Kevin is a seasoned C-Suite with over 25 years of experience and an expert in cognitive science.
Tell me more about CureMetrix.
CureMetrix is focused on deep learning and medical imaging. We’ve created a technology that takes medical imaging, finds anomalies and classifies them as either suspicious or non-suspicious. Our first detection target is breast cancer.
How exactly does CureMetrix’s AI help detect breast cancer?
We pair deep learning with computational physicists, computer-vision specialists, and radiologists; the AI is learning from the images and from cancer experts. We apply science, physics, and big data techniques; we run an algorithm that reads images, runs its process and tells us what it found. CureMetrix scores the anomalies that we find and we unpack the score to explain what it means. It’s the score that the AI assigns which indicates whether an anomaly is suspicious: indicating the likelihood that a patient is breast cancer positive or negative.
What AI algorithm / tech stack are you using at CureMetrix?
Our algorithms are all developed in-house and are uniquely clear-box (vs. black box), lightweight, and extremely accurate
Has IoT come into the picture for CureMetrix at all?
What trends do you see in AI and saving people's lives?
There are a number of players jumping on AI for health applications. These applications will inarguably augment the capabilities of doctors. With AI, doctors can apply real-time augmented intelligence to solve medical problems and save people’s lives.
How can AI for cancer help enterprises financially?
Through cost reduction. Currently, the cost of false cancer positives costs the US $4B a year. By better predicting and identifying breast cancer, AI will help reduce those costs by eliminating unnecessary procedures and catching disease earlier.
What challenges do you see in AI for cancer?
There are three things that come to mind.
- Not all AI’s are made equal. Doctors, hospitals, and investors will have to identify quality AI technologies over the noise.
- Doctors’ acceptance and adoption. Doctors fear that technology may replace them and may be resistant to AI technology adoption. Doctors will need to understand that technology is here to augment their abilities.
- Trust between doctors and technology. The gap in collaboration between the health and the tech industry needs to be filled. We must work together.
Can you share your favorite AI use case?
In the course of our learning and study, we’ve collected 500,000 images to train, test, and validate to make sure our technology is working. During this time, we found a patient who was diagnosed with breast cancer in 2010. We went back to 2007 and ran our AI and looked at the images from that time period. CureMetrix was able to detect breast cancer from the 2007 images and track its progress over time.
Bottom line: CureMetrix was able to predict breast cancer. That’s big!
“Mammograms, self-exams, and early detection of cancer save the lives of women.” - Kevin Harris
That’s impressive. Is there anything else you’d like to share?
This is just the beginning for us. As a startup we’ve had to pick one thing to focus on - the early and accurate detection of breast cancer in mammograms - and get it right.The algorithm and platform we’ve developed will let us target other modalities, parts of the anatomy, and diseases. We’re excited to see how our technology can help improve patient outcomes in the years to come.
How can we learn more about you and CureMetrix?
You can read more about us here.