AI for Medical Imaging needs to be Generalizable. CT Images vary depending upon the vendor, age and acquisition protocol. AI is not valuable to a clinician if it only works some of the time. Our AI has built in redundancy to make it generalizable to almost any images thrown at it.
Artificial Intelligence is transforming industries at a breakneck pace, and the entrepreneurs driving this innovation are at the forefront of this revolution. From groundbreaking applications to ethical considerations, these visionaries are shaping the future of AI. What does it take to innovate in such a rapidly evolving field, and how are these entrepreneurs using AI to solve real-world problems? As a part of this series, I had the pleasure of interviewing Chris Wood. Chris is a long-time innovator in medical imaging, whose career has been shaped by both personal experience and a deep commitment to diagnostic precision. His interest in imaging began as a child, after his mother’s inconclusive breast cancer diagnosis led to a course of aggressive treatment that ultimately saved her life. That early encounter with diagnostic uncertainty became a formative influence, leading him to co-found Confirma in 2000, a company that developed the first computer-aided detection software for breast MRI. His subsequent ventures, including Clario and RevealDx, have focused on solving targeted problems in imaging workflows and enhancing diagnostic accuracy, particularly in the detection of lung cancer. With RevealDx, Wood is now at the forefront of integrating AI into medical imaging, aiming to reduce unnecessary procedures, lower healthcare costs, and accelerate life-saving diagnoses.
Thank you so much for joining us in this interview series. Before we dive in, our readers would love to learn a bit more about you. Can you tell us a bit about your childhood backstory and how you grew up?
It wasn’t until recently that I realized why my entire career focused on diagnostic imaging. Even as a little kid, I loved science and I studied physics in college, but I always thought I just “happened” to focus my work on improving the performance of radiologists. I realize now that my choice was driven by the most important event in my childhood. My Mother was diagnosed with Breast Cancer when I was 11. This was devastating to my family, causing huge financial and emotional stress. What I remember as most difficult for me, however, was that the diagnosis was not conclusive. There was a disagreement among the pathologists about whether she even had cancer. To me, at 11, that meant something else could be wrong and they did not know what it was. Ultimately, she opted for very aggressive treatment, and it had the desired effect. She remained cancer-free for the rest of her life.
In 2000 I co-founded Confirma, which created the first Computer Aided Detection software for use on Breast MR images. Breast MR is highly sensitive, often referred to as the “definitive negative test”. Confirma helped drive adoption of this test, which is something I remain very proud of.
Can you share the most interesting story that happened to you since you began your career?
I worked for a brief time in image guided neurosurgery. Transporting, scanning and clamping on human cadaver skulls into place for training, is probably the strangest thing I have ever done.
None of us are able to achieve success without some help along the way. Is there a particular person who you are grateful towards who helped get you to where you are? Can you share a story about that?
For me, instead of an individual, it would be all the “old timers” I got to work with when I first started at Picker International in 1989. Picker X-Ray was founded by James Picker just after the invention of the X-Ray machine, and when I started working there, I got to learn from the best. This was a company who had built air-dropped X-Ray labs for the Army in World War II and the Korean War. The company knew how to build exceptional medical imaging equipment very efficiently, and the employees had decades of knowledge about the business, as well as the technology, that they were more than willing to share with the new engineers.
Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?
Among a few of my favorite quotes from learning ample life lessons in my journey scaling and creating successful companies in imaging technology innovations is, “When one eye is fixed upon your destination, there is only one eye left with which to find the Way.” –Joe Hyams, Zen in the Martial Arts. This quote reminds me to solve the problems in front of me. Don’t focus on the long term financial rewards of building the business, but the rewards of the little successes that bring you closer to your goal.
MedTech companies take years to build so you need to pace yourself a bit and enjoy every minute of it. Embrace change, and know that new challenges are always around the corner.
You are a successful business leader. Which three character traits do you think were most instrumental to your success? Can you please share a story or example for each?
First, I like to dive deep into a problem. This is something that bigger established companies can have a hard time doing. For example, I started a company called Clario that just built a worklist. Nearly every company in the market was building a worklist, an image viewer, and an archive. By focusing on just the worklist, we built something that was much better than anything in the market. Our superpower was that the company focused on these specific problems without distractions.
Second, is approaching problems with curiosity. I have been working my entire career to do more of this, because the reward is creative solutions.
Third is empathy… specifically for the patient. I worked at Moffitt cancer canter in graduate school. This early exposure to the impact we have on patients is something I have carried with me. There were times I would write software and shortly after, see the impact it would have on patient care. After seeing that, I couldn’t imagine working in a different field.
Ok super. Let’s now shift to the main part of our discussion. Share the story of what inspired you to start working with AI. Was there a particular problem or opportunity that motivated you?
At Moffitt, we had a grant from NASA to apply satellite image analysis software to medical images for the purpose of quantifying disease by using segmentation. The software worked well, but when the data does not follow a normal distribution, we found that “Neural Networks” could work better. This can happen in medical imaging, and was the basis for my masters thesis.
Describe a moment when AI achieved something you once thought impossible. What was the breakthrough, and how did it impact your approach going forward?
Many in our industry were impressed with the results from Alex Krizhevski and Geoffrey Hinton when they made a giant leap in performance at the ILSVRC competition. This was deep learning’s debut in image recognition. Within 4 years, these deep learning computer vision techniques surpassed human performance.
By 2014, Billions of dollars were being invested in Medical Imaging AI companies. We are now at the point where we know how to apply this fantastic deep learning technology to diagnostic decision making, and it is starting to be adopted by Radiologists who know their jobs aren’t going anywhere.
Talk about a challenge you faced when working with AI. How did you overcome it, and what was the outcome?
In healthcare, the main challenge is almost always getting good quality data that is suitable for training the AI. The medical imaging AI world has lots of examples of quickly trained AI running amuck. This happened a lot during COVID. For example, while attempting to train AI to recognize COVID using X-Ray Images, researchers trained AI using images from patients with and without COVID. What the researchers failed to recognize was that most of the COVID patients were imaged laying down, while most of the normal patients were imaged standing up. The resulting AI was able to identify how the patients were imaged, not if they had COVID.
We had to build our AI using a hand curated set of features, and trained with a smaller (yet higher quality) set of training images.
Can you share an example of how your work with AI has had a meaningful impact (on others, on business results, etc)? What was the situation, and what difference did it make?
While at Confirma, the breast cancer detection company, we would occasionally receive letters from patients who told them that their doctor credited our software for finding cancer they would have normally missed.
My current company, RevealDx, is starting to assemble similar stories that have similar patient impact, but we also know that we will be saving significant healthcare costs. Today, patients can be imaged 5–10 times before reaching a diagnosis, and a large percentage of lung biopsies are negative. We know that our technology can improve the diagnostic process, resulting in significant cost savings. This is why we are currently eligible for reimbursement by health insurance.
Based on your experience and success, can you please share “Five Things You Need To Know To Help Shape The Future of AI”?
1 . AI for Medical Imaging needs to be Generalizable
CT Images vary depending upon the vendor, age and acquisition protocol. AI is not valuable to a clinician if it only works some of the time. Our AI has built in redundancy to make it generalizable to almost any images thrown at it.
2 . Asserting an ethical-first approach
In any form of business AI practice, ethics should take precedence. As procedures, policies, and restrictions are constantly debated and decided on with various AI integration, it’s important to abide by the developing precautions and security standards for the safety of all users involved.
3 . AI needs to be clinically impactful
For AI to be truly valuable, patient outcomes need to be improved. When that happens, insurance begins to reimburse, and an AI company can become viable.
4 . AI needs to be Clinically Validated
AI for healthcare needs peer reviewed validation. This is a time consuming process, but a lack of publications will prevent sales from ever happening.
5 . Adapting and Maximizing
AI continues to get huge investments for applications outside medical imaging. It is our primary job to review all significant breakthroughs, and identify those that can be used to improve our diagnostic challenges.
When you think about the future of AI, what excites you the most, and how do you see your work contributing to that future?
I think AI will drive the personalization of medicine. We will know through genetic and other testing which individuals are at high risk for disease, and with AI, we can screen using imaging without overburdening health care. This means moving more toward disease prevention and early stage diagnosis.
Our company is focused on the biggest cancer killer, Lung Cancer. If you were to pick one thing to reduce cancer deaths, it would be accelerating the diagnosis of Lung Cancer. This disease is incurable when found at the later stage, but curable when found early. We are excited about the cost savings our product will generate, but we are even more excited about the lives that will be saved through early detection.
What advice would you give to other entrepreneurs who want to innovate in AI? Can you share a story from your experience that illustrates your advice?
Find something that is hard or impossible for a human to do. In my case, I have been hearing clinicians complain about lung nodules for 30 years. The guideline published by the American College of Radiology (Lung-Rads), which is designed to help categorize these nodules, has grown from 750 words at its inception, to over 1,800 words in the 2022 version, and it is bound to grow again. It is not possible for a busy radiologist to apply this many rules to one finding in an exam that is already exceptionally complicated to understand. This is exactly the type of problem that AI can assist with.
Is there a person in the world, or in the US with whom you would like to have a private breakfast or lunch, and why?
When I was younger I may have picked someone who could help me, or someone who would give me some special knowledge I could use for success. Today, it would probably just be a young person who has a problem with their business that they want to discuss and brainstorm.
How can our readers further follow your work online?
The best way to stay connected with my ongoing work with Reveal Dx’s initiatives is to follow me on LinkedIn where I often share ongoing milestones and breakthroughs to stay tuned on.
Thank you so much for joining us. This was very inspirational, and we wish you continued success in your important work.
Chris Wood Of RevealDx On Pushing the Boundaries of AI was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.