Penn Center for Innovation

I-Corps Blog: Automated Anatomy Recognition

Posted on July 17, 2015 — pennadmin

Week 1

We help physicians make better treatment decisions through the segmentation and quantification of medical images. Automatic Anatomy Recognition is an innovative technology for segmenting and quantifying medical images. Through collaboration with Penn’s Department of Radiation Oncology, we identified a first application for this technology in radiation therapy planning. Our participation in I-Corps will help us identify other areas where we can enhance clinical decision-making.
Week 1 forced to us put aside our technology and spend more time thinking about our customers. We considered the following 4 questions:

  1. 1 What problem do you solve?
  2. 2 Who’s got the problem? (i.e. who are your users?)
  3. 3. What is the impact of the problem?
  4. 4. Who is willing to pay to solve this problem?

These questions helped us realize that we have 2 distinct value propositions that are offered to different customer segments. We are ready to start interviews!

Week 2

The first week of interviews helped us role out some customer segments and value propositions. While cardiologists use quantified information from medical images to make clinical decisions, we were unable to identify unmet needs that our technology could solve. We originally though that our technology could result in faster treatment response, but realized that we had no clinical data to support such a value proposition.

The biggest “Aha!” came during an interview with an industry expert in PET imaging. He helped us identify applications in medical oncology for our technology – specifically, producing more reliable information from PET images to guide treatment decisions. He also recommended that we interview medical oncologists and radiologists who specialize in molecular imaging. Our first interview with a medical oncologist helped us better define the problems. We look forward to additional discussions on this topic!

Week 3


Category: i-corps, i-corps-team-posts