AI development support through machine learning using DICOM data
We support AI development using DICOM images such as intravascular imaging and CT scans.
We are available to answer your questions at any time.
It is also optimized for LCD tablets, allowing you to create annotation data efficiently.
The annotation data can be used for machine learning using general-purpose engines (Detectron, YOLO, etc.).
You can also request us to create a machine learning engine for 3D information data (CT, MRI).
To the Clinical Research CoreLab Team
Provides tools for manual measurement of all DICOM images, including IVUS, OCT, CT, and US.
Measurement data is stored in a database.
Since it is managed locally, security is maintained within the research team.
Core lab analysis types that require specific measurements with large amounts of data can be automated using AI.
・Example : CT measurements before TAVI (annulus area measurement, calcification volume within each coronary cusp, coronary height, etc.)
・Example: Automatic extraction, volume, and thickness measurement of plate-like calcifications in OCT.
Automatically determine the cause of acute coronary syndrome (Calcified Nodule, Plaque Rapture, Plaque Elosion)
(Joint research with Dr. Takuya Mizukami and Dr. Akikazu Yamamoto, Department of Clinical Pharmacology, Showa University School of Medicine)
(昭和医科大学, 臨床薬理学教室 水上拓也先生, 山本明和先生と共同研究)
For medical device developers (companies, research teams)
We will contribute to the creation of 3D data to verify and investigate the compatibility and operation of medical devices currently under development with living bodies.
Create 3D data (obj, stl) from CT data, etc. (Can be created manually or fully automated).
It is possible to create models using XR devices and 3D printing to verify the equipment under development.
・Example: Creating a vascular model from 3D data of the iliac artery and its calcification, and verifying catheter passability and stent tracking. (*We do not undertake 3D printing.)
・We are also happy to discuss the sale of only the STL data for the required parts.
