Pathology is the area of medicine focused on making a definitive diagnosis of a serious disease through laboratory examination.
“If you go to the doctor with a cough, they might send you to a radiologist to look at your lungs,” PathAI Co-Founder and CEO Andrew Beck explained. “If they’re concerned and don’t know what it is, they’ll take a piece of tissue and send it to pathology to find out.”
PathAI is a company born from Beck’s interest and background in this area of medicine. The company combines the methodology of machine learning with vast swaths of medical data to help doctors make better diagnoses. In today’s technologically advanced world, the pathology lab hasn’t really evolved. The use of microscopes and glass slides are still commonplace.
PathAI is looking to change that, and digitize the pathology lab using machine learning and other software tools. Their platform automates the handling of monotonous tasks – and helps pathologists make a more precise diagnosis – by removing the potential for human error when determining a diagnosis.
Beck, who earned an MD from Brown Medical School, did his residency in pathology at Stanford, where he saw machine learning’s ability to accomplish a number of different tasks in the lab.
“I knew the future was going to involve a lot of data-rich pathology. There’s so much new technology to generate lots of data. There’s also imaging methods, molecular methods, and there’s more and more types of data. So, you need machine learning to make sense of it.”
After his residency, Beck earned his PhD in machine learning at Stanford University – specifically Biomedical Informatics, which focuses on using health information technology to improve healthcare. At the Stanford Artificial Intelligence Laboratory, he worked on computer vision for pathology.
Roughly five years ago, Beck came to Boston, where he started a research lab at Harvard Medical School and the Beth Israel Deaconess Medical Center to focus on using machine learning to improve pathology.
“What I saw over the past three years or so was really the convergence of cloud computing – which gave everyone incredible power – and computer vision core algorithms really started working for things like self-driving cars and facial identification. And there’s just better and better tools for digitizing pathology data.”
Beck conducted a study in his lab, led by postdoctoral fellow Dayong Wang who is now at PathAI, to see if these methods could be applied to pathology, and got early indications that AI and pathology could work well together. Seeing these positive results, he began to consider how he could make the most impact with this technology. A company would be the solution.
He ended up connecting with his Co-Founder and CTO Aditya Khosla, who was just finishing up his PhD in Computer Science at MIT.
Beck read an article about Khosla’s research and reached out with an email to invite him for coffee. “I convinced him that this was an interesting problem, and an important problem worth solving,” Beck explained. As a result, the duo founded PathAI in early 2016.
“The company started very small. We were actually based in my basement in Brookline with me, Aditya and Ryan McLoughlin (previously a software engineer at Google) until the end of last year. We then moved to our current place in Inman Square, where we started building out the technology platform and hiring folks.”
Beck, Khosla, and the team have been building a platform that can take in pathology data (typically digitized glass slides), annotate that data at scale, and produce “really high-quality deep-learning models to solve a whole array of problems in pathology.”
The company’s technology is currently being developed for applications to be used in three different ways:
- Empowering pathologists to make more accurate, standardized diagnoses using a consistent, AI-driven methodology.
- Accelerating drug development with pharmaceutical companies by increasing analytical capabilities and automating certain tasks, so scientists can focus on the more strategic biological implications of drugs.
- Applying the technology in global health. With a project funded by the Gates Foundation, PathAI is currently developing AI-driven applications that can provide pathologic diagnoses to developing nations at a low cost.
PathAI raised an $11M Series A round of funding led by David Fialkow of General Catalyst in November. The company has raised a total of $15M in funding to date.
The biggest challenge for the company, as Beck discussed, is to help transform the medical practice of pathology – in which diagnoses today are made using glass slides and microscopes – into a data-driven digital medical specialty where glass slides are digitized and diagnoses are made augmented by artificial intelligence.
For now, the company focuses on the aforementioned three areas, but in the future, Beck sees a world where his platform can set a new standard in pathology.
“Instead of just using a microscope, now they’re going to have the world’s most powerful AI pathologist assisting them in making diagnoses faster and more accurately.”
Images provided by PathAI.