91.1 F
San Fernando
Thursday, Mar 28, 2024

Diagnosis by Data

Beyond Limits Inc.’s artificial intelligence once gave NASA’s Mars Rover the ability to navigate Martian terrain. But the Glendale-based company sees many more applications for its software here on Earth – including solving some of the biggest challenges in healthcare. “This technology has been proven in very difficult environments in space,” Kevin Taylor, Beyond Limits’ new vice president of business development for healthcare, told the Business Journal. The software was developed at NASA’s Jet Propulsion Laboratory in La Canada-Flintridge. Caltech, which runs JPL, licenses the technology to Beyond Limits through a revenue-share partnership. The software is already used by clients in finance, energy and logistics. But Beyond Limits sees its next great potential in healthcare. Taylor was hired in March to oversee the expansion of the firm’s clients in the medical field, which he anticipates will include pharmaceutical companies and hospital systems. “Because the healthcare market is so large, you have these massive segments that have some really daunting problems,” Taylor said. “With the power of artificial intelligence, you can strip out costs and create efficiencies.” The market for artificial intelligence in healthcare is expected to be around $23 billion by the mid-2020s, according to a report by Allied Analytics, a full-service market research company based in Portland, Ore. Firms of all sizes are eager to capitalize on that growth; Amazon.com Inc., Apple Inc. and Google parent company Alphabet Inc. are all reportedly building out a strategy for entering the sector, while startups focused on AI applications in healthcare have raised more than $2.1 billion in venture capital, according to CB Insights, a technology market research company. But Beyond Limits’ technology could give it an edge over the competition, as it is able to generate not only insights but an explanation of how it came to find them. “From that perspective, (Beyond Limits) has a lot of potential to be very appealing to physicians,” Jennifer Bendfeldt, a manager at ECG Management Consultants Inc. in San Francisco, said. “They’re onto something with the cognition.” Beyond AI Beyond Limits’ technology complements “symbolic” AI, which mimics human reasoning, with what Taylor described as “numeric” AI. Numeric AI uses machine learning to solve narrowly-defined problems in clean sets of data, while symbolic AI can generate insights from “messier” data by processing information in a manner similar to that of the human brain. “It represents how we as humans think,” Taylor said. “How do we reason? What rules do we bring to the table or to a decision that allow us to generate hypotheses and make complex decisions, especially when the data is messy or there’s ambiguity?” Those strengths are well-suited for solving problems in a clinical setting, where physicians must make decisions about patient care based on a many different factors. Additionally, Beyond Limits’ AI software has the ability to show how it came to its conclusions, evading the “black box” issue that is a major downside to other types of artificial intelligence in medicine. “Physicians want to understand why the output is what it is,” Bendfeldt said. Beyond Limits’ technology could give them better intel as to what the software is thinking than other types of AI software, as it mimics human thought processes in terms of data recognition. The technology also could have significant utility in precision medicine, as it will be able to sift through a patient’s genetic information, laboratory results, disease progression history and diagnostic images to create a comprehensive patient profile that enables the creation of targeted treatments. “All of that can get brought together in a way that no single clinician could do on their own – it’s far too sophisticated and complex,” Taylor said. “But we can do that with our tools, which then can help a clinician make the best treatment decision possible for a patient given the unique qualities of that patient.” Market resistance AI’s utility does not preclude it from resistance to adoption. In a clinical setting, already time-crunched physicians might be hesitant to spend more time learning or using new technology if it requires that must cut time with patients to do so, noted Bendfeldt. Doctors over the last decade have been burned by the frustrations of implementing electronic health records, which provided more data but made their work more burdensome. “The trick is to find a way to seamlessly work intel into existing workflows,” Bendfeldt said. “Some (physicians) are a little leery of trusting systems to make their lives easier when they haven’t experienced that in the past.” Even if physicians do embrace the technology, training them on it will still be an expensive endeavor, she added. Integrating it into existing technical systems poses a daunting hurdle as well, noted William Studebaker, president of ROBO Global, the San Francisco-based creator of an exchange-traded fund for robotics, AI and automation companies. “When people talk about AI taking over the world, I kind of laugh,” Studebaker said. “We really don’t have the integration skills to do that yet. It’s hard enough to integrate a new payroll system or database in your company.” Beyond Limits has an advantage on this front, Chief Marketing Officer Dann Wilkens said. The company builds custom software solutions to address a company’s specific problems as well as to complement its existing IT infrastructure, he said. “For some AI systems, the first thing you have to do is buy a supercomputer,” Wilkens said. “Our technology is designed to work on systems that can fit in a space capsule. It has to be economical in a computing sense.” Ultimately, companies in every sector – healthcare included – will be forced to adopt the new technology as it grows more ubiquitous, Studebaker believes. “Robotics and AI are must-have technologies,” he said. “Either you’re using them to power your business or you’re going out of business. The only way to survive is to use AI to crunch data.” Next steps Beyond Limits closed a $20 million Series B funding round last June, led by BP Ventures, the investment arm of oil and gas giant BP plc. It plans to expand its workforce from 52 to 140 by the end of the year, and intends to do most of its hiring locally, Wilkens said. “We aren’t a Silicon Valley company – we’re a Southern California company,” he said. “We like to say we have one foot in Caltech and the other in Hollywood.” As for its healthcare business, Taylor sees potential clients in every segment of the industry, from insurance to providers to logistics. The rising costs associated with an aging population and policy reform demand sweeping change across the board, he said. “Healthcare is a wildly complex industry that has a lot of needs right now,” Taylor said. “The costs just continue to grow. … All of these things come together to put tremendous pressure on each of the different players within this industry to change how they behave.” Disorganization in every sector of the healthcare system – and hospitals in particular – present a massive opportunity for technology that can improve the way systems function, ROBO Global’s Studebaker said. “The business models are going to change because there’s so much waste in healthcare – everything in a hospital is inefficient,” he said. That shift will not be incremental adjustments in the context of healthcare companies’ current business model, he asserted. Rather, the industry will undergo a data-driven transformation, one that he believes will require the kind of technology Beyond Limits offers. “It will take time,” he said. “But the fact that we’re bringing in the cognitive reasoning piece to these complex problems is going to be transformational.”

Featured Articles

Related Articles