DICOM to Cloud

The Rapid Onset of Cloud-Native PACS

The incumbent PACS model

PACS technology has been an integral part of radiology since the early 1980’s and has provided a way to hold and view large volumes of medical imaging data as radiology has transitioned away from analogue image acquisition. It has thus enabled the growth of diagnostic imaging to serve so many more patients.

By breaking down the physical and time barriers associated with traditional film-based image retrieval, distribution, and display, it has significantly improved the ability to display, manipulate and share images with the interpreting doctor and the referring doctor. As a result, it has improved the efficiency and quality of care for patients and at the same time improve the working life and productivity of the highly valued medical specialist interpreting the images. 

Early PACS development was expensive and mostly financed by the federal government including the National Institutes of Health (NIH) and the U.S. Army Medical Research and Materiel Command 

Two decades ago, it appeared that it was only for tertiary hospitals and research centres and it seemed that it was designed only for a setting where the PACS and the modalities it served were located within the same facility and operated on a Local Area Network (LAN) 

I recall the first PACS install I was aware of was a children’s hospital in Sydney and this was approximately 20 years ago. Its direct upfront investment costs alone in today’s dollars would be $12-15M before including the substantial indirect operating costs 

At the time, I was running a diversified radiology group in Australia that operated across most states and had hundreds of locations. I often thought about how would we ever be able to afford to purchase and operate such a system across Wide Area Networks (WAN). It was difficult to even contemplate due to the large upfront and operating expenditure and even more problematic was our distributed clinics in varied settings from large public and private hospitals to large and small community clinics. 

Fast forward 20 years and it is now commonplace in the private and public settings as well as in hospitals and the community. In fact, it is a critical component of the infrastructure for any medical setting that is utilising medical imaging modalities and very few organisations can contemplate operating their services without it. 

The advent of improvements in the bandwidth to shift the data across a WAN, as well as dramatic improvements in hardware capacity, both in storage and in processing capacity and the cost of these aspects declining over time has enabled a system that was once developed with large single institutions in mind, to be adapted to address a distributed PACS network such as what exists today in many countries. 

These advances, however, have not been able to fully address many challenges that remain with the incumbent on-premise hardware-dependent PACS operating model.  

These challenges include: 

  • coping with bandwidth that still struggles to cope with the data size of DICOM studies that has been further exacerbated by the increase in the images per study and the resolution which combined, drives up each study’s data size 
  • hardware cost and maintenance which remains expensive and often involves an upfront outlay of capital as well as ongoing maintenance and periodical hardware refreshes 
  • expensive software spend upfront as well as either annual licence fees or a high cost per study click rate or even both 
  • data durability responsibilities and vulnerability to secure the data, 
  • it is difficult to quantify the actual real costs of the “On-premise PACS model” as there are many areas of other costs. For example, data centres, IT support staff focused on maintaining the system, etc 
  • major step costs and challenging change management if the business is growing organically  
  • healthcare imaging organisations that are growing via mergers and/or acquisitions often experience significant difficulties in bringing together multiple PACS onto a single platform because of the complexity of migration and hardware capacity including the developing model of imaging operations that are multinational in nature

These challenges continue today. 

The cloud has been broadly available as a commercial option for some time now, with the establishment of the large operators commencing commercially available operations around 15 years ago 

The key point this raises is the two different types of cloud hosting.  

  1. The one more commonly seen today is a non- cloud-native PACS being hosted in the cloud  
  2. A cloud-native PACS which is the future of PACS.

In the next article, we’ll cover the differences between what it means to be a cloud-based PACS vs a cloud-native PACS. 

We’ll go into detail about the “lift and shift” model of the existing on-premise PACs model, and why it isn’t always the best solution.

Introducing AdvaPACS

AdvaPACS is a fully featured cloud-native PACS built and managed by AdvaHealth Solutions.


  • Alan McCarthy

    Alan has been in the healthcare industry for over 30 years in leadership roles in the provision of healthcare services as well as leading major MNC medical companies, including Cardinal Health, Carefusion and Philips. He has also shown his entrepreneurial skills in building valuable healthcare assets in Australia, New Zealand and Asia. He is passionate about continuous improvement and finding value for a country’s healthcare spend.