EnvoyAI Partner:

Website: www.vuno.co
Year Founded: 2014
Area of Focus/Speciality: Radiology, Emergency Medicine, Regular Medical Checkup; Body region is not restricted but currently we focus on chest (for lung disease screening) and hand (for bone-age assessment).
Email: hello@vuno.co
Hospital Associations: Asan Medical Center(Seoul, South Korea), Severance Hospital(Seoul, South Korea), Korea University Hospital(Seoul, South Korea), Bundang Seoul National University Hospital(Seongnam, South Korea), Kangbuk Samsung Hospital(Seoul, South Korea), Ajou University Hospital(Suwon, South Korea), Sejong Hospital(Bucheon, South Korea), and Ewha Womans University Hospital(Seoul, South Korea).
Company Headquarters: Seoul, South Korea
Number of Algorithms: 2 machines - ILD quantification and PTX detection.


About VUNO Inc.

“Putting the world’s medical data to work”: VUNO innovates clinical workflow by solving medical problems using data-driven machine learning technology. Based on massive data collected from the hospitals, we develop software that provides intelligent assistance to streamline the decision making for clinicians.


Reference Sites:

  • VUNO-Med LungQuant: Asan Medical Center(Research purpose only)
  • VUNO-Med PTX Detective: N/A
VUNO Inc. Leadership
Yeha Lee CEO

Yeha Lee


Hyunjun Kim CSO

Hyunjun Kim


Kyu-Hwan Jung CTO

Kyu-Hwan Jung


VUNO-Med LungQuant : Chest CT | Lung | Disease Pattern Quantification

Vuno is directly responsible for product and clinical representations

  • screenshot_ILD.png

This algorithm automatically quantifies ILD(Interstitial lung disease) by classifying each pixel of lung parenchyma in chest CT scan into 6 pattern subtypes. The subtypes are normal, emphysema, ground-glass opacity, reticular opacity, honeycombing and consolidation. This algorithm visualizes the quantified lung CT scans with color map which enhance the understanding of disease diffusion.

  • Applicable study types: Lung CT scans(Both HRCT and thin section CT scans)
  • Indication for Use: Analytical research for ILD such as prognosis prediction, automatic diagnostics or contents-based image retrieval.
  • Training Data Source: Asan Medical Center(Seoul, South Korea), National Jewish Health Center(Denver, USA)
  • Number of Studies Trained On: 92 cases
  • Statisical Performance: Dice index with gold standard : 0.82
  • Type of Model Used: Generative Adversarial Network for Semantic Segmentation
  • Regulatory Status: Research use only

VUNO-Med PTX Detective : Chest Radiograph | Lung | Lesion Detection

Vuno is directly responsible for product and clinical representations

  • screenshot_PTX.png

This algorithm automatically detect pneumothorax in PA view chest radiograph and visualize the detected region on the original chest image as a heatmap. The algorithm also provides the confidence score of pneumothorax detection in percentage.

  • Applicable study types: Chest X-ray(PA View)
  • Indication for Use: Triage in emergency rooms.
  • Training Data Source: Asan Medical Center(Seoul, South Korea), Bundang Seoul National University Hospital(Seongnam, South Korea)
  • Number of Studies Trained On: 400 cases
  • Statisical Performance: Sensitivity 95.7% @ Specificity 97.0%
  • Type of Model Used: Weakly and Semi-supervised Convolutional Neural Network
  • Regulatory Status: Research use only