Partner Spotlight: 4DMedical

A Q&A with Chuck Hatt, Vice President, Medical & Clinical Affairs

​4DMedical's groundbreaking lung imaging technology is transforming pulmonary care.

Since 2013, this global medical technology company and long-standing OSIC member has been helping pave the way for more precise, personalized, and effective respiratory care through the use of advanced imaging and AI solutions. Its unique and non-invasive imaging technologies provide clinicians with unprecedented insights into pulmonary function, ultimately helping improve the diagnosis and treatment of lung disorders, including unexplained dyspnea, asthma, COPD, fibrosis and other restrictive airway diseases.

We spoke with Chuck Hatt, 4DMedical’s vice president of medical and clinical affairs, about the company’s commitment to advancing lung health, its cutting-edge work, and how the OSIC Cloud Data Repository has directly contributed to bringing its revolutionary solutions to market. Following is an excerpt from our conversation.
​

Tell us more about 4DMedical’s work. How is your company making a difference in the pulmonary space?

4DMedical is making significant strides in the realm of pulmonary diagnostics with our latest innovation, IQ-UIP™ software. This tool, which was granted Breakthrough Device designation and FDA clearance in 2024, is specifically designed to address the critical challenge of earlier diagnosis for IPF. IQ-UIP is an AI imaging solution trained to automatically assess chest CT scans for the presence of radiological Usual Interstitial Pneumonia (UIP). This radiological finding and key diagnostic criteria are often linked to idiopathic pulmonary fibrosis (IPF).

As a rare disease, IPF diagnosis has historically been a prolonged process for patients. It takes an average two years and involves consultations with multiple doctors, as well as several CT scans. The condition itself is often fatal, with a median survival of just 3–5 years post-diagnosis. Diagnostic delays have been a notable barrier to effective care.

The introduction of IQ-UIP represents a transformative advancement in this field. By leveraging deep learning technology, the software offers highly precise UIP detection (Figure 1) that can be deployed at scale across a health system. It automatically identifies patients for referral to specialty care. IQ-UIP can be part of a provider’s population health effort to ensure timely interventions, ultimately aiming to improve survival rates.

​​Relatedly, a recent published study showed that IQ-UIP similarly discriminated survival when compared with expert radiologist-determined UIP classification (Figure 2). IQ-UIP exemplifies how technology can bridge existing gaps in healthcare, offering hope for better management and outcomes for patients.

Figure 1. The ROC curve for IQ-UIP performance testing for FDA clearance. IQ-UIP achieved an ROC-AUC of 0.966 with 92% sensitivity and 90% specificity using an optimal confidence threshold for UIP classification. The test dataset (n=804 subjects) was multi-center including IPF, other ILDs, pneumonia, COPD, other non-fibrotic pulmonary conditions, and normal controls and was ground-truthed by a panel of 5 expert thoracic radiologists. 

---

Figure 2. Kaplan-Meier survival curves for an interstitial lung disease clinical cohort according to IQ-UIP radiomic UIP classification (A) and expert thoracic radiologist visual UIP classification (B). This study showed that IQ-UIP similarly discriminated survival when compared with radiologist-determined UIP. Chung JH, et al. A Deep Learning-Based Radiomic Classifier for Usual Interstitial Pneumonia. Chest. 2024 Feb;165(2):371-380. doi: 10.1016/j.chest.2023.10.012.

---

Why did you join OSIC?

OSIC unde​rtook a daunting task to compile the leading database of ILD imaging data in the world. The fact that many members leverage the data is a testament to OSIC’s success on that front. Beyond the data, OSIC has played a valuable role in creating a community across AI imaging companies, pharma, and academic members. This community shares knowledge and experience far more than we would day-to-day without the coordinating and mediating efforts of OSIC. We know that OSIC also has strong ties into patient advocacy groups. We look forward to working with OSIC to see how 4DMedical can leverage those connections to advance R&D studies, clinical pilots of new solutions, and daily care.

What have you been able to accomplish by using the OSIC Cloud Data Repository?

4DMedical has worked with the OSIC repository since joining in 2021. We’ve regularly used OSIC as a source of data for our internal development efforts. We’ve also been able to partially utilize OSIC data for regulatory activities. Advances in the OSIC repository over time have continued to make it easier to curate specific usable datasets. OSIC was one of the key data sources for the development of IQ-UIP along with 4DMedical’s in-house repository and other external sources.  We can definitively say that without OSIC data being readily available, IQ-UIP would have taken longer to bring to market. OSIC has directly contributed to accelerating the time it takes to bring solutions to market for the benefit of physicians and patients.

Are there any future projects you’d like to tease?

In addition to aiding population screening for earlier detection of fibrotic conditions, 4DMedical is also advancing quantification of interstitial lung abnormalities (ILAs). This aids diagnosis, staging, and tracking therapy response across a broad spectrum of ILDs. 4DMedical is in the later stages of development and regulatory clearance for our next-generation Lung Texture Analysis product. This product provides automated detection, visualization, and quantification of ILAs as well as other key imaging features. Our membership in OSIC has been critical, as we’ve utilized data from a variety of ILDs present in the OSIC database during the development and validation of LTA.

Outside of OSIC, we are also excited about our ongoing work with CT:VQ™ technology, which is currently in clinical studies. This innovative product enables existing CT equipment to provide detailed ventilation and perfusion data without the need for traditional nuclear VQ scans or contrast agents. By doing so, it aims to reduce costs and improve access to advanced diagnostics. It offers a faster and more efficient alternative to understanding a patient’s lung function in detail, which is often critical for treatment decisions and post-treatment follow-up in a wide range of pulmonary conditions.