Dhruv Patel, 20, got addicted to reading health-based research papers at a very young age. These readings exposed the young student of biotechnology at the University of Maryland to the increasing number of brain injuries in the U.S. His curiosity continued to rise and it eventually led him to develop a technology to turn around the diagnosis of Alzheimer’s disease, a brain condition that slowly affects memory, thinking ability, and other cognitive functions. Mr Patel, who spoke about his innovation at the 10th edition of TEDx Gateway recently talks to The Hindu about his plans:
What drove you to focus on Alzheimer’s?
It all started with my research on brain trauma injury, particularly concussions, that is a huge problem in the U.S. We could not accomplish our research due to the complexities of the condition. We realised Alzheimer’s was equally grave, affecting 50 million people globally, where research has been stagnated. A family member had probable Alzheimer’s, and my co-founder, Christopher Look lost his great-grandmother. That drew our focus to Alzheimer’s.
How does this technology aid in Alzheimer’s diagnosis?
The device works by analysing a patient’s brainwave using a variety of mathematical analysis and compares it to brainwaves of people who don’t have Alzheimer’s. We have combined portable electroencephalogram (EEG) with machine learning and artificial intelligence (AI). Using machine-learning characterisation, the algorithm can decide, to a certain level of confidence whether the new brainwave belongs to an Alzheimer’s patient or a healthy patient.
The idea is to find subtle differences in people who have Alzheimer’s and those who don’t and are normally ageing, with the help of this technology. The intention is also to decode and then study those differences.
Once the analysis is completed, the results are provided to the physician instantly. Physicians at present do a cognitive assessment, use questionnaire method, order MRIs, Cerebral Spinal Fluid (CSF) test. The problem is that these are not accessible and cannot be used to make AD screening routine. Plus, there aren’t enough doctors and money for this. This portable technology makes it easier. The headset device can be connected to any laptop at the physician’s office and used.
At what stage are you currently at? When will this technology be available?
We are in clinical development stages at present, looking to conduct a clinical pilot study to gather more data to improve the accuracy of the algorithm. Once accomplished, we will submit it to the regulatory agencies. It may take us another two to four years to commercialise the technology. We are still experimenting with algorithms. At present, we have 85% accuracy with the technology, but we are still in the R&D phase. We plan to conduct a large-scale clinical study next year and plan to market the technology in the next two-three years.
What resources went into developing the technology?
We had to read a lot of medical and engineering research papers. We were a team of five engineers. We read these research papers carefully to comprehend information. We also used software and hardware, such as Open BCI (an open-source brain-computer interface platform). We’ll be using other software like Memory MD, Matlab and Python.
What role do Electroencephalogram (EEG), machine learning and Artificial Intelligence (AI) play in this technology?
By combining portable Electroencephalogram (EEG) with machine learning and artificial intelligence (AI), the idea is to find subtle differences in people who have Alzheimer’s and those who are normally ageing. We also try to decode and then study these differences. The electrical signals released from the brain are analysed using computer algorithms which can detect differences between those with Alzheimer’s and those without Alzheimer’s. At present, it works with a laptop or a computer. In the future, we are looking to pair it up with mobile phones as well.
How long did it take for you to develop this technology?
Well, with the amount of research, reading and to invent our own, it took us two years to develop this technology.
Is the use of this technology only restricted to Alzheimer’s or can it also be used for other diseases?
The dataset only represents Alzheimer’s currently. However, we have created a platform through which diagnostics can generally be provided for other neurodegenerative diseases also.
Apart from early diagnosis, what are the other benefits of this technology?
Usually, patients with Alzheimer’s are diagnosed late. Early diagnosis gives patients more time to plan financially, and legally, and they also have the ability to enter into clinical trials for novel therapeutics, thus giving some hope, while at the same time bringing the rest of the world closer to a cure.
If brought into the market, how cost-effective will this be?
The device is less expensive than MRIs and PET scans. The data acquisition does not cost anything at all. I think it is with EEG, the combination of software and low-cost hardware, that Alzheimer’s diagnosis becomes more accessible.
