Last Updated:
25th January 2025
AXIR-CX is an AI-powered, CE Class IIa chest X-ray diagnostic screening software that detects approximately 10 lung abnormalities, aiding in the diagnosis of tuberculosis and pneumonia in adults. It highlights abnormal regions on chest X-rays and generates reports with confidence scores for each finding.
Independently evaluated and widely used globally in healthcare facilities and community-based TB screening programs, AXIR-CX seamlessly integrates with various PACS systems and interprets images from different X-ray machines. Its user-friendly design ensures compatibility with existing workflows, providing an efficient and adaptable solution for chest X-ray analysis.
Certification
CE Certified Class IIa
Development Stage
On the Market
Deployment
Shipped with RadiSen's AI-box, an edge computer with the AI installed in it.
Intended Age Group
16+ years
Target Setting
AXIR-CX can be used in primary screening sites, in public or private healthcare facilities as it detects lung abnormalities in addition to reporting TB and pneumonia.
Current Market
AXIR-CX is being used in many haspitals in South East Asia, India, and Africa.
Input
Can be used to read images from any chest X-ray machine and model
Chest X-ray image format: DICOM, JPG
Chest X-ray type: Posterior-anterior chest X-ray, Anterior-posterior chest X-ray
Output
Output includes:
Heatmap
Dichotomous output indicating whether each abnormality is likely present or absent
Probability score for TB
Probability score for each abnormality
Location of each abnormality
Based on the internal test dataset, AXIR-CX has a threshold of 0.3 for TB. Different thresholds are applied in several sites where AXIR-CX is deployed based on the sensitvity level required by sites.
Results are presented in a structured report.
Lung abnormalities included in the TB Score: Loculated pleural effusion, Nodule, Opacity, Pleural effusion
Additional findings reported by the product: Atelectasis, Fibrosis, Loculated pleural effusion, Mass, Nodule, Opacity, Pleural effusion, Prominence in hilar region, Pneumothorax
Hardware
Radisen's proprietary AI-box will be provided with the artificial intelligence software. AXIR-CX can be deployed in two ways: 1) Work in between the x-ray console pc and PACS 2) Independently with its own AXIR-CX viewer.
Server
Online & Offline (with Radisen's AI box)
Integration with X-ray Systems
Integration with PACS and Legacy Systems
AXIR-CX can be integrated with third party X-ray systems and PACS systems through SCP and SCU to facilitate the exchange of dicom images from the console viewer and to the PACS
AXIR-CX can be integrated with third party X-ray systems and PACS systems through SCP and SCU to facilitate the exchange of dicom images from the console viewer and to the PACS
Software
Image viewer provided by Radisen or integration with any third-party console viewer.
Processing Time
It generates a report in 10 seconds on average
Data Sharing & Privacy
As AXIR-CX can be used offline, there is no mandatory data sharing requirement with Radisen. There is an option to de-identify data.
Software Updates
We update the product regularly and updates are smoothly installed in discussion with the clients and our support team. Annual license subscription covers the updates.
Price
Installation and training are given free of charge. We provide license based pricing for unlimited image readings. The license subscription also covers maintenance. We provide an initial two years licnese including the AI-box with in less than 3,000 US dollars. For mass deployments and distributions, please contact us through contact@radisentech.com .
Product Development Method
Semi-supervised deep learning (CNN)
Training
Over 300,000 quality x-ray images and expert annotations are used for training.
Reference Standard
We used a combination of ground truths while developing the AI.
For Tuberculosis: we used GeneXpert test results as golden standard,
For Nodule/Mass: used x-rays of lung cancer patients confirmed by biopsy.
For other abnormalities: from manual annotations of multiple senior radiologists.
Publications
Evaluations: Evaluation report of RCRT (https://www.rcrt.or.th/ai_test_report/)
Publications: Please find a list of peer reviewed publications here https://www.radisentech.com/en/publications/
This website works best with browsers other than Internet Explorer.