Axial3D Helps Surgeons Save Lives with 3D Medical Models Powered by AWS

Axial3D helps surgeons save lives using detailed preoperative 3D medical models. Using AWS, Axial3D creates and delivers 3D-printed models in a day or two, a task that previously would have taken up to eight weeks. An AWS Partner Network (APN) Public Sector Partner, Axial3D used Amazon SageMaker, Amazon EC2, and Amazon S3 to help surgeons at Belfast City Hospital successfully perform a kidney transplant that saved two lives.

Across the globe, surgeons are increasingly using 3D-printed anatomical models to prepare for complex operations by creating detailed surgical plans. However, the process of segmenting 2D images into multiple labeled regions to produce a 3D model is often time-consuming, requiring clinicians to be away from treating patients.

Axial3D, a medical 3D-printing technology provider, aims to solve this problem. The company offers patient-specific, accurate 3D-printed models that can be created quickly. “We specialize in converting images to 3D-printed files that surgeons use for preoperative planning when they need a better understanding of surgeries,” says Daniel Crawford, founder and chief strategy officer for Axial3D.

An AWS Partner and a member of the Public Sector Partner Program, Axial3D runs its web-based 3D modeling application on AWS. Crawford says, “Using AWS, we have the agility and scalability we need to offer our customers fast access to high-quality 3D printing.”

Axial3D relies on Amazon Elastic Compute Cloud (Amazon EC2) to host its web platform, which surgeons use to easily and quickly place orders for a 3D-printed model. Axial3D uploads 2D images from MRI or CT scans to Amazon Simple Storage Service (Amazon S3) and then runs machine learning models based on Amazon SageMaker. “Using Amazon SageMaker, we can quickly deploy new models and perform fast inference of large medical image datasets,” says Crawford.

One clinician taking advantage of Axial3D to perform complex surgeries is Dr. Tim Brown, a renal transplant surgeon at Belfast City Hospital in Belfast, Northern Ireland. Brown recently treated a young female dialysis-dependent patient with end-stage kidney failure. The woman required a second kidney transplant and was scheduled to receive a living-donor transplant from her father, but tests revealed a cyst—potentially cancerous—on his kidney. Brown says, “She only had an estimated six months to live, so we had to act quickly as far as operating on her father.”

Brown used Axial3D to generate an accurate 3D model of the father’s kidney in advance of surgery. Using a physical 1:1 scale model of the father’s donor kidney, including the cyst, Brown conducted a laparoscopic donor nephrectomy with an ex-vivo partial nephrectomy, followed by a living-donor transplantation of the repaired kidney into the daughter. “We studied the model closely, and it reduced variability from 2D imaging and helped us plan the operation more precisely. It was almost like running through a dress rehearsal before a performance, making sure we knew exactly what to do,” Brown says. “Relying on Axial3D, we could prepare better for this very complex surgery. As a result, we completed the transplantation successfully and were able to save two lives.”

Using AWS, Axial3D has transformed the process of producing these 3D-printed models in a day or two, a task that previously would have taken at least six to eight weeks. “By hosting our applications on AWS and taking advantage of Amazon SageMaker to enable the training of machine learning algorithms, we have automated the process of converting MRI and CT images from 2D to 3D,” says Crawford. “Amazon SageMaker specifically helps us reduce the amount of segmentation required during model creation. This enables us to support large volumes of medical images and speed up the process of creating and delivering anatomical models.”

Brown adds, “With Axial3D, we used an interface to simply click a few buttons and upload patient imaging data easily. Axial3D then delivered the model in less than two days.”

Axial3D customers like Belfast City Hospital can empower and educate patients before surgeries by using 3D models to demonstrate anatomy and reference specific problems. “We showed the 3D model to our kidney donor, and it helped him clearly understand his condition,” says Brown. “When we can educate patients by showing them exactly what’s going on inside their bodies, it removes some of the apprehension they may have had about a surgical procedure.”

Axial3D is expanding its use of the 3D technology to support additional use cases. “We offer a web application that is applicable to any clinician because it enables medical scans to be sent directly from a hospital, processed on AWS, and then sent back to the hospital,” Crawford says. “We are now seeing more cardiologists and urologists using our technology to improve patient outcomes. Cardiology in particular involves extremely complex anatomy, so it’s critical that surgeons have an accurate presurgery planning tool. That’s what we’re providing, thanks to AWS.”