Radiologist viewing image

Big Data in Radiology: Is The Future of Imaging a Number?

Reading Time: 3 minutes read

The application of algorithms will advance preventative healthcare.

Radiologists, stop and savor the images you are reading. They might disappear from view.

In the not-so-distant future, radiologists might analyze ‘numbers’ rather than images. This radical change has the potential to change not only the role of radiologists, but also to advance the goal of preventative care. Here’s how big data in radiology might change the future.

Radiologist viewing image

Today, physicians order imaging exams to detect the presence of a specific and usually singular cause or disease. The pixel data that is captured by the imaging modality is assembled (or reconstructed / displayed) in an image that is meaningful to the human brain. Radiologists are educated to recognize, understand, and analyze the shapes, shades, and colors within that image in order to render a diagnosis.

Algorithms take first pass at analysis

Advancements in imaging analytics are making algorithms increasingly capable of doing these interpretations currently done by radiologists. Algorithms can analyze the pixel and other bits and bytes of data contained within the image to detect the distinct patterns associated with a pathology. The outcome of the algorithmic analysis is a metric. In the current early stage of imaging analytics, these metrics complement the analysis of the images made by radiologists, and help them render a more accurate or faster diagnosis.

For example, it is possible today to calculate bone density by applying an algorithm on any CT image of a bone. The resulting number is then compared with a threshold metric to determine whether the patient is at risk of fracture. If the number is below the threshold, a doctor can prescribe a regular intake of calcium or other preventative measure. The screening for ‘low bone density’ is made automatically without a dedicated and additional exam. It is determined simply by leveraging an existing CT examination performed on a patient. This is an important first step into preventative care.

The development of these automated analysis tools is already under way. Research teams and start-up companies across the world work every day to produce new algorithms to cover more body parts and pathologies. It won’t take long before radiologists are equipped with thousands of predictive algorithms to automatically detect the patterns of the most standard diseases. This application of advanced data analysis holds the exciting prospect of preventing diseases.

Big data in radiology gets smarter through image analysis

Evolving to this level of preventative care requires not only expertise in imaging analytics to develop the algorithms, but also access to huge libraries of images to refine the algorithms. Like other artificial intelligence (AI) tools, the algorithms will evolve, becoming more knowledgeable and more accurate as they analyze more cases.

Carestream can play a key role in this evolution. We manage hundreds of billions of images in our medical imaging repositories in the cloud and at national and regional healthcare data centers; and the volume grows daily. These images can be raw material for clinical leaders to develop, test, and validate new algorithms.

In compliance with local regulations and Personal Health Information privacy laws (all data are anonymized), we are developing collaborations with large regional healthcare providers to understand how big data tools can be applied to the imaging examinations stored in our data centers. Potentially, we can find applications that could open new doors to population health management.

We also are exploring the application of algorithms as an ‘on-demand’ cloud service. In this model, radiologists could get the help of a personal, well- trained algorithm to help with their routine work, resulting in potentially higher accuracy of diagnosis and faster results.

Looking further ahead, it is possible that radiologists will no longer look at images at all. Instead they might analyze only the outcomes of the algorithms. Algorithms could replace the  advanced clinical applications in their PACS that reconstruct data into images and display them. Imagine the impact this evolution would have on the role of radiologists as well as their education and required skill sets.

This emerging science has real implications for advancing preventative care . This is turn has the potential to lower overall costs for private and government healthcare payers, and advance population health management. As a company, Carestream is excited about leveraging our vast archives of data and our expertise in imaging analysis for the benefit of our customers and their patients. The healthcare industry is much closer today to advancing the goal of preventative care. #HealthIT #radiology

Want to learn more about the potential role of big data in radiology? Read the blog by Dr. Eliot Siegel.

Patrick Koch


—Patrick Koch is Carestream’s General Manager of Medical Imaging Solutions for Western Europe


This site uses Akismet to reduce spam. Learn how your comment data is processed.