by Rutesh Shah

Savvy enterprises can mimic success of cognitive AI in healthcare

Nov 18, 2020
AnalyticsArtificial IntelligenceHealthcare Industry

Cognitive technologies are being deployed to solve the world’s biggest challenges. Here’s how enterprises can put cognitive AI to work.

virtual brain / digital mind / artificial intelligence / machine learning / neural network
Credit: MetamorWorks / Getty Images

Businesses across all sectors are taking a keen interest in the potential of artificial intelligence (AI) to address its most pressing challenges. AI is already known for its ability to speed up processes, streamline operations, and of course to crunch vast quantities of data faster than a human ever could. But when it comes to systems that can think for themselves? This reality is closer than you might think.

Cognitive AI assimilates data from multiple sources, in different formats, and is able to weigh up these data to form insights. This type of AI differs from others in its ability to mimic the way the human brain works. Cognitive AI systems are interactive, contextual, and, crucially, adaptive, in that they learn and evolve dynamically as new information comes to light. Far from replacing humans, AI is being taught to work alongside humans, helping to enhance the work we do or fulfill needs in other ways. Early adopters of cognitive technologies view them as critical to the future success of their organization and the ability to digitally evolve.

To understand the business implications of cognitive AI, look no further than the healthcare industry. Researchers are using cognitive AI to analyze blood samples, metabolism, speech and language patterns, and handwriting to understand risk factors associated with Alzheimer’s disease, resulting in programs that can diagnose the disease six years earlier than previously possible. A federation of 30 business, healthcare, and research institutions are developing cognitive AI models that identify brain tumors. And startups like MyndYou are using the brain as an AI sensor to help care providers assess and monitor elderly patients with a platform that monitors speech, walking, and driving over time to identify changes that indicate deteriorating physical or cognitive abilities. With these science-fiction like use cases, it’s little wonder that IDC predicts worldwide spending on cognitive and AI systems will reach $77.6B in 2022.

It is of course early days for cognitive AI in healthcare and, in time, history will judge its long-term success. Nonetheless, the early indications are hopeful and positive, to the degree that we can already extract some early lessons that other sectors can learn from. By examining the drivers and enablers for cognitive AI in healthcare, particularly at a research level, enterprises in other markets can identify ways to grow and improve their own business processes.

The need and availability of vast data sets

Without massive data sets, there is no place for AI, never mind cognitive AI – but readily available databases or spreadsheets are enough to get started. In the healthcare world, medical data is already vast and is growing at 36 percent CAGR through 2025, owing to advances in wearables and other IoT-enabled devices, medical imaging and real-time data production. One of the biggest enablers for cognitive AI in healthcare has simply been the sheer volume of data that are generated. Enabling connected systems to have access to this aggregated, anonymized data that already exists about patients allows cognitive AI to spot health trends and patterns, especially when combined with real-time health monitoring information (such as from wearables) and environmental data.

Assimilating disparate data, extracting insights, and turning them into actionable intelligence is the common digital challenge affecting all sectors. In the insurance industry, for instance, we are already seeing cognitive AI used to gather vast amounts of structured and unstructured data to improve the accuracy of underwriting, remotely process claims, streamline operations, and drive down costs. In the future, thanks to cognitive AI, the data streams we generate could be shared with insurance companies to automatically adapt premiums according to the choices we make and handle insurance claims in real-time based on events that take place.

Context is king

Improving customer experience is one of the biggest drivers for the adoption of cognitive technologies. We are all familiar with retail experiences that have been (mis)guided by algorithms. We especially notice when it misfires, for instance when we are shown ads for something we have just purchased and are unlikely to buy again. (I only need one washing machine, thank you.)

By contrast, cognitive AI is a much subtler tool. Take as an example Navigating Cancer, the world’s first patient-centered platform for oncology, which supports more than one million patients and thousands of cancer care providers to drive better outcomes through cognitive analytics. Intelligent analytics is helping improve assessment of “at risk” patients, drive down costs for patients, and speed up the overall process.

There is clear demand for cognitive AI in retail too, such as when intelligent agents offer 24/7 customer service support. Cognitive technologies are responsible for hyper-personalized, omni-channel customer experience, which is still more aspiration than reality for the majority of retailers. It can also help retailers make savvy business decisions in hyperlocal contexts. For example, by predicting which stores will struggle and taking remedial actions earlier, or by contrast, spotting spikes in demand and being able to deploy people and goods to satisfy that demand.

How to work faster, quicker and smarter

The final major driver for cognitive AI in healthcare – as in other industries – is the pressure to automate where possible: streamline operations, drive down costs, and boost productivity. In healthcare revenue lifecycle management, many processes are still manual and largely unautomated. A system underpinned by cognitive AI, however, could streamline the diagnostics made by doctors in real-time as the doctor types their notes, with the rest of the required back-end processes including billing, which currently occupies time that could be spent attending to patients. Or consider how hospital pharmacies currently serve a large hospital. Today, prescriptions are routed by AI engine to the nearest pharmacy, regardless of its capability to process the request in a timely fashion, which depends on staffing, capacity, inventory levels and how busy the rest of the hospital is.

Other industries are taking advantage of the winning combination cognitive AI offers: big data analytics for the “heavy lifting” involved in sorting through vast amounts of data, paired with the deep learning and neural networking technologies that apply human-like cognitive function and learn as they go. For example, in banking, Eigen Technologies uses natural language processing to extract relevant meaning from documents, eliminating repetitive, manual document processing. Its cognitive platform recognizes nuances in meaning, context and the idiosyncrasies of human language.

Data protection and anonymization

Protecting health data is paramount. Understandably, security concerns topped the agenda when it came to how cloud computing and big data analytics could be used. Enterprises can also learn a lot from the health sector by examining how it overcame barriers to cognitive AI adoption. The federated, anonymized nature of cognitive computing makes it particularly adept at preserving data privacy, navigating laws around data protection and any industry specific regulations that need to be adhered to.

Among the most highly regulated industries, healthcare’s adoption of cognitive technologies has been spurred on by advances in cloud computing services, such as the growth of HIPAA and GDPR-compliant platforms. Similarly, in industries such as financial services, cognitive AI is being deployed to sift through financial data, resulting in a  dramatic increase of the speed and accuracy of anti-money laundering (AML) investigations or identifying fraudulent credit card transactions in real-time . Other highly regulated sectors are wising up to the benefits of cognitive technologies in improving system performance, risk reduction, and security enhancement.

Ironically, it is the healthcare market itself that could perhaps best emulate the early success of cognitive AI. Outside of research facilities and closer to patients, cognitive AI is still relatively untapped. Over the past few months, in reaction to the pandemic, patient care outside of COVID-19 transitioned to telehealth almost overnight. The geographical distances between the doctor and patient grew, meaning that local context became blurred or was lost completely. In the future, this gap could be filled by cognitive AI systems, as a support to medical practitioners.

For instance, if a number of patients start to present with respiratory symptoms, a doctor would examine the patient’s notes and consider causes such as asthma, allergies, or any number of other illnesses, including of course the illness on everyone’s mind, COVID-19. In this situation, an AI system would be able to parse together information from other sources to highlight likely causes. For example, is there a forest fire in the region that could be contributing to an uplift in respiratory cases? Is the pollen count particularly high in the area? Or could this be linked to infectious disease transmission? AI has the potential to help answer these and many more questions to enable healthcare providers deliver more nuanced, personalized and informed opinions.

Ethical use of AI technologies

While AI and robotics can help in many ways and get enterprises excited with its benefits, it is important to think rationally about the use and impact of AI technologies. Enterprises need to have ethics built into the idea of why a certain piece of technology, equipped with AI, is being developed. It is important to review the outcomes of AI technology in order to fully understand its behavior and make sure that it’s not violating our (human) moral compass or at minimum the value of your enterprise. Remember, just because we can, doesn’t mean we should.

We are on the cusp of the “cognitive era.” It’s difficult to over-emphasize the strategic importance of cognitive systems for businesses. Early adopters of cognitive technologies routinely cite not only rapid ROI but cognitive’s importance to their company’s strategic vision and competitiveness. Ultimately, cognitive AI is helping to solve the world’s toughest challenges and in so doing, is reimagining the digital age as we know it. Before long, there won’t be a business anywhere that escapes disruption, positive or negative, related to cognitive computing.  

AI is evolving faster than you think. Now is the right time to plan, experiment with, and adopt AI to realize its full potential.