Oakland University looks to the cloud to detect early signs of Alzheimer's disease

Dr. Erik Fredericks at Oakland University in Rochester, Michigan is using the Google Cloud Natural Language API to prototype analysis for speech changes signaling the progression of Alzheimer’s disease. He sees the potential for cloud computing in higher education and research as “limitless” — and puts Google Cloud (Google Cloud) tools at the core of his work.

Dr. Erik Fredericks is an Assistant Professor in Oakland University’s Computer Science and Engineering Department. After attending a Google Faculty Institute, he realized the potential of cloud computing to power a vast range of applications that are revolutionizing information technology, and that his students needed to master these tools. Since 2016, he’s been using Google Cloud (Google Cloud) APIs in both his classroom teaching and research work.

Providing Insights to Support Alzheimer's Patients

Dr. Fredericks is using Google Cloud APIs to develop tools that may enable earlier diagnosis for people with Alzheimer's. “We are in the initial stages of prototyping a smart home for medical purposes to monitor and analyze speech changes, which can be early indicators of Alzheimer's disease,” he says. He is partnering with Dr. Katey Price from Albion College in Michigan, who has a PhD in health communications and has worked with Alzheimer's patients.

The pilot is using the Google Cloud Natural Language API, which reveals the structure and meaning of text with a pre-trained machine learning model, and an Android voice recording app developed by Dr. Fredericks' lab, to collect, and analyze speech data from subjects in their homes. "The thought is with daily monitoring, and with recording and analyzing conversations with patients over time, we may be able to pick up early signs of Alzheimer’s disease," he says. "We may be able to improve patients’ quality of life if we can detect symptoms earlier."

Graduate students Kate Bowers, Reihaneh Hariri, and Seth Jones are assisting with developing the pilot study, including researching wearable devices for patient state detection. Undergraduate student Nicholas Kerns developed a prototype virtual reality application for training Alzheimer's caregivers.

Dr. Fredericks sees cloud computing's capacity to fast-track advances in medicine and other fields. He says a research project of this potential scale and scope would normally need to be carried out on premises with costly, high-powered computers — but Google Cloud manages that work remotely. "We're using Google as a heavy processing engine to do computations that otherwise would not be feasible,” he says. “With Google Cloud, the processing capabilities are nearly limitless."

"We're using Google as a heavy processing engine to do computations that otherwise would not be feasible. With Google Cloud, the processing capabilities are nearly limitless."

Rocketing Students into the Cloud

Dr. Fredericks initially chose the Google Cloud for its ease of use. "It's been a wonderful addition to my classes," he says, pointing out the real-world applications for computer science majors preparing for careers in business, science, technology, education, research, and other fields.

His undergraduate system administration students are learning the skills necessary to analyze, deploy, manage, and troubleshoot enterprise computing infrastructures. Students develop these skills in Google Cloud using Google Compute Engine, which lets them access virtual machines running in Google's data centers and worldwide fiber network, and Google BigQuery, a cloud data warehouse for analytics with built-in machine learning capabilities. BigQuery includes BigQuery ML which allows data scientists and analysts to build and operationalize machine learning models on structured or semi-structured data, directly inside BigQuery, using simple SQL and at a fraction of the cost and time compared to traditional ML systems. Google Cloud serves as the platform for his web design course as well.

Dr. Fredericks also teaches Software Engineering for Cyber-Physical Systems in the nationally award-winning Oakland University Center for Cybersecurity. Students work with a variety of Google Cloud tools including Google Cloud BigQuery, Cloud Speech to Text, and VisionAPI. The speed at which students can access and analyze information - makes learning about cyber-physical systems even more engaging and exciting.

"They used their Raspberry Pis [a credit-card sized computer on which students learn to code] to sift through the GitHub repository in seconds," Dr. Fredericks says. "They are amazed that their tiny little microcomputer can handle that using Google Cloud tools." By bringing Google Cloud to the classroom, faculty are less reliant on IT for maintenance and security, and students can experiment and stretch their skills without worrying about breaking computers on premises.

That's one reason why he puts Google Cloud at the core of his university classes, so his students can learn the skills needed to master the tools and explore their own possibilities.

Dr. Fredericks participates in the Google Cloud Education Credits program, which provides faculty and student credits to use toward Google Cloud tools. That’s a huge plus for cash-strapped students. "They really like not having to use a credit card," he says. "The credits are free and if they run out, I can request more. It’s not the end of the semester for them."

When it comes to using cloud computing for teaching and research, Dr. Fredericks believes the sky’s the limit. "It's easy, the students like it, and there's a lot more I could be doing with it," he says. "We look forward to expanding our use of Google Cloud tools in the future."