As attention to mental health increases and our understanding of its complexities improves, New Research Led by a doctoral candidate at Cornell Tech Dan Adler We know there is no one-size-fits-all solution to the mental health symptoms we experience in our daily lives, and using artificial intelligence, Adler is identifying trends to improve our understanding of the field in order to make symptom detection and treatment more effective.
New research led by Cornell Tech highlights the complex challenges and opportunities of using artificial intelligence to support mental health tracking and precision medicine. While the study found that AI cannot currently be trusted for such tracking, it raised important questions for future research, including the potential for bespoke solutions tailored to target populations and the challenges inherent in trying to implement broad-brush diagnoses and solutions for large, diverse groups of people.
Adler's research paper, Published in npj Mental Health Researchhas looked at how technology such as smartphone data can help measure behaviors related to mental health. For example, smartphones can track GPS data to monitor mobility, which is closely linked to depression. Previous researchers have published papers showing that people who are more mobile throughout the day are less likely to suffer from depression than their more sedentary counterparts.
Adler's research also uses AI to find correlations between behavior and mental health. While some studies have argued for consistency in such measurements, his team is using a combination of AI and a variety of other tools, including faculty advisors. Tanzeem ChoudhuryProfessor of Computing and Information Sciences and the Roger and Joelle Burnell Professor of Integrative Health and Technology, they focus on larger, more diverse populations. Their research reveals that there is no single set of behaviors that can uniformly measure the mental health of all individuals, a finding that highlights the importance of individualized measurement in mental health care.
Despite mental health clinicians' dedication to supporting their patients, Adler points out that there are significant challenges to care, especially when it comes to measurement. Traditionally, mental health diagnoses and assessments have relied heavily on self-reported information, clinician observations, and supplemental information from family and friends. This approach often complicates accurate diagnosis and treatment evaluation.
Measuring mental health is inherently complex, and because patients progress differently, there is often a lack of objective tools available to clinicians. Adler points out the limitations of previous pursuits of more objective measures, such as biomarkers in the brain and smartphone measurements he studied. “Research continues to highlight that mental health is not that simple,” he said, emphasizing that while data-driven methods are promising, mental health remains a highly personal and subjective experience.
“We've used AI tools to find connections between behavior and mental health, but we've found that these tools aren't very accurate,” Adler said of the paper. His research shows there are conflicting signals in the data, suggesting that a one-size-fits-all approach to measuring mental health isn't effective. Instead, Adler advocates for precision medicine and personalized tools that can tailor care to an individual's triggers and needs.
For example, his paper shows that heavy mobile phone use may be associated with depression in older adults, while less mobile phone use may be associated with depression in younger adults, demonstrating that additional context is needed to understand exactly how behaviors affect mental health.
“The potential of wearable sensors and smartphones may lie in their ability to account for differences, track symptoms and support precision treatment tailored to the individual's disease trajectory,” Choudhury says.
Adler's engineering background and Cornell Tech's interdisciplinary environment create a unique environment to explore solutions in the context of multiple disciplines and perspectives. Influenced by his personal experiences in the mental health care system, his work is driven by a passion to advance technological solutions to these challenges and create a more effective system of care for both patients and providers.
He emphasizes the importance of real-world impact in academic research, which is something Cornell Tech and Choudhury have been working on for some time. Human Cognitive Computing Groupis focused on driving a technology-enabled future of wellbeing.
As for future research, Adler sees great potential in using AI to address challenges in access to care. For example, Adler said new large-scale language modeling tools could fill gaps in mental health services. But he cautions against uncritical adoption of such technology. Technologists need to implement guardrails to ensure these systems provide helpful guidance, not harmful guidance, he argues.
Adler envisions a balanced approach to AI in mental health care, serving both as a means to fill gaps known to exist in the health care system and as a means to complement existing care practices. Adler believes that AI can be used to improve efficiency by handling administrative tasks and summarizing information, but believes it is important to evaluate these tools to truly enhance care delivery.
