At Deep Breathe, we combine advanced artificial intelligence with clinical expertise to revolutionise lung ultrasound diagnostics.
Experience the transformative power of Deep Breathe, where cutting-edge AI and the convenience of portable lung ultrasound technology unite to revolutionize the diagnosis of respiratory diseases. Our clinician-driven and data-centric approach sets us apart in our pursuit of healthcare innovation. By leveraging our large datasets with expert clinical and engineering oversight, we intend to reshape the landscape of lung imaging and diagnosis.
Deep Breathe is at the forefront of revolutionizing lung health diagnostics with a team that uniquely blends software developers, engineers, clinicians, and ultrasound experts across Canada. Our diverse expertise, operating from London, Waterloo, and Vancouver, enables us to leverage cutting-edge AI and medical imaging technology, fostering a culture of innovation and cross-disciplinary collaboration.
Dr. Rob Arntfield, our visionary founder, is a world-renowned thought leader and expert in point-of-care ultrasound. With a distinguished background in critical care and emergency medicine, Rob’s passion for improving patient outcomes through technology has driven Deep Breathe’s mission to democratize lung health diagnostics. His leadership and innovative spirit continue to inspire our groundbreaking work.
Lung ultrasound videos for AI training in our proprietary dataset
Peer-reviewed publications as leaders in this space
Patents filed, protecting our innovations
The chest X-ray market size in the USA we are addressing
PneumoLens
Best in class lung sliding AI model currently undergoing regulatory validation.
This study examined the real-time diagnostic accuracy of artificial intelligence-assisted lung ultrasound (AI-LUS) in identifying absent lung sliding, crucial for pneumothorax diagnosis. Results indicated high sensitivity and moderate specificity for AI-LUS, suggesting its potential utility in clinical settings.
This study introduced a deep learning-based solution to automate the assessment of pneumothorax through lung ultrasound (LUS) by detecting lung sliding artefacts. Using data from two academic hospitals, researchers trained a binary classifier on M-mode (motion-mode) images extracted from B-mode (Brightness-mode) videos.
This multicenter study aims to automate the crucial differentiation between A-line (normal parenchyma) and B-line (abnormal parenchyma) patterns on lung ultrasound (LUS) using deep learning. By training a customized neural network on a large dataset of labeled LUS images, the researchers achieved impressive results.
Contact us for any inquiries, collaborations, or information. Your inquiries are valuable to us as we work.
Clean and standardize your ultrasound data
Customized SaaS AI diagnostic solutions
Disease-specific lung diagnostics (patent pending)
Deploy hardware-optimized solutions
Clean and standardize your ultrasound data
Customized SaaS AI diagnostic solutions
Disease-specific lung diagnostics (patent pending)
Deploy hardware-optimized solutions
