Mental health is increasingly part of the national conversation—but the reality behind the numbers tells a more urgent story. Despite a growing ecosystem of digital tools, mental health apps, and employer wellness initiatives, the overall state of mental health in the US is stark. In fact, 1 in 4 adults (nearly 60 million) have experienced a mental illness in the past year. And those are just the ones we know of.
As we observe Mental Health Awareness Month, it’s clear that we’re missing a lot when it comes to the critical information needed to diagnose and treat these conditions. But it’s not all bad news—artificial intelligence (AI), particularly tools that can analyze unstructured clinical notes, may offer a powerful path forward. New research from John Snow Labs, Oracle Health, and the Children’s Hospital of Orange County explores just that.
The Missing Link: Unstructured Data
The research compared the detection of neuropsychiatric symptoms—such as anxiety, memory loss, agitation, and mood disorders—using structured electronic health record (EHR) data versus structured data augmented with unstructured clinical notes. After analyzing more than 109,000 patients using Oracle’s Real-World Data platform, linked to national claims databases, the findings were sobering.
Key mental health events were routinely overlooked when relying solely on structured data. In fact, the number of suicide and self-harm events doubled once unstructured notes were included. Symptoms like irritability and hallucinations were often recorded only in narrative form, highlighting how much is lost in translation when relying on codes alone.
Structured EHR fields—think diagnosis codes or medication lists—serve as the skeleton of a patient’s health history. But when it comes to mental health, it’s often the nuances documented in unstructured notes that reveal the bigger picture.
These clinical narratives capture patient behavior, family input, mood fluctuations, and cognitive concerns that may never make it into a billing code. For pediatricians and general practitioners, reluctance to assign a definitive mental health diagnosis—due to stigma, uncertainty, or scope of practice—often results in key symptoms remaining in the margins.
If we want to really understand the state of mental health, we must capture the full patient journey. Structured data is a necessary part of that, but alone, is insufficient. The details that clinicians jot down during encounters, or the way they describe subtle symptom progression, contain vital insights we can’t afford to ignore.
Hybrid AI Models are a Bright Spot
Natural Language Processing (NLP) and Large Language Models (LLMs) are two areas that have potential to greatly improve the mental health space. These AI technologies can analyze unstructured text at scale, surfacing mental health indicators that traditional tools miss. In the aforementioned study, incorporating unstructured notes led to a 20% increase in identified outcome events—a significant leap in progress.
That said, building these systems is no small feat. Training AI to extract clinically relevant insights requires annotated data, clinician oversight, and significant compute resources. And the reality is, many health systems lack the internal capacity to manage such initiatives on their own.
The challenge is compounded by the complexity of mental health coding itself. Unlike a broken bone, which comes with a clear diagnostic code and treatment path, mental health symptoms often fall into gray areas. Without precise language or billing incentives to document these events consistently, it’s no surprise we’re not capturing the full spectrum.
The most promising approach lies in hybrid AI architectures that combine the precision of rule-based NLP with the flexibility and reasoning capabilities of LLMs. These pipelines can extract both overtly stated symptoms and more subtly implied conditions, achieving near expert-level accuracy without relying solely on structured fields.
In practice, that means better detection of at-risk patients, faster insights for researchers, and more informed care pathways. LLMs fine-tuned on medical language can flag early signs of distress even if words like “depression” or “anxiety” are never explicitly mentioned. This approach allows AI to scale across healthcare settings, from academic centers to community clinics, and adapt to new data with minimal retraining.
What’s Next?
As the incidence and severity of mental health conditions continue to rise, it’s clear that traditional methods of tracking and treating these disorders are no longer sufficient. The good news is that much of the information we need to paint a more comprehensive picture is at the fingertips of our healthcare professionals. We just need to find better ways to put it to work.
With AI-powered analysis of both structured and unstructured clinical data, healthcare systems now have the opportunity to uncover what was previously missed—leading to earlier diagnosis, smarter interventions, and more comprehensive care. The challenge is enormous, but the benefits can be life-changing. And it’s about time we leverage the tools we have to meet this crisis with the precision and urgency it deserves.
David Talby
David Talby is CEO for Pacific AI and CTO for John Snow Labs.
