Radical Reads

Radical’s Sanjana Basu on the future of digital health care: COVID-19 has created health care’s long-awaited digital moment  (Toronto Star)

“In response to COVID-19, health care is entering a phase of rapid digital transformation… Trends that we expected to play out over three to five years are now happening in real time. Health care is riding a wave of innovation, where novel digital applications and business models are rapidly becoming mainstream — and this expansion will continue until digital adoption in health care nears equilibrium with the consumer sector.

Ironically, the most disruptive health crisis in recent history has created the optimal conditions for a revolution in how we use technology to care for one another.”

Radical Commentary: This Op-Ed by Sanjana Basu of Radical’s investment team, explains our health thesis which you can read part 1 here and part 2 here. We are now seeing a rapid acceleration of digitization across healthcare. This first wave will lay the groundwork to collect data and create opportunities for a second wave characterized by the application of AI solutions that can help deliver personalized and predictive healthcare. Radical is investing in companies delivering in both waves.

AI for Medical Imaging: Breakthrough shows that AI-accelerated MRIs are interchangeable with traditional MRIs  (Facebook AI)

“In a first-of-its-kind clinical study, the team of computer scientists and radiologists demonstrated that AI can generate equally accurate and detailed MRIs using about one-fourth of the raw data traditionally required for a full MRI. Since less data is required, MRI scans could run nearly 4x faster.”

Radical Commentary: This breakthrough by Facebook AI and radiologists at NYU could have a number of positive outcomes for the healthcare industry, including:

• Significantly less discomfort for patients. The traditional MRI scan process can take up to one hour, and patients need to be still for the entire duration. Patients who are in severe pain or are claustrophobic may even require sedation to get through the traditional process.
• Higher patient throughput per day, reducing wait times, and potentially allowing doctors to use MRIs in place of X-rays and CT scans for some cases. Note that MRI machines don’t use ionizing radiation.
• Faster diagnosis, allowing for a faster response from medical teams.

There are several start-ups focussed on using AI to automate diagnosis of medical images using anomaly detection, whereas this project is focussed on a different part of the workflow: creating medical images faster.

The team trained the AI model on low-resolution and high-resolution scans, providing the model with sufficient historical information to predict how a new scan will look with less data. Interestingly, radiologists could not distinguish which images were produced with AI and which came from the slower traditional scans.

The next step for the project is to prove that the technology works equally well for all use cases and with other types of MRI scanning machines. To speed up this process, the team has made this initiative an open source project, allowing for a faster path to FDA approvals through collaboration with other researchers and hardware vendors.

AI Powered Robots: I think… therefore I might be a material handling robot: interview with Ted Stinson  (DC Velocity)

“The Covariant ambition is to build a universal AI. We call it the Covariant Brain. It is meant to be essentially the cognitive system for a robot — the “brain” that gives it the ability to see and reason and act on the world around it. We chose to focus on warehouses initially because the logistics market offers such a great opportunity to deploy these capabilities to automate jobs that are tedious and, thus, hard to fill. Things like order picking in an e-commerce or grocery or apparel warehouse are great examples. These jobs are clearly repetitive in nature, but at the same time, every pick has a degree of variability and change, which is what makes warehouse environments so challenging.”

Radical Commentary: Ted Stinson, the Chief Operating Officer at Radical’s portfolio company, Covariant, offers a look at how AI-powered robotics platforms may reshape the future of supply chain management. Covariant was co-founded by renowned AI roboticist Pieter Abbeel out of his lab at University of California Berkeley, along with his PhD graduates, including CEO Peter Chen.

Covariant is building a universal AI platform capable of helping robots see, reason and interact with the world around them. Stinson discusses some of the technical hurdles Covariant has tackled to effectively scale the use of robots in real world environments, helping the company land partnerships with industrial robotics companies ABB and Knapp. As Stinson puts it, the benchmark for this technology is performance that’s on par with manual human processes.

AI and Drug Discovery: Broad Institute launches academic-industry cell imaging consortium to speed drug discovery and development  (Broad Institute)

“The Imaging Platform at the Broad Institute of MIT and Harvard, together with industry and non-profit partners, has launched a new collaboration to create a massive cell-imaging dataset, displaying more than 1 billion cells responding to over 140,000 small molecules and genetic perturbations.

This microscopy image dataset, which would represent the largest collection of cell images generated by Cell Painting, will act as a reference collection to potentially fuel efforts for discovering and developing new therapeutics.”

Radical Commentary: When developing a new drug, there is a large information gap between early stage simulations and how human biological systems will respond to a drug treatment. Comprehensive microscopy datasets provide researchers with a library of cellular mechanics which help researchers better understand the impact of drug compounds on cells. Automated systems powered by deep learning systems, such as the ones developed by the Broad Institute, enable this to happen at a scale previously not manageable by human researchers. The dataset assembled by this consortium addresses a bottleneck in the drug-discovery process and may unlock additional value in the biotech sector.

AI Strategy: How to Win with Machine Learning   (Harvard Business Review)

“Many companies are already working with AI and are aware of the practical steps for integrating it into their operations and leveraging its power. But as that proficiency grows, companies will need to consider a broader issue: How do you take advantage of machine learning to create a defensible moat around the business — to create something that competitors can’t easily imitate?

…moving early can often be a big plus, but it’s not the whole story. As we discuss, late adopters of the new technology can still advance — or at least recover some lost ground — by finding a niche.”

Radical commentary: Founders are frequently asked by investors to explain their technology moat. The authors clearly define three strategies for defending machine learning-based companies from imitation: 1) creating or accessing sufficient training data, 2) an ability to quickly generate and incorporate feedback data, and 3) prediction quality. Although big tech and other first-movers would seem to have an advantage over new entrants in these areas, differentiating the purpose and context for using machine learning can create tremendous value in a space that is defensible for new companies.

Companies looking to employ machine learning can benefit from untapped sources of training data, or new sources of feedback data that enable faster learning and better predictions. Additionally, the authors outline the benefit of specializing predictions for a particular niche such as hardware or a specific market segment to outperform competitors. As the authors of this piece conclude, many of the machine learning opportunities in the future will come from precisely engineered products that are highly adapted to specific purposes and contexts.

***
Editor’s Note: We will continue to use this platform to share without commentary articles focused on data and the use of it to illustrate and illuminate racial injustice. Because you cannot fix problems you cannot see or understand.

The economic impact of closing the racial wealth gap  (McKinsey)

“The persistent racial wealth gap in the US is a drag on Black Americans as well as the overall economy. New research quantifies the impact of closing the gap and identifies key sources of this socioeconomic inequity.”