The RadioGo^® system featured on this website is currently in the advanced R&D phase and has not yet received medical device clearance/approval. The content on this page is intended exclusively for scientific research exchange and collaboration discussions regarding intelligent medical imaging algorithms. The system does not possess functions for clinical diagnosis, treatment recommendations, or radiation dosage prescription.

RadioGo^®

AI-AVM System – AI-Assisted AVMs Targeting System for Radiosurgery
Advanced R&D Phase, Not Commercially Available

Powered by localized medical imaging datasets, we develop AI models and system integrations that enable partner hospitals to adopt advanced smart healthcare applications.
Our current focus encompasses three core AI imaging research frameworks:

Clinical Brain Imaging Decision Support System

Brain Imaging Anomaly Detection System

Automated Endoscopic Surgery Detection System

With AI assistance, medical staff workloads can be significantly reduced,
enhancing overall healthcare quality.

With AI assistance, medical staff workloads can be significantly reduced, enhancing overall healthcare quality.

AI-MRI

•Product introduction

AI-MRI is a brain MRI intelligent reading system that combines edge/cloud computing technology and medical image AI analysis, providing structural and multi-stage auxiliary diagnosis.

AI – MRI

• Main characteristics

Supports DICOM imaging format, compatible with both Web and App platforms
Automated classification and reconstruction of multiple MRI sequences (T1w, T2w, DWI, etc.)

Three-stage imaging analysis with abnormality triage mechanisms
Co-designed and validated in collaboration with clinical physicians

• Core Modules

AI-Atlas
Brain Functional Region Segmentation Module

The AI-Atlas module has the ability to segment functional areas of the brain in a structured brain, and the functional areas of the brain can be marked through standard anatomical atlases (such as DKT Atlas). The visualization results help clinicians analyze the spatial relationship between the lesion and key functional areas (such as language and motor cortical areas) before the operation.

Abnormality Triage System
Emergency Case Prioritization Module

After AI completes the image analysis, this module can actively mark the high-risk cases that need to be prioritized according to the scale of the foci, the degree of abnormality of image characteristics and the positioning results. Sorting is based on the integrated multi-modal image output (such as foci size, position, symmetry and functional area intersection and other indicators) to assist doctors to quickly focus on clinical urgent needs.

Lorem ipsum dolor sit

	AI-Atlas	Abnormality Triage System
Function	Brain region segmentation	Emergency case prioritization
Input	Standard anatomical atlas	AI imaging analysis results
Output	Visualized functional brain regions	Risk distribution & priority ranking
Advantage	Supports surgical planning & risk assessment	Quickly highlights high-risk key cases

AI-AVM

AI – AVM

• Product introduction

Supporting physicians in the precise planning of brain radiotherapy
Improving Clinical Decision-Making Efficiency and Accuracy

AI-AVM is a cloud-based intelligent medical software designed specifically for patients with arteriovenous malformations (AVM).

By integrating T2-weighted MRI and TOF (Time-of-Flight) MRA, the system automatically performs lesion contouring, tissue segmentation, and preliminary surgical risk stratification.

These imaging analysis capabilities enable radiologists, neurosurgeons, and oncology teams to quickly understand lesion characteristics and spatial relationships during radiotherapy planning.

Powered by deep learning, AI-AVM precisely segments the nidus, normal brain tissues, and cerebrospinal fluid (CSF), and computes quantitative metrics such as volume and tissue proportions—providing an objective foundation for treatment planning.

The system also supports Spetzler-Martin grading, generating AVM size-related scores based on imaging features, while allowing clinicians to supplement location and venous drainage information to complete the grading process.

All analytical results are fully visualized and can be exported as DICOM-RT structure sets (RTSS) for seamless integration with downstream planning workflows and multidisciplinary team communication.

AI-AVM serves as a clinical decision-support tool, and all analytical outputs require final confirmation and necessary adjustments by qualified medical professionals.

Although the system is not intended for diagnostic use, it significantly enhances the efficiency and safety of radiotherapy planning, helping clinicians anticipate treatment risks and potential toxicities.

AI-AVM is a powerful enabler on the path toward precision medicine and intelligent medical imaging analysis.

Product Features

Image Segmentation & Labeling	Clinical Analysis Metrics	Spetzler–Martin Grading
Automatic annotation of ROI, Nidus, Brain, and CSF	Provides tissue volume (CC) and proportional (%) measurements	Assist in judging the default score of AVM size disease
Preserves original imaging data without alteration, ensuring integrity	Calculates total AVM volume and maximum diameter	Allows clinicians to supplement functional cortex location and venous drainage information
Color-enhanced segmentation for improved visual clarity	Generates DICOM-RT structures for treatment planning reference	Provide a risk assessment process to assist surgical decision-making

• Indications for Use

01

AI-AVM is intended for the automatic labeling, visualization, and volumetric quantification of segmentable brain structures and lesions from a set of MRI images.
The system is used by trained medical professionals and serves as an assistive tool for structural MRI assessment.
All outputs require final review and any necessary adjustments by qualified clinicians.
Clinical decision-making and patient management should not rely solely on the results provided by this system.

02

This system is designed specifically for patients with unruptured cerebral arteriovenous malformations (AVM).
It is suitable for patients who have previously undergone Gamma Knife radiosurgery as well as those who are scheduled to receive radiotherapy.

03

By importing T2-weighted MRI and TOF images, the system automatically performs image alignment and overlay processing, reducing the need for clinicians to manually switch between different imaging sequences.
The AI model automatically labels suspected lesion locations and clearly differentiates tissue types:
Red: Nidus (lesion vessels)
Green: Normal brain tissue
Blue: Cerebrospinal fluid (CSF)

04

The system also provides lesion volume, distribution ratios, recommended radiation dose references, and maximum lesion diameter, serving as important parameters for surgical planning and risk assessment.
Clinicians can use the interactive image viewer—supporting rotation and zoom—to quickly inspect ROI regions.
All analysis results can be exported as Excel reports or image snapshots for clinical communication or medical record documentation.

Notice

AI-AVM is an assistive medical imaging analysis system and does not provide diagnostic results.
All analytical outputs are intended for reference by qualified medical professionals only.
Lesion segmentation results generated by the system must be reviewed and interpreted by physicians with anatomical expertise, such as radiologists, neurosurgeons, or neuroradiologists.
The system is designed exclusively for use with T2-weighted MRI and TOF MR angiography in the evaluation of cerebral arteriovenous malformations (AVM).
It is not intended for use in the following circumstances:
Lesions with a maximum diameter smaller than 2 mm
Poor imaging quality
Presence of hemorrhage
Lesions located in or adjacent to the leptomeningeal (pial) region
Before use, medical professionals must ensure that the imaging data meets the system’s analytical requirements and specifications.

RadioGo®