Artificial Intelligence Makes Waves in Breast Cancer Research & Awareness

Artificial Intelligence (AI), machine learning and data mining technology interface

We’re excited to share an interview with Dr. Mark Traill, who is the owner of MammoBot LLC. Mammobot is a cloud-based startup that offers a double reading of screening 2D or 3D mammograms using artificial intelligence algorithms. Clients can upload mammogram studies that become part of the investigational protocol prior to pending FDA approval of the algorithm. Review of the mammogram study by AI algorithms has the potential of detecting early breast cancer that was missed on the initial human review of images.

Ambra Health: When did Mammobot begin? 

Dr. Mark Traill: I started Mammobot a little over one year ago with my partner, Dr. Jordan Toman. I had been looking for access to develop artificial intelligence algorithms. We started testing the workflow with Qure.Ai in Mumbai. I was able to successfully upload test files using Ambra Health. Qure then could access files, run them through a chest algorithm, and return results through the cloud in a HIPAA compliant fashion. I then approached Cure Metrix located in La Jolla, CA who had a 2d mammography algorithm where MammoBot studies could be processed. Recently, we signed an affiliation with the University of Michigan. 

Ambra Health: Who are the target users of Mammobot? Is it designed for patients or providers? Both? 

Dr. Mark Traill: Both! But let’s start with providers first. We’re focusing on providing this technology to hospitals, imaging centers, and second opinion groups. Most of the radiologists in those facilities are dealing with a giant backlog of mammography imaging. There’s a real shortage of trained mammographers and many are very burnt out and overwhelmed by the high volume of studies they are expected to read. Eventually, we would like patients to be able to upload studies for an additional AI (and physician) second opinion.

Ambra Health: We’ve certainly heard that burn-out is a big problem in the radiology community. Do you think an AI algorithm like MammoBot could reduce human error in these situations?

Dr. Mark Traill: Absolutely. There’s a significant error rate on routine mammography and many early cancers are missed. An AI algorithm can help provide clarity by reducing redundant imaging and unnecessary biopsies in healthy women. And on the other end, it can also help detect cancer much earlier, preventing some of the high costs associated with treating later stage cancers.

Ambra Health: As a physician founder (and radiologist) you certainly saw a need for mammograms to be read in a faster, cheaper, and more accurate way. But what about the technical aspect? What were you looking for in terms of a solution that could transfer the images?

Dr. Mark Traill: I had some familiarity with cloud image sharing prior to MammoBot as I used such services to securely offer second opinions to colleagues. Most importantly, I was looking for a solution that was HIPAA compliant. In terms of my other requirements, they revolved around the large file size of mammography exams. I needed a platform that could upload these large images quickly and allow an annotated version to be sent back to me that I could review again on a viewer in turn. Ambra’s been really helpful in getting the image management side of things up and running.

Ambra Health: Thank you for sharing MammoBot’s story with us! We look forward to seeing the platform grow and improve the detection and treatment of breast cancer.

Catherine Slotnick, Marketing Manager

About Catherine Slotnick

Catherine Slotnick is a passionate healthcare marketer with a deep interest in the latest & greatest in the Health IT space. As Ambra Health's Director of Marketing, Catherine primarily focuses on creating and sharing thought leadership content in the radiology and informatics space. Catherine graduated from the University of Virginia with a BA in Psychology & Art History. When she's not writing, she enjoys cooking and petting dogs that aren't hers.

Medical Imaging Software Buyer's Guide

Leave a comment

Your email address will not be published.