Return to Podcasts

Revolutionary Technology for Detecting Early Signs of Alzheimer’s Through Speech Patterns With Winterlight Labs

Published on
July 5, 2022

Life science thought leaders, startup founders, and hubs are creating and innovating, from drug discovery to the final stage of getting life-saving and life-altering therapeutics and treatments into the hands of patients. In our Making Impact Podcast series, we will explore technologies and initiatives that are making a huge impact in their communities, locally and globally. 

Recently, our Co-Founder, Helen Kontozopoulos, met with Liam Kaufman - CEO and Co-Founder of Winterlight Labs - to discuss technological innovation in pharma, and how they’re changing lives using speech-based, digital biomarkers to quantify neurology and psychiatry for early Alzheimer's detection. 

In 2015, Winterlight Labs was founded with a mission to improve the lives of those suffering from Alzheimer’s disease and other speech-altered illnesses. By developing a technology that assesses cognitive health through speech, the company has been able to track changes and identify treatment responses. As CEO, Liam Kaufman puts it, “We’re measuring memory. [This] can have a huge impact on not just detecting problems earlier, but also being able to measure those symptoms in shorter time periods to be able to run trials faster, and find treatment sooner.” 

Using Technology for Early Diagnosis and the Benefits of Proactivity 

During the research phase, one of the first things Liam did was approach HCPs who all had similar notes on those suffering from dementia - noticeable speech differences. A disease like Alzheimer's typically starts 10 to 20 years before a diagnosis is made and once an individual has had that diagnosis, they’ve already reached a critical threshold. 

Though proactivity cannot cure certain diseases, early detection can reduce decline and prove to be practical for future preparedness. This includes lining up caregivers or arranging wills, which can prove helpful for loved ones and the patient themselves before the disease progresses. “At this stage where we think we can help people the most is really in clinical research. A lot of pharmaceutical companies are looking to come up with different treatments to stop the disease from getting worse [and] think intervening as early as possible will stop that damage from happening and progressing,” Liam says. 

Human-Centered Technology in Pharma 

Early on, a pain point Liam found for Alzheimer's detection was how cumbersome and subjective pen and paper tests were to quantify Alzheimer’s disease. “In neurology and psychiatry, there’s been this challenge where a lot of the tools that we use to measure efficacy or whether the treatment works have been really blunt and not very sensitive, but in the last five to ten years, that started to change pretty dramatically,” Liam says. 

In the summer of 2021, Winterlight Labs published research about their proprietary technology, showing that “with about two minutes of speech, [they] could create severity measures that were equivalent to an assessment that took about 35 to 40 minutes for a clinician to administer.” And as AI and machine learning improve, the potential for more effective and powerful mechanisms becomes possible. 

As Winterlight Labs moves into the real world with their technology, it is imperative that measurements are done effectively. Relying on different types of machine learning allows for quality assurance and valid samples. With this in mind, Winterlight Labs is able to address the needs of patients and clinicians with a subtle tool more sensitive in nature and able to detect changes. 

Accessibility is Key

Another challenge Liam outlines that patients face is travel time. Previously having worked with a Cognitive Neurologist, Liam says one of the issues the patients “had was they’d come from places that were four to five hours away from the doctor’s office.” When accounting for the thousands of different patients across North America, the sum of hours wasted became an apparent issue. 

This challenge came to the forefront even more so with the pandemic. As Liam notes, “People weren’t coming to the clinic in person or showing up in clinical trials, and quickly we saw a demand to be able to do these assessments remotely and at home.” COVID was the push for accessibility, cutting down hours spent behind the wheel and instead, allowing for the use of owned devices to administer assessments. “It became more practical for people to participate in research…this could be a way to diagnose and monitor people remotely.” 

Using Existing Tech to Aid Other Areas of Pharma Research 

Beyond Alzheimer’s disease, Winterlight Labs has taken on initiatives in multiple therapeutic areas with the use of their technology. Through their research, Liam identifies that speech changes are evident “not just in Alzheimer’s disease, but in schizophrenia, depression, different types of dementia, and Parkinson's.” Increasingly, they’ve found themselves supporting these different spheres which prove to have more effective treatments. What does this mean to Liam? “A real opportunity to help people in the real world, not just in clinical research and that’s something we’re really excited about.” 

Making Impact

Winterlight Labs' technology is a testament to the benefits AI and machine learning can pose in the life sciences. Not only does it shed light on the progressive impact technology has in the healthcare industry, but it proves a valuable tool to enhance the quality of human lives. 

This is the first installment of ODAIA’s new Making Impact Series, where we will go on to interview other thought leaders who are using cutting-edge technology to innovate in the life sciences. To stay in the loop about upcoming episodes, follow us on Linkedin.