Towards a Better and Accessible Healthcare Industry with AI – Hear from Margaretta Colangelo

Towards a Better and Accessible Healthcare Industry with AI – Hear from Margaretta Colangelo

Bala Priya C
·Jan 5, 2022·

9 min read

Subscribe to my newsletter and never miss my upcoming articles

Play this article

Table of contents

  • Advances in Healthcare
  • Leveraging AI for Healthcare Applications
  • Enabling Quality Healthcare for All
  • Tech and AI for Good
  • AI in Drug Discovery
  • Prospective Guest and Recommended Resources

We, at the AI Time Journal thank Margaretta Colangelo from Jthereum for taking part in the AI and Robotics for Healthcare Interview Series and sharing insights on the applications of AI in medical diagnosis, drug discovery, precision health and personalized medicine.

Advances in Healthcare

Bala: According to you, what are the most significant advances in healthcare in the recent years?

Margaretta: There have been many significant advances and next generation AI is now successfully implemented in many areas of healthcare including medical imaging, drug discovery, clinical trial design, medical research, and biomarker development. There are many examples of areas where AI is excelling. I’m on the advisory board of the AI Precision Health Institute (AI-PHI) at the University of Hawaii Cancer Center and I follow their work closely.

They are having success using AI in breast cancer detection and risk assessment, diagnosis of difficult melanoma cases, aging and frailty research, pancreatic and liver cancer, and algorithms to estimate specific fat regions including visceral, liver, pancreatic, and subcutaneous fat in young children because obesity on children increases risk of cancer as adults.

Scientists at the AI-PHI are also collaborating with NASA to develop AI powered technology that astronauts can use to monitor their health on long duration space missions. As a result of microgravity, astronauts can lose as much as 50% of their muscle mass in less than 6 months in space. Astronauts may also develop osteoporosis while on long duration missions such as the journey to Mars resulting in vertebral or hip fractures. The AI-PHI is working with NASA to develop technology that will enable astronauts to precisely modify their nutrition and adjust their fitness training to minimize muscle loss and loss of bone density in space.

Leveraging AI for Healthcare Applications

Bala: What motivated you to focus on leveraging AI for healthcare applications?

Margaretta: My background is in technology. I’ve worked in software companies in Silicon Valley developing emerging technologies for over 30 years. During that whole time I thought about how we could leverage advanced technologies like AI and nanotech to help eliminate sickness and disease.

When I was young I cared for both of my parents when they suffered and died from very serious terminal illnesses. I’ve lived my entire adult life without my parents. The experience of caring for my parents made me ​acutely aware of how much pain and suffering people endure when they get sick and I became very interested in learning how we could apply advanced technologies to keep people healthy.

​I’ve been waiting all of these years for technical progress, and now, technology is finally at the point where we can use it to help people stay healthy.

Bala: AI in Healthcare is a promising direction because of advances in research, and easier access to compute. Could you please give an example of how AI can be used in medical diagnosis?

Margaretta: Researchers at National University of Singapore used AI and inexpensive equipment to identify live cancer cells in less than 35 minutes with 95% accuracy. They were able to discriminate between normal and cancer cells, discriminate between benign and metastatic cells, and classify cancer cells that originated from different organs.

The ability to analyze single cells is one of the holy grails of health innovation for precision medicine and personalized therapy.

The researchers developed a novel protocol for single-cell classification based on intracellular pH. Since cancer cells have lower intracellular acidity than normal cells, the AI algorithms were able to identify cancer cells by examining the level of acidity. This novel technique is an important advance because traditional techniques such as fluorescent probes induce toxic effects and kill the cells. The new technique involves using a pH-sensitive dye that changes color according to the level of acidity and does not harm cells.

Using AI, they were able to quantitatively map the unique acidic fingerprints so that the cell types examined can be easily and accurately identified. Using AI they were able to analyze single cells faster, cheaper, and with higher accuracy, and most importantly the AI technique does not kill the cells that are being analyzed. After analyzing the cell, the AI generates a report with analysis. The paper was published in APL Bioengineering.

Enabling Quality Healthcare for All

Bala: If AI systems for medical prognosis and diagnosis can be deployed at scale, Remote Diagnosis-as-a-Service would be possible. Could you tell us how this would improve accessibility to quality healthcare in developing regions?

Margaretta: AI can help extend the reach of care outside of hospitals​ and​ expand access to care in places where doctors are inaccessible​. ​AI can enable anyone with a smartphone to have access to healthcare​. This will help​ people living in the developing world and people who work in remote environments​. Low cost portable devices embedded with AI are already assisting healthcare workers screen, assess, and diagnose patients​ in remote locations, even in places without electricity or internet access. ​

We can use AI to decrease disparities in healthcare in developing regions. For example, there’s a global shortage of radiologists. 2/3 of the planet’s population, 4 billion people, don’t have access to radiologists.

Consider this: On the entire continent of Africa, if you remove Egypt and South Africa, there are only 6 pediatric radiologists. In Nigeria there are fewer than 60 radiologists for 190 million people. In Mexico there are only 4,000 radiologists for 130 million people. In Japan there are only 4,500 radiologists for 126 million people. In Liberia there are only 2 radiologists for 4.9 million people. 14 African countries have no radiologists at all. But, one hospital in Boston (Massachusetts General Hospital) has 126 radiologists, and there are over 34,000 practicing radiologists in the US.

It takes over 10 years to train a human radiologist so the only way that we are going to be able to decrease this disparity is using AI embedded in inexpensive portable devices to analyze medical images.

Tech and AI for Good

Bala: Tech and AI for Good – this notion has been driving many innovations. Could you let us know your views on this dictum?

Margaretta: My favorite example of AI for Good is a company called Arterys. This company developed an application that has helped doctors treat over 50,000 newborn infants born with potentially fatal heart defects. Arterys was the first company granted FDA clearance to use cloud-based AI in a clinical setting. Its cardiac application for infants is used in over 100 hospitals around the world and is considered by doctors to be one of the top of AI solutions available.

The team at Arterys developed AI based technology that made it possible for doctors to have a full view of a newborn baby’s heart, see blood flow, and accurately diagnose newborn babies. Prior to Arterys nobody had developed a reliable, non-invasive diagnostic tool that enabled physicians to visualize and quantify blood flow for cardiovascular disease. It takes a human radiologists an hour to analyze a cardiac image. It takes Arterys 15 seconds to analyze the same image with equal accuracy.​

The founders of Arterys thought that medicine should be powered by data and that doctors should be empowered with every piece of information they need to make the best decisions for their patients. They also believed that every patient should have access to the best technology regardless of where they live or how they live.

They believed that AI applications should be created, distributed, and used openly, and data needs to be aggregated in a central location, and open and available to everyone. So they built such a platform.

They developed the world’s first online medical imaging platform and changed the way image analysis is done and the infrastructure used.

Today Arterys has 6 FDA clearances and radiologists around the world can receive automatic and very precise measurements of medical images of brain, heart, lung, liver, breast, and everything in between.

Bala: How do you think the healthcare industry can embrace the power of AI, so that their synergy can help us all lead better lives?

Margaretta: Over the long term the healthcare industry could use AI to promote life long precision health. We could get to the point where AI could predict risk of disease at birth and help people avoid disease throughout their lives. This includes slowing down biological aging and improving mental health. This could eliminate so much suffering and improve the quality of life for people all over the world and for all future generations.

AI in Drug Discovery

Bala: Could you tell us about a few advances in drug discovery that have been facilitated by AI research?

Margaretta: A very important milestone in AI for drug discovery was published in a landmark study in Nature Biotechnology in 2019. The study detailed how Insilico Medicine succeeded in using AI to design a new molecule from scratch in 21 days and validate it in 25 days. This was the first time that using AI in the Pharma industry was validated in practice. Insilico Medicine used a generative approach combined with deep reinforcement learning to design and validate a new drug candidate end-to-end in 45 days. This is 15x faster than Pharma companies.

This achievement confirmed that AI could accelerate drug discovery which is very important because the traditional drug discovery process often takes decades and in preclinical phases the failure rates are over 99%.​​

​Insilico Medicine completed the preclinical experiments required to nominate a preclinical candidate in under 18 months.​ Insilico’s AI platform is exceptionally good at finding the molecular targets in specific diseases and inventing new chemistry. Insilico was the first company to​ ​​demonstrate the power of AI to accelerate the pace of scientific R&D and had a big impact on the Pharma industry. Today, 7 out of the top 30 pharma companies use Insilico’s AI platform including Merck KGaA and UCB. Insilico Medicine also built a drug discovery team in China and started multiple therapeutic programs targeting novel, difficult, and previously undruggable targets.

Bala: Could you suggest a guest whom you would like us to feature in this series?

Margaretta: I suggest that you interview Dr. Geetha Manjunath, the Founder, CEO, and CTO of NIRAMAI Health Analytix. NIRAMAI uses AI to analyze thermography images to detect breast cancer. AI is the core of the technology because the human eye cannot interpret so many color points on an image. The device can detect tumors 5x smaller than clinical exams. The device has been used to screen over 32,000 women and is used in over 30 hospitals and diagnostic centers across India. The device is non-contact, radiation free, low cost, and portable.

Bala: Could you suggest any podcasts, and newsletters that’d help us stay up to date with the recent advances in tech?

Margaretta: I’ll be hosting a podcast series focused on AI in Healthcare at AI Time Journal. I​ also publish a weekly newsletter on AI and DeepTech. If you subscribe to my newsletter, you’d receive the newsletter in your inbox every week.

Share this