Last Updated:
12th January 2025
Determining the activity of pulmonary tuberculosis (TB) on chest radiographs is difficult. Medical IP's DeepCatch X TB deep learning software can help determine TB activity and differentiate active TB from healed TB on chest radiographs comparable to expert clinicians. The model-driven activity score reflects bacilli burden and treatment response. MEDICAL IP also has other solutions available that allow for the detection of COVID-19 or pneumonia.
Certification
Korean MFDS Approved (overseas market only)
Development Stage
On the Market
Deployment
Online & Offline
Intended Age Group
15+ years
Target Setting
Primary health centres, general hospital (above primary level), teleradiology companies, government/public sector, e.g. national TB programme, private sector, high-TB burden countries
Current Market
South Korea, South East Asia
Input
Can be used to read images from any chest X-ray machine and model
Chest X-ray image format: DICOM
Chest X-ray type: Posterior-anterior chest X-ray
Output
Output includes:
Heatmap
Probability score for identifying active TB from healed TB
Score trajectory during anti-TB treatment
Lung abnormalities included in the TB Score: Abscess, Air fluid level, Blunted costophrenic angle, Cavity, Chest wall invasion / destruction, Consolidation, Interstitial markings, Lymphadenopathy, Mass, Nodule, Opacity, Pleural effusion, Prominence in hilar region
Hardware
CPU: Intel Core i5 or higher
RAM: 8 GB or higher
GPU: NVIDIA GeForce 1000 Series (3 GB) or higher
HDD: 50 GB memory or higher
Offline: memory supporting Compute Unified Device Architecture (CUDA)
Hardware Requirements: TiSepX Lung / TiSepX COVID / TiSepX TB -> 8GB GPU
Server
N/A
Integration with X-ray Systems
Integration with PACS and Legacy Systems
DeepCatch X TB can be integrated with general X-ray systems. It can be transmitted via PACS.
It is possible to integrate the product with the client’s legacy picture archiving and communication system (PACS) for on-premise type only (e.g., PACS, HIS, and OCS, etc).
Software
Windows OS (7 or higher), NVIDIA CUDA.
Processing Time
Less than 1 minute per X-ray
Data Sharing & Privacy
DeepCatch X TB is operated only for on-premise, so, de-identification of the data is not required.
Software Updates
As soon as the module is developed, an update is made after the test. Regular updates happen at least once a year, but irregular minor updates might happen if abnormality function is discovered.
Please contact company for information: tech@medicalip.com
Please contact company for information: tech@medicalip.com
Product Development Method
Generative AI
Training
9,836 adult chest X-rays
Reference Standard
Culture and smear
Publications
1. Qin ZZ, Van der Walt M, Moyo S, et al., Computer-aided detection of tuberculosis from chest radiographs in a tuberculosis prevalence survey in South Africa: external validation and modelled impacts of commercially available artificial intelligence software [published correction appears in Lancet Digit Health. 2024 Sep;6(9):e604. doi: 10.1016/S2589-7500(24)00176-6]. Lancet Digit Health. 2024;6(9):e605-e613. doi:10.1016/S2589-7500(24)00118-3
2. Kim HJ, Kwak N, Yoon SH, et al., Artificial intelligence-based radiographic extent analysis to predict tuberculosis treatment outcomes: a multicenter cohort study. Sci Rep. 2024;14(1):13162. Published 2024 Jun 7. doi:10.1038/s41598-024-63885-0
3. Hwang EJ, Jeong WG, David PM, Arentz M, Ruhwald M, Yoon SH., AI for Detection of Tuberculosis: Implications for Global Health. Radiol Artif Intell. 2024;6(2):e230327. doi:10.1148/ryai.230327
4. Lee S, Yim JJ, Kwak N, et al., Deep Learning to Determine the Activity of Pulmonary Tuberculosis on Chest Radiographs. Radiology. 2021;301(2):435-442. doi:10.1148/radiol.2021210063
5. Kim JY, Lee S, Park H, et al., Post-treatment Radiographic Severity and Mortality in Mycobacterium avium Complex Pulmonary Disease. Ann Am Thorac Soc. 2024;21(2):235-242. doi:10.1513/AnnalsATS.202305-407OC
This website works best with browsers other than Internet Explorer.