Project OPUS seeks to uncover X-ray markers to aid in IPF diagnosis

Use of AI may help to find disease signs evident in imaging

Joana Vindeirinho,PhD avatar

by Joana Vindeirinho,PhD |

Share this article:

Share article via email
A person holds an X-ray image at chest level in this illustration.

Carestream Health and the Open Source Imaging Consortium (OSIC) are collaborating in the OSIC Pulmonary Understanding Study, dubbed Project OPUS — a pilot study to uncover X-ray markers for the early diagnosis of idiopathic pulmonary fibrosis (IPF).

Project OPUS will make use of artificial intelligence, or AI, to investigate relationships between clinical indicators and imaging analyses, and IPF risk and prognosis.

The aim is to discover early disease signs that are evident in X-ray analyses, so that this more available imaging method can be used effectively for the early diagnosis of IPF patients. Carestream, a company specialized in medical imaging, including X-ray systems, is participating in and funding the project.

“We believe there is a strong possibility that X-ray, when coupled with artificial intelligence, can enable earlier diagnosis and eventually disease management of IPF,” Luca Bogoni, PhD, head of advanced research and innovation at Carestream, said in a company press release.

“This could offer opportunities for the development of new clinical solutions leveraging X-ray, a widely available and more affordable imaging modality than CT,” Bogoni added.

Recommended Reading
Umbilical cord stem cells

Radiotracer May Allow for PF Diagnosis, Monitoring via PET Scans

Searching for X-ray markers with the help of AI

The idea behind the use of AI in clinical settings is to leverage the ability of computers to screen large amounts of information and, through the use of programmed-in mathematical algorithms, come to conclusions about that date. Once a computer learns how to make such determinations, it would then be able to detect patterns and associations between pieces of information that humans could not, due to the sheer scale of data volume.

Harnessing the power of AI to improve diagnostics and disease monitoring is a hot topic of research in many diseases, and OSIC has been involved in several initiatives focused on fibrotic lung diseases, including IPF.

IPF, a progressive disease of unknown origin characterized by scarring (fibrosis) in the lungs, to date has no cure. However, there are treatments, such as anti-fibrotic medications, that can significantly slow its progression.

Given the availability of such treatments, an early and accurate diagnosis is crucial for a better long-term prognosis.

X-rays are a common method for diagnosing IPF and other pulmonary diseases, in combination with other tests, such as blood work and lung function tests. However, it currently might not be enough for a correct diagnosis — at least not in the early stages of IPF — which can lead to misdiagnosis and resulting hardships for patients.

High-resolution CT scans, known as HRCT, are more powerful imaging tools, both for diagnosis and for monitoring disease progression. However, they are more expensive and thus, less widely available than X-rays.

“Today the standard of care is HRCTs, yet two-thirds of the world do not have access to this modality,” said Elizabeth Estes, OSIC’s executive director.

As a first step in advancing X-rays as a potential diagnostic tool, Project OPUS will provide AI experts and other researchers with data from IPF patients. Among that data will be electronic medical records, X-rays, HRCT scans, and weekly forced vital capacity readings.  Forced vital capacity is a lung function parameter that measures the total amount of air a person can exhale after a deep breath. These data will be made available through the OSIC Data Repository.

Researchers and experts then will study and design algorithms that connect imaging measures and clinical indicators with disease risk and prognosis factors to assess patterns in the data. The goal is to uncover new disease signs in X-rays that are now unnoticed, which can become potential biomarkers for the early diagnosis of IPF.

We believe there is a strong possibility that X-ray, when coupled with artificial intelligence, can enable earlier diagnosis and eventually disease management of IPF.

OSIC aims to recruit 100-200 participants worldwide initially. Ultimately, the company hopes to expand the project to include up to 2,000 patients. All participants will have allowed access to their imaging and clinical data leading up to their IPF diagnosis.

“If this pilot [study] uncovers markers that are currently undetected in X-rays, it could lead to earlier detection and treatment of patients with IPF,” Estes said, adding, “The implications for improving patient outcomes could be significant.”

Beyond Carestream and OSIC, Project OPUS is being supported by a collaboration from the American Lung Association, the European Pulmonary Fibrosis Federation, Action for Pulmonary Fibrosis, and patientMpower, a company specialized in remote monitoring devices, which will provide the lung function readings collected by the project.

Your PF Community

Woman laying down reading

Visit the Pulmonary Fibrosis News forums to connect with others in the PF community.

View Forums