Image32 Simplifies Medical Image Sharing using Google Cloud Platform

Image32 makes technology that helps doctors and patients quickly, easily and securely share digital medical images such as MRIs and X-rays. Several components of Google Cloud Platform are key elements of the company’s business model, from Google Compute Engine for image processing to Google App Engine for fast, secure access.


Today’s medical world is increasingly digital and interconnected — until you hit the world of medical imaging, which looks like a frontier outpost in comparison.

Although images such as X-rays, CT scans and MRIs are almost always stored digitally, these images are difficult to access and share. Complicated rules, incompatible software and unwieldy file sizes mean that medical images end up stored in a separate silo. As a result, such images are often not integrated into patient electronic health records, hindering medical professionals’ ability to build a consolidated view of each patient.

Image32 is trying to bring medical images out of isolation with cloud-based software that makes uploading medical images fast, simple and secure. “Think of it as a kind of YouTube for medical images — our software is specially built to make it easy to view and share medical images,” says Bob Pellican, co-founder and CEO of Image32. The application encrypts personal health information as needed for regulatory compliance, compresses the images in a lossless format, and stores them in the cloud. ”Patients and providers can view the images and share them quickly and easily with each other,” says Pellican. This means that doctors can review images before the patient even comes in the door, and patients can upload and manage their own medical information. And with the average angiogram at about 3 gigabytes and MRIs similarly data heavy, it also means that Image32 needs serious computing horsepower and large amounts of storage to support their business model.


The company didn’t want the expense of building an on-premises data center or the accompanying management and maintenance chores. Google Cloud Platform‘s combination of power and affordability was the perfect alternative. Image32 uses several different components of Cloud Platform. When a patient or provider uploads an image, Google Compute Engine processes the images using Image32’s lossless compression technology and sends the file to Google Cloud Storage, which currently houses more than four million high-resolution medical images. From there, Google App Engine acts as the company’s main server, accessing files as needed.

In addition, the fact that Cloud Platform is HIPAA-compliant makes it even more appealing, says Pellican. “Being able to run on a HIPAA-compliant cloud infrastructure with an attractive cost structure helps us offer a more attractive product and service than our competitors who have sunk time and money into in-house data centers,” he says.


One of Image32’s chief requirements is to provide a fast, easy and reliable experience for the people uploading images, and Cloud Platform delivers. Images take between 10 seconds and 5 minutes to process, depending on the complexity of the image. “Many of our clients report that the chief bottleneck is likely to be their own Internet connection rather than anything platform-related on our side,” says Pellican. “It’s reliable too. We’ve had 8 months in a row of 100% uptime, with millisecond response times.”

Cloud Platform also obviates the need for anyone in the company to deal with infrastructure and network operations. While other Infrastructure as a Service (IaaS) options still require customers to spend time on tasks like load balancing or building server instances, Google builds those capabilities into Cloud Platform. “That lets our engineers focus on developing and improving our app rather than keeping the lights on,” says Pellican. “In teams that I’ve worked with at previous jobs, deploying and maintaining the technical environment can take up to 10-15% of engineering resources, so offloading that task is no small achievement.”

As a startup, Image32 also needed a platform that’s easy to scale. The company’s client base varies widely, from individual patients who want to manage their own images to clients running clinical trials with thousands of images, so Image32 needs a platform that lets them quickly respond to new projects and request for proposals. “For example, let’s say we’ve put in a proposal to upload the medical images of a state health department that has 10 years of archives, comprising more than a terabyte of data,” says Pellican. “If we win that project, we suddenly need to upload and store a large amount of data for a client that isn’t going to want to wait.” Cloud Platform gives the company the agility to quickly access and use technology on demand. In Pellican’s scenario, the company could scale up parts of the systems to handle uploading and processing images and scale back down as soon as that part of a project is finished.