capabilities:

• - Machine Learning Support: Utilizing image features, this system can automatically generate annotations via machine learning, streamlining the entire analysis process.
• - Quantitative Measurement: Users can measure lines and areas, and calculate area/volume based on labels, thereby ensuring accurate assessments driven by physical scale settings.
• - Particle and Fiber Analysis: The enhanced functionality caters to detailed analyses such as calculating the area, diameter, circularity, aspect ratio, and elongation of particles, in addition to offering automated generation of fiber orientation maps and histograms for textile material evaluations.

User-Centric Design

The integrated system supports high-precision AI segmentation even with minimal annotations. Features like project management functionalities contribute to reproducibility and operational efficiency, saving time and improving productivity. The user interface can be toggled between Japanese and English, making it accessible for a wider audience.

Upcoming Presentation

BioNet Lab Inc. will showcase the SEGMAT at the 81st Annual Meeting of the Japanese Society of Microscopy, taking place from June 9 to 11, 2025, at the Fukuoka International Congress Center in Fukuoka City.

Ideal Use Cases

SEGMAT is perfectly equipped for 3D segmentation and structural extraction of images obtained from FIB-SEM and optical microscopes. Specific applications include:

• - Particle Image Analysis: Analyzing particle size and shape based on electron microscope images.
• - Fiber Material Evaluation: Visualizing fiber orientation trends and performing directional analysis using structural tensors.
• - Quantifying Segmentation Results: Assessing area, volume, and orientation metrics based on segmentation results.

Explore SEGMAT Further

For those considering the adoption of SEGMAT, BioNet offers feasibility assessments by evaluating the applicability of this advanced system on provided image data. Prospective users are encouraged to reach out for preliminary evaluations.

System Requirements

The recommended operating environment includes:

• - OS: Windows 11 (64-bit)
• - CPU: Intel® Core™ i7 11th gen or higher, clock speed of 2.1GHz or more
• - GPU: NVIDIA® GeForce RTX 3060 with at least 12GB VRAM
• - Memory: Minimum 32GB (16GB minimum)
• - Storage: SSD of at least 1TB (minimum 20GB free space)
• - Monitor: Full HD (1920 × 1080) resolution recommended

The new SEGMAT system stands poised to elevate the standard for image analysis across research and manufacturing sectors. For further information on pricing or system details, visit the official product page or reach out directly to the BioNet Lab team.