Every time a doctor updates a patient record, a laboratory releases a test report or a new research paper gets published, the tsunami of healthcare data rises to new heights. At an estimated rate of 36% a year, healthcare data is growing even faster than media, manufacturing and financial data—and so is the need to use this vital information to improve healthcare quality and control costs.
However, two formidable barriers stand in the way of unlocking the full potential of Big Healthcare Data. First, most healthcare data—such as free-form physician notes—is unstructured, making it difficult to analyze and share electronic health records. In addition, the US Health Insurance Portability and Accountability Act (HIPAA) strictly regulates access to protected health information and penalizes breaches of patient privacy.
“When it comes to healthcare data, it’s a very interesting challenge,” says Dr. Karuna Joshi, Associate Professor of Information Systems at the University of Maryland, Baltimore County (UMBC) and UMBC Director of the Center for Accelerated Real Time Analytics. “Most of it is textual and very sensitive data that cannot be shared freely. However, this huge amount of data needs to be shared, especially with the COVID-19 situation. So how do you analyze the data so that privacy and security are not violated?”
Automation
Provided cloud services for a groundbreaking healthcare research study to automate HIPAA regulations and expedite access to encrypted data
Solutions for students
Will provide an education solution on the cloud to help professors instruct computer science and electrical engineering students
The data dilemma is further amplified by the sheer volume of healthcare data that must be accessed. “We have billions of health-related records and we need systems to process this information automatically because it’s too time consuming for humans to do manually,” says Dr. K. Joshi.
Dr. K. Joshi and her data analytics team at UMBC saw a pressing need for research tools that could accelerate and automate access to Big Healthcare Data, despite the hurdles of unstructured data and privacy rules. But to support this complex research, they needed computing resources beyond the capabilities of typical laptop, desktop and IT legacy systems. “I was introduced to cloud technology and virtual computing through IBM,” says Dr. K. Joshi, “and it was a natural progression to collaborate with IBM Cloud for Education.”
When Dr. K. Joshi and her research team started investigating ways to analyze highly regulated and encrypted healthcare data, one of the key questions revolved around encryption. “You have to understand and make decisions quickly in the healthcare domain, but if you start decrypting the huge volumes of data involved, it will take you days to decrypt,” says Dr. K. Joshi. “So how do you let the data stay encrypted but at the same time let the algorithms and access control mechanisms work? Because when you have this huge amount of encrypted data and you want to query it without decrypting, it’s going to be a computational nightmare.”
By developing new approaches using encryption and semantic reasoning, Dr. K. Joshi’s research team was able to take small snapshots of the data and decrypt small sections of the record to pull out the information required without devoting significant time for decryption. “Something innovative that we are trying to do here is use AI approaches and deep learning to access the information without decrypting the data,” says Dr. K. Joshi.
Using the IBM Cloud® for Education tool to refine this strategy, the UMBC team recently published two important research papers on this topic, including, “A Semantically Rich Knowledge Graph to Automate HIPAA Regulations for Cloud Health IT Services” for the Big Data Security 2022 Conference and “Delegated Access Control using Attribute-Based Encryption,” sponsored by the Office of Naval Research.
“Our semantic reasoning and encryption approaches have been very well received by the research community,” says Dr. K. Joshi. “Companies have policies and organizations have rules about how the data should be encrypted and sent across the internet. These rules cannot be automated now and our goal is to automate these policies.”
She continues: “The cloud is a great solution for this work because it is available at any time. Everyone can share it easily. Because the computational requirements for our research will be huge, the cloud will be very important in our work, and the research community is excited about the idea, as well as the healthcare community we interface with.”
Dr. Anupam Joshi, is the Oros Family Professor and chair of the Computer Science and Electrical Engineering (CSEE) department at UMBC. For years, he has used IBM Cloud for Education to support his research projects.
“For example, I am currently doing a project with the US and Japan to understand social isolation in older populations,” says Dr. A. Joshi. “This study relies on audio data gathered from conversations, and the IBM Cloud analysis tools do a great job of giving us the tone and sentiment of conversations, while preserving privacy.”
Dr. A. Joshi also advises graduate students that use IBM Cloud for research projects and thesis development. “The feedback from my students is uniformly positive,” he says. “For example, when one of my students ran into a problem, his issue was fixed in an hour. Graduate students are on a timeline and they are deeply appreciative that a company the size of IBM turns around requests so quickly.”
Based on his research experience with IBM Cloud for Education, Dr. A. Joshi chose the Education Solutions on IBM Cloud platform to help teach the high-level UMBC computer engineering course on operating systems. “Outside of algorithms, operating systems is one of the hardest courses to get through in a good computer science curriculum,” says Dr. A. Joshi. “To perform the tasks this course demanded, we are transitioning this course to IBM Cloud to take advantage of the cloud’s scalability.”
While taking the course, students are required to make changes to the Linux® kernel and reboot the operating system with their own kernel. “Compiling a Linux kernel can take a lot of time on your own machine and you need to set up a ‘big box,’” says Dr. A. Joshi. “Because of the elastic demand, we can spin up the virtual machines on the cloud throughout the semester when projects are due.”
UMBC students will also have opportunities to work on virtual machines in the IBM Cloud, where they can experiment, manage their work and gain VM experience in preparation for computer engineering jobs.
Cloud-supported computer education addresses the fact that many students are from underserved populations and can’t afford powerful laptop computers. “UMBC’s message has been ‘inclusive excellence.’ We take all comers and try very hard to give them the means to succeed,” says Dr. A. Joshi. “Instead of students having to buy expensive, state-of-the-art machines, we have the resources on IBM Cloud to give them access to hardware and software that they may not have on their own machines.”
In addition to IBM Cloud for Education and Education Solutions on IBM Cloud, other IBM solutions and services contributing to UMBC’s research and teaching include:
- IBM Cloud Bare Metal Servers, which run IBM Cloud for Education workloads on fully dedicated servers for maximum performance and security-rich single tenancy.
- IBM Cloud Satellite® Infrastructure Service, a distributed cloud solution that deploys and runs applications on public clouds, on-premises data centers and edge computing systems.
- IBM Cloud App ID to add authentication, secure back ends and APIs, and manage user-specific data for mobile and web applications.
- IBM Cloud Object Storage with flexible, cost-effective and scalable cloud storage in remote locations for unstructured data such as healthcare records and imaging data.
- IBM Db2® on Cloud, a managed SQL cloud database with AI capabilities that offers a dedicated operations team, point-in-time recovery and independent scaling for enterprise applications.
- Red Hat® OpenShift® on IBM Cloud to automate updates, scaling and provisioning with the resiliency to handle unplanned surges.
IBM Cloud for Education is the culmination of years of collaboration between IBM and colleges and universities across the Americas and Europe. “IBM Cloud for Education is based on a solution IBM developed almost two decades ago called the Virtual Computing Lab (VCL). North Carolina State was the lead institution and UMBC participated early on,” says Dr. A. Joshi. “It was one of the first cloud solutions developed specifically for education at a time when the term ‘cloud computing’ hadn’t yet caught on.”
“IBM provided me with access to the VCL and that’s how I was introduced to cloud technology and virtual machines,” says Dr. K. Joshi. “IBM has also been very receptive to collaborating with researchers and exchanging information and ideas. If I have an idea, I know I can find someone who can look into it.”
As an example, she remembers when an IBM lawyer came to UMBC for a talk and later met with her to discuss some of the legal challenges related to data compliance. “The lawyer spent half an hour with me and that is very rare when you have the counsel of a big company discussing research ideas. That’s another reason why our relationship with IBM Cloud works.”
In the fall of 2022, Dr. K. Joshi will be teaching senior-level data analytics and anticipates using resources from IBM Cloud for Education to help her students. “I have been looking into providing my students with access to IBM Watson analytics tools. They would like to learn the software so they can be experts when they graduate, and that is one of the collaborations I am hoping to set up.”
In addition to supporting academic research and classroom teaching, IBM Cloud for Education has also been developed as a general-purpose cloud platform. It can support the business side of universities and an applications lab residing within the platform can integrate with other IBM services of interest to faculty members including AI, blockchain, IoT and other solutions.
IBM Cloud for Education is also supported by the IBM Center for Advanced Studies (CAST), which manages university relations for one of the largest IBM development communities, including UMBC and many other North American institutions of higher learning.
“We think IBM Cloud for Education is the right way for the future, especially for specialized courses that need special hardware and software combinations such as computer engineering and computer security,” says Dr. A. Joshi. “We will continue exploring the possibilities and tapping the enthusiasm we have for this computing resource.”
UMBC (link resides outside of ibm.com) is a public research university known for innovative teaching, high-impact research, and a supportive community that empowers and inspires inquisitive minds. UMBC serves 14,000 undergraduate and graduate students and is one of the country’s most inclusive education communities. US News & World Report ranks UMBC #6 nationally in both undergraduate teaching and innovation, and NSF ranks UMBC among the nation’s top universities in federal research support, particularly in NASA and social sciences funding. The university contributes to economic development through government and industry partnerships advancing entrepreneurship, workforce training, K–16 education, and tech commercialization.
Legal
© Copyright IBM Corporation 2022. IBM Corporation, Hybrid Cloud, New Orchard Road, Armonk, NY 10504
Produced in the United States of America, January 2022.
IBM, the IBM logo, ibm.com, Db2, IBM Cloud, IBM Cloud Satellite, and IBM Watson are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the web at ibm.com/legal/copyright-trademark.
Red Hat® and OpenShift® are registered trademarks of Red Hat, Inc. or its subsidiaries in the United States and other countries.
The registered trademark Linux® is used pursuant to a sublicense from the Linux Foundation, the exclusive licensee of Linus Torvalds, owner of the mark on a worldwide basis.
This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates.
The performance data and client examples cited are presented for illustrative purposes only. Actual performance results may vary depending on specific configurations and operating conditions. THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided.