WhatsTheBigData

Presiding over the ceremonial opening of the new IBM Watson Health global headquarters in Cambridge, Mass., IBM’s senior vice president Mike Rhodin highlighted the sometime-neglected focus of the effort to mine the ever-increasing quantities of health data. “We know that technology alone isn’t the answer,” said Rhodin. “At its core, Watson Health provides the means to orient the entire system around us.”

In a telephone conversation before the event, Dr. Lynda Chin, associate vice chancellor for health transformation at the University of Texas system, voiced a similar perspective: “Technology and innovation are the instigators for change, but they alone won’t do it. We have to think about implementation, about translating the technology into desired outcomes. Implementation is never just a technology play.”

Before assuming her current position in April, Dr. Chin was the founding chair of Genomic Medicine and scientific director of the Institute for Applied Cancer Science at The University of Texas MD Anderson Cancer Center. Two years ago, IBM and MD Anderson announced the Oncology Expert Advisor (OEA), based on IBM’s Watson data analytics engine, an expert system enabling clinicians to “uncover valuable insights from the cancer center’s rich patient and research databases.”

Dr. Chin reports that MD Anderson has by now developed two “apps,” each dealing with a different type of cancer, and is in the process of developing a third one, with each successive cancer-specific solution taking less time to develop. The ultimate goal is to make these solutions available to MD Anderson’s national and international network, so general oncologists in remote hospitals and clinics could tap into its accumulated and evolving expertise.  “To show that the OEA is a knowledge democratization tool, we have to build a network cloud infrastructure to support it. The OEA will not be useful if it doesn’t fit into the everyday life of the general oncologist.”

To achieve that goal, MD Anderson has also partnered with PwC for the development of the cloud information interchange and with AT&T for a secure, dedicated network. It is now piloting its first network link, to one of its network partners in New Jersey.

The integration with the general oncologist’s workflow is moving the expert system from a research reference resource and clinical decision support tool to helping manage the care of specific patients. “The OEA is trained to simulate the exchange between a physician and an expert,” says Dr. Chin. “So for the OEA to work, it has to be connected to the EHR system so we can learn about the patient. The OEA is trained not only to understand the profile of the patient in terms of what is the appropriate evidence-based treatment options but also sharing the experience in managing patients on that type of therapy and helping the general oncologist manage it. It’s as if the oncologist has the ability to call up the expert 24/7 to ask for advice.”

Still, one of the lessons learned so far is that “there will always be a question the OEA was not trained on,” so a teleconferencing component has been built into the system.  Other lessons include the need to provide mobile-device-based solutions, the challenge of teaching the OEA the relative value of each piece of information, and that the expert system “is very valuable from a learning perspective,” as a teaching tool for doctors in training. It also turned out that the OEA is useful in helping research nurses screen patients for clinical trials. Before, the nurses were often considering only the trials they knew about. Now they have at their disposal a clinical trial recommendation engine that screens through all the available trials and an expert system that helps with monitoring the patients participating in the clinical trial.

The development of the OEA is a never-ending journey. Healthcare is a complex and constantly changing endeavor involving research and practice, experiments and established procedures, professionals, institutions, and providers of all sorts, and most important of all, the people they serve—both patients and everybody else trying to keep themselves healthy.  Over the last decade healthcare has gone, at long last, through rapid digitization, transforming mounds of paper into electronic records and introducing computers to many aspects of the physician’s work.

As in other fields, the introduction of computer technology provides opportunities for reducing costs and increasing quality and effectiveness, while at the same time increasing the potential for errors caused by over-reliance on technology and automation. Similarly, while digitization facilitates the collection and sharing of practical knowledge and research expertise, it also produces mountains of data that threaten to impede rather than accelerate progress.

IBM Watson helps in processing and analyzing this data and presenting it as confidence-level-ranked suggestions and recommendations.  At the IBM event on September 10, Dr. Jeff Burns described how OPENPediatrics doesn’t tell the physician “do this,” but rather “tells the doctor what to think about.” OPENPediatrics is a Boston Children’s Hospital-led initiative bringing medical knowledge to pediatric caregivers worldwide (currently reaching 900 hospitals in 127 countries). IBM and Boston Children’s Hospital plan to develop “solutions for commercialization, initially pursuing applications in personalized medicine, heart health and critical care,” leveraging Watson’s genomic, image, and streaming analytics capabilities.

At the new Watson Health headquarters in Cambridge, Mass., Dr. Watson—and 700 other IBM employees—will be joining more than 600 Massachusetts-based life sciences companies and research organizations employing about 60,000 people. IBM plans to open there an interactive Watson Health Experience Center (a demonstration center for IBM customers) and establish a dedicated Health Research lab.

“We have to do it as a community,” Mike Rohdin declared at the event. Other participants, executives from Yale University, Sage Bionetworks, Medtronic, CVS Health, Modernizing Medicine and Teva Pharmaceuticals, echoed the sentiment. And Like Rhodin, they stressed putting people at the center of their efforts to improve healthcare, highlighting the specific goal of helping patients manage their disease.

Digitization—and smart analysis of the data it generates—helps in building a community around shared knowledge. The competition for profits and prestige among healthcare providers, however, while driving the innovation that may lead to better healthcare, also could stand in the way of cooperation and knowledge sharing. It may also lead to hasty development of technology-based solutions without a careful evaluation of the actual benefits and potential risks. Let’s hope that IBM and its partners will do everything they can to uphold the medical community’s tradition of controlled experiments and contributing to the ever-growing public repository of knowledge of what works and what doesn’t work in healthcare.

Originally published on Forbes.com

Phil Orr and Max Wessel, Sapphire Ventures:

..we believe three characteristics of VR will help it carve out its position in the world of computing. These characteristics hint at Neal Stephenson’s metaverse. Namely, they are presence, scale, and interaction design.

• Presence is an Oculus-coined term that effectively translates to a sense of being there. With presence, comes focus and attention. It’s impossible to ignore a piece of content when it’s your reality. It’s also proven that presence changes the way people internalize information and react emotionally to stimulus. That level of intellectual involvement will impact what we can do with software.
• Scale is something we all understand. It’s difficult to imagine what the interior of a house will look like from a diagram on the screen of an iPhone. It’s easy for an architect to help you imagine how the interior of a house might look when you can walk around inside of it.
• Interaction design changes fundamentally when your body becomes the controller. Today, when you use a computer for training or simulation, you interact with it using a mouse or a keyboard. It can inform you about different environments, but it does little to teach you how to execute tasks physically in the real world. VR changes that. When a system is designed to monitor movement on hundreds of different points on your body, it can help train you to do everything from perform surgical motions to throwing a football.

Etay Maor, IBM

Wired recently reported that hackers posted a “data dump, 9.7 gigabytes in size… to the dark web using an Onion address accessible only through the Tor browser.” The data included names, passwords, addresses, profile descriptions and several years of credit card data for 32 million users of Ashley Madison, a social network billing itself as the premier site for married individuals seeking partners for affairs.

“I want to show you the dark side of the net,” Etay Maor told me when we met last month at the IBM offices in Cambridge, Massachusetts. He then proceeded to give me a tour of the Internet’s underground, where cyber criminals and hackers exchange data, swap tips, and offer free and for-fee services. “Information sharing is a given on the dark side,” said Maor, “but for the good guys, it’s not that easy.”

Maor is a senior fraud prevention strategist at IBM and has watched the dark side of the Web at RSA, where he led the cyber threats research lab, and later at Trusteer, a cybersecurity startup which IBM has acquired in 2013 for a reported $1 billion. His focus is cybercrime intelligence, specifically malware—understanding how it is developed and the networks over which it is distributed. Maor is an expert on how cyber criminals think and act and shares his knowledge with IBM’s customers and also with the world at large by speaking at conferences and blogging at securityintelligence.com.

The Web is like an iceberg divided into three segments, each with its own cluster of hangouts for cyber criminals and their digital breadcrumbs. The tip of the iceberg is the “Clear Web” (also called the Surface Web), indexed by Google and other search engines. The very large body of the iceberg, submerged under the virtual water, is the “Deep Web”—anything on the Web that’s not accessible to the search engines (e.g., your bank account). Within the Deep Web lies the “Dark Web,” a region of the iceberg that is difficult to access and can be reached only via specialized networks.

Maor first demonstrated to me how much cybercrime-related information is available on the Clear Web. Simply by searching for spreadsheets with the word “password” in them, you can get the default password list for many types of devices and other things and places of interest to criminals. There is easily accessible information that may have been posted to the Web innocently or by mistake. But there is also a lot of compromised information (e.g., stolen email addresses and their passwords) available on legitimate websites that provide a Web location for dumping data.

Then there are forums for criminals, some masquerading as a benign “hacking community” or “security research forum,” promoting themselves like any other business and/or community, including a Facebook page, and covering their costs or even making some money by displaying ads. One such forum had 1,200 other people accessing it when Maor showed it to me, demonstrating how, with a few clicks of the mouse, you can find lists of stolen credit card numbers including all the requisite information about the card holder.

Maor proceeded to introduce me to Tor, the most popular specialized network providing anonymity for its users, including participants in the underground economy of the Dark Web.  It was developed in the 1990s with the purpose of protecting U.S. intelligence communications online by researchers at the US Naval Research Lab which released the code in 2004 under a free license. It has 2.5 million daily users, some with legitimate reasons to protect their identities, and others who are engaged in criminal activities.

Tor is based on Onion routing, where messages are encapsulated in layers of encryption. The encrypted data is transmitted through a series of network nodes called onion routers, each of which “peels” away a single layer, uncovering the data’s next destination. The sender remains anonymous because each intermediary knows only the location of the immediately preceding and following nodes. The final node in the chain, the “exit node,” decrypts the final layer and delivers the message to the recipient.

While Tor is used by people with legitimate reasons to hide their identity, it (and similar networks) also facilitates a thriving underground economy. This is where you can buy firearms, drugs, fake documents, prescription drugs or engage in pedophilia networks, human trafficking, and organ trafficking. Maor paraphrases Oscar Wilde: “Give a man a mask and he will show his true face.”

Tor is also home to rapidly growing “startups,” offering fraud-as-a-service. A decade ago, says Maor, cybercrime “was one-man operation.  Today, it’s teamwork.”  Furthermore, the whole process, from coding the malware to distributing it to working with money mules, can be easily outsourced.  Everything a cybercriminal might need is now available on the underground forums, some components of the process as a free download, others as a for-fee service, including cloud-based services with guaranteed service level agreements (SLAs). The menu of cybercrime options has grown beyond financial fraud tools, to include advanced targeting tools, Remote Access Tools (RATs), and health care and insurance fraud tools and services.

The explosion of data about us, our lives and our workplaces on the Clear Web has helped the denizens of the Dark Web circumvent traditional online defenses such as passwords. “Fifteen years ago,” says Maor, “it took a lot of work to breach a company. Today, I can go on Linkedin and find out exactly what is the structure of the company I’m interested in.” Knowledge of the reporting structure of a specific company helps criminals’ “social engineering” efforts, manipulating people into performing certain compromising actions or divulging confidential information. Once criminals get to know their targets (e.g., by connecting on Linkedin), the victims may open an email or attachment that will infect their computer and provide the desired access to the company’s IT infrastructure.

Cyber criminals are taking advantage of the abundance of data on the Web and its success at connecting and networking over 2 billion people around the world. 80% of cyber attacks are driven by highly organized crime rings in which data, tools and expertise are widely shared, according to a UN study on organized crime, generating $445 billion in illegal profits and brokering one billion-plus pieces of personally identifiable information annually.

Data and networking—aren’t they also great tools in the fight against cybercrime? Not so much. Corporations and security firms have been reluctant to share cybersecurity intelligence. Only 15% of respondents to a recent survey said that “participating in knowledge sharing” is a spending priority.

There have been some efforts to change that, such as the establishment of industry-specific Information Sharing and Analysis Centers (ISACs) and the cross-industry National Council of ISACs.  The Department of Homeland Security and other government agencies are working to promote specific, standardized message and communication formats to facilitate the sharing of cyber intelligence in real time. The Cybersecurity Information Sharing Act (CISA), a bill creating a framework for companies and federal agencies to coordinate against cyberattacks, is being debated in Congress.

Alejandro Mayorkas, the Deputy Secretary of Homeland Security recently said: “Today’s threats require the engagement of our entire society. This shared responsibility means that we have to work with each other in ways that are often new for the government and the private sector. This means that we also have to trust each other and share information.”

IBM has taken a big step towards greater engagement and information sharing when it launched in April the IBM X-Force Exchange. It is a threat intelligence sharing platform where registered users can mine IBM’s data to research security threats, aggregate cyber intelligence, and collaborate with their peers. IBM says the exchange has quickly grown to 7,500 registered users, identifying in real-time sophisticated cybercrime campaigns. “I’m a fan,” security guru Bruce Schneier responded when I asked him about X-Force Exchange.

“The security industry must share information, all the time, in real time,” says Maor. “It’s a change of mindset, but it has to be done if we want to have some sort of edge against the criminals.”

Originally published on Forbes.com