AWS Startups Blog

Axial3D: Powering the Revolution in Medical 3D Printing with Amazon SageMaker

Guest Post by Roger Johnston, CEO, Axial3D

For surgeons, having fast access to detailed, reliable patient data is vital for success in the operating theater. Too often in complex cases, 2D imaging does not provide enough detail or insight to clinicians to give them full certainty of their approach in surgical plans.

Of course, in most cases, CT and MRI scans give clinicians an acceptable level of insight to give them the confidence to make a decision and proceed with a surgery plan. However – for the three million complex cases that need to be operated on each year, a 2D scan doesn’t always cut it for planning surgery and communicating your course of action to your patient.

Axial3D provides clinicians with the insight and confidence they need to create a surgical plan in the form of high-quality, patient-specific 3D anatomical models.

3D printing standardizes the approach in interpreting the patient’s anatomical detail and takes away unnecessary variability from one surgeon’s anatomical interpretation from another. These models can be held in the surgeon’s hands and fully scrutinized – allowing them to create and practice a surgical plan before they set foot in the operating theater.

“Being able to utilize a 3D model and hold it in your hands prior to surgical intervention is very insightful for the patient – it helps them understand surgical interventional risks and benefits,” says Matthew Lawson, MD, a neurosurgeon and Stroke Medical Director at Tallahassee Memorial Healthcare. “3D models have been shown to increase patient consent rates and aid in patient understanding of the planned procedure.”

Currently, the technology has not gained huge ground outside of a core user-base, many of which are early adopters. This is soon to change, however, thanks to innovations that break down the barriers to affordable, high-quality patient-specific models in lightning-fast turnaround times – or as we like to call it, the revolution in medical 3D printing.

The Challenge – Understanding the barriers to accessing medical 3D printing

The application of utilizing 3D printing to create patient-specific models for pre-operative planning is still in its infancy. In fact, Gartner recently announced that only ~3% of hospitals and research institutions have 3D printing capabilities on site. More hospitals are adopting the technology each year, take the new 3D print lab at Newcastle’s RVI for example, however it’s still not nearly as widespread as it could be.

One of the reasons why the technology has not been more widely adopted, and often seen as being one of the largest bottlenecks for the routine production of 3D printed anatomical models, is the availability of radiologists or biomedical engineers available to segment the 2D images. The segmentation process is the partitioning of an image into multiple labeled regions – locating objects and areas of interest in images. This can be an extremely time-consuming process and take clinicians (such as radiologists) away from treating patients for hours at a time.

The Goal – Removing the bottlenecks in medical 3D printing

With over three million complex procedures taking place each year, automation is essential if 3D printing is to become a go-to pre-surgical routine in healthcare. The segmentation process currently takes anywhere between 4-10 hours per printed model, but we aim to reduce this by platforming on AWS and using machine learning algorithms to augment and automate the process of segmenting the 2D images and turning them into 3D printable objects. This combination allows us to deliver near-instantaneous results and removing the main bottlenecks associated with medical 3D printing.

The Solution – Making 3D printing routine in healthcare

By platforming on AWS, Axial3D is helping to transform surgery by providing previously unavailable insights to clinicians for preoperative planning across the world. Axial3D achieves this by creating highly accurate 3D printed models of specific parts of the patient’s own anatomy, that allow clinicians to better prepare for surgery. This results in better clinical outcomes for patients, significantly reduced planning and operating time (and costs) for the surgery and patients who are much better informed about their condition and the proposed surgery.

“Certainly, in my own practice, with Axial3D we have created a fantastic interface where I can click a couple of buttons and upload patient imaging data and they then turn around and deliver the model it a couple of days,” says Dr. Tim Brown, a consultant transplant surgeon at Belfast City Hospital.

Axial3D uses EC2 to host the infrastructure that allows surgeons to easily and quickly place orders to request a 3D printed model. We store the images on S3 and record metadata about them on DocumentDB allowing us to quickly and easily track and sort our data. This scalable storage allows us to deal with large volumes of medical images quickly, in turn speed up the process of creating the 3D printed models so that we can produce and ship patient-specific 3D anatomical models within 48 hours.

By applying machine learning to medical image segmentation we have reduced our processing time to a few minutes. We are able to quickly deploy new models to SageMaker as they become available, facilitating rapid testing of new architectures and benchmarking performance of algorithms over time. The auto-scaling features of SageMaker allow us to process thousands of images simultaneously; automatically triggered by an upload to S3. The efficiency savings provided by platforming on AWS are passed on to our end-users, enabling us to provide the highest-quality segmentations and 3D printed models at lightning speed and incredibly low cost.

With a worldwide network of customers and partners, Axial3D has been able to transform how we approach the application of 3D printing for healthcare. This novel approach is driving the global adoption of the technology within healthcare institutions and across medical facilities. Our partnership with AWS has enabled us to provide a super-quick ‘DICOM to model’ service to clinicians wherever they are in the world, 24/7/365. The effect this has on patient care is game-changing. No longer will radiologists have to spend hours segmenting images to make them 3D-printable. No longer will surgeons need to wait weeks on a 3D printed model being produced and shipped. No longer will hospitals be expected to pay upwards of thousands of dollars for a single 3D printed model.

Surgeons working with Axial3D have reported increased levels of patient understanding, better intra-team communication, and positive patient outcomes when using a 3D printed anatomical model in planning for surgery.

“Axial3D’s services have allowed us to use 3D printing for complex cases without the cost of building our own lab and hiring new staff,” says Dr. Christopher Lockhart, a consultant cardiologist in adult congenital heart disease at Royal Victoria Hospital. They manage the entire 3D printing process, and we focus on treating patients.”

The revolution in medical 3D printing has begun. Will you be one of the 25% of surgeons expected to use a 3D printed model in pre-operative planning by 2021?

Join the movement that gives you increased confidence ahead of complex operations, enhanced insights into complex pathologies, and better patient communication. Try a free patient-specific model from Axial3D and explore the benefits for yourself.

 

Interested in ML on AWS? Contact us today!