Imagine that after a phone call with your mother, you learn that she is concerned about early signs of Alzheimer’s disease.
She tells you that she will soon take part in an assessment, where she will wear an IoT sensor for a few weeks that listens any time she speaks and sends the data to a doctor. From that data, the doctor will be able to tell her whether it’s likely or not that she has Alzheimer’s — all by using AI.
This speculative scenario is not far from what several researchers at Anhui University of Science and Technology and Chuzhou University in China believe could happen as a result of their experimentation. The team used the data to pinpoint spectrogram features, helping identify Alzheimer’s disease with the assistance of AI. The researchers noted in their report that using speech data can reduce medical costs and “can be collected in a non-invasive manner so that the patient’s data can be collected in real-time and accurately.”
Most research regarding the use of AI to diagnose Alzheimer’s disease centers on imaging recognition in patient scans — like the study in India that developed a deep learning technique with magnetic resonance imaging and PET scans from Alzheimer’s patients.
AI’s capabilities to analyze and draw insights from healthcare data is progressing every day and is not limited to Alzheimer’s. Four machine learning algorithms — decision tree, support vector machine, ensemble classifier and Bayes’ classifier — were used in India to predict the severity of Parkinson’s disease. Research assessments published in the National Library of Medicine even found that AI can predict the periods when a patient might die due to heart disease.
New York City tech companies are at the forefront of AI and healthcare research. Built In spoke with a tech leader at Merck who is finding novel ways to use AI and machine learning in the field of disease biology.
Merck is a global biopharmaceutical company that has been inventing for more than a century.
What excites you about the work you’re doing?
What excites me about the work I’m doing is the potential impact it can have on patients and people. We are focused on using AI/ML approaches and methodologies to gain a better understanding of disease biology and uncover the underlying mechanisms of diseases. By achieving this, we can identify targets that can intervene in disease states and help move them toward a healthier state.
What’s the coolest project you’ve worked on as a data scientist focused on disease biology?
The coolest projects I have worked on as a data scientist at my company are the new approaches and methodologies we are using to discover what drives Alzheimer’s disease. This project stands out to me because it is incredibly motivating to work on something that has the potential to impact many people, and it is personally rewarding since my mom suffered through Alzheimer’s disease.
What makes this project even more exciting is that we have the opportunity to validate our models in the wet lab. This means that we can test the predictions and insights generated by our AI/ML approaches in real-world experiments.
“This means that we can test the predictions and insights generated by our AI/ML approaches in real-world experiments.”
This feedback loop between data analysis, AI/ML and experimental validation allows us to learn what we understand well and where we need more data or better approaches. It gives me hope that we are making progress and will ultimately gain a better understanding of the disease, leading to improved treatments or interventions.
Why is AI exciting for your team in this context?
One of the reasons why this work is so exciting is that we are fortunate to be living in a time of exceptional growth in new data modalities and exponential growth in both data volume for studying molecular and cellular biology and medical data. This increase in data is coupled with novel AI/ML methodologies and the exponential increase in compute power, creating a perfect alignment of data, AI/ML methodologies and compute capacity.
The combination of these factors allows us to tackle complex problems and make advancements in understanding diseases and finding potential drug targets. It’s an incredibly exciting time.
What do you love most about the culture on your team?
The strong emphasis on collaboration and the opportunity to work with people from diverse backgrounds and training. Our team is highly interdisciplinary, with individuals who bring unique perspectives and expertise to the table. This creates an environment where everyone’s contributions are valued and respected.
For example, we have team members with backgrounds in biology, computer science, physics, and statistics, among others. This diversity of skills and knowledge allows us to tackle complex problems from multiple angles and produce innovative solutions.
The collaborative nature of our team also fosters a sense of humility and continuous learning. We recognize that we have much to learn from each other, and this mindset encourages open dialogue and knowledge sharing. It’s incredibly stimulating to be surrounded by colleagues who challenge and inspire each other.
Another aspect of our team culture that I appreciate is the scientific rigor and commitment to quality. We place great importance on conducting rigorous research and ensuring that our methodologies and findings are robust. This commitment to excellence not only pushes us to produce high-quality work but also instills confidence in the impact and validity of our findings.
Overall, the collaborative and interdisciplinary nature of our team, combined with a focus on scientific rigor and commitment, creates an engaging and stimulating work environment. It fosters a sense of unity and shared purpose, and I truly believe that it enhances the quality of the work we produce.
