For decades, treating major depression has involved a frustrating cycle of trial and error, with patients cycling through medications and therapies while clinicians rely on symptoms and intuition more than biology. A growing body of research now suggests that brain scans could soon help match people to the treatment most likely to work for them, turning what has long been guesswork into a more precise science. If these tools reach everyday clinics, they could shorten suffering, cut costs, and reshape how I, and many others, think about mental health care.
Why depression treatment still feels like educated guesswork
Major depressive disorder is diagnosed based on how a person feels and functions, not on what shows up in a lab test or scan. Two patients can both meet criteria for depression yet differ dramatically in sleep, appetite, thinking speed, and motivation, which helps explain why the same antidepressant can lift one person and leave another unchanged. Without biological markers to guide decisions, clinicians often rotate through medications and psychotherapies, hoping to land on the right combination after months of trial and error.
That approach is especially painful when symptoms are severe or when people have already tried several treatments without relief. I have heard many patients describe feeling like “guinea pigs” as they move from one prescription to the next, each with its own side effects and waiting period. The emerging promise of brain-based tools is that they could give clinicians a more objective way to predict who will respond to which option, using patterns in brain activity and connectivity rather than relying solely on checklists and clinical hunches.
Six depression “biotypes” and what they reveal in the brain
One of the clearest signs that depression is not a single disease comes from work led by Jun and Williams, who used functional MRI to identify six distinct “biotypes” of depression based on how brain networks behave. Their team found that certain patterns of connectivity, especially in circuits involved in emotion and cognition, clustered together in ways that lined up with different symptom profiles. In another analysis, Jun and Williams showed that using fMRI brain imaging improved their ability to classify patients, including those identified as having the cognitive biotype, which is marked by problems with attention and thinking speed, compared with relying on symptoms alone, and they reported these findings in a study of depression biotypes.
Jun and colleagues have argued that, to their knowledge, this is the first time researchers have been able to demonstrate that depression can be explained by different disruptions in brain function, and that some of these biotypes do not respond to standard antidepressants. In their report, Jun described how certain subtypes, such as those with more pronounced cognitive impairment, may need different strategies than traditional serotonin-focused drugs, which helps explain why some patients fail multiple medications. By tying specific symptom clusters to measurable brain signatures, the work on these six biotypes offers a roadmap for tailoring treatment to the underlying circuitry rather than treating all depression as the same condition, a point underscored when Jun noted that some biotypes simply do not respond to standard antidepressants.
From one-size-fits-all to scan-guided choices
As these brain-based subtypes become clearer, the vision is to move away from a one-size-fits-all model and toward scan-guided decisions in routine practice. Reporting on this shift has highlighted how Aug and other commentators describe a near-future in which clinicians can use a quick brain scan to determine the best treatment for depression, rather than defaulting to the same first-line drug for everyone. In that coverage, Aug emphasized that, Soon, medical professionals will be able to conduct a quick brain scan to determine the best treatment for depression, replacing the current one-size-fits-all approach, a claim that reflects growing confidence that imaging markers can be translated into practical tools, as described in a report on six depression subtypes.
That vision aligns with broader efforts to build decision support systems that combine brain data with clinical information. Aug has also described how Researchers are testing tools that integrate brain scans, depression ratings, and other health data to suggest whether a patient is more likely to benefit from medication, psychotherapy, or neuromodulation. In those accounts, Researchers are also testing another decision support tool that uses data from a patient’s brain scan, depression rating scales, and other clinical information to guide treatment choices, reflecting a push to embed neuroscience into everyday care rather than keeping it confined to research labs, as detailed in coverage of personalized mental health care.
Historic hints that scans can predict who needs meds or therapy
The idea that brain imaging can forecast treatment response is not entirely new. More than a decade ago, researchers used positron emission tomography to look at resting brain metabolism before treatment and then tracked how people responded to medication or psychotherapy. In one historic study, After having an initial baseline PET scan, 33 patients completed 12 weeks on escitalopram, a common antidepressant medication, while another group received cognitive behavioral therapy, and the team found that patterns in the anterior insula and other regions predicted who would do better with which option, as reported in a landmark analysis of how PET scans can guide depression treatment decisions.
A related line of work used functional imaging to focus on the insula, a region involved in emotional awareness and bodily states, to see whether its baseline activity could steer treatment selection. In that research, investigators reported that if a patient’s pre-treatment resting brain activity was low in the front part of the insula, on the right side of the brain, they were more likely to respond to cognitive behavioral therapy, while higher activity in that same area predicted a better response to antidepressant medication, a pattern that suggested a single scan could indicate a differential response to these treatments, as described in a study showing how a scan predicts whether therapy or meds will best lift depression. Those early findings were small and preliminary, but they laid the conceptual groundwork for today’s more sophisticated, network-level approaches.
Reversing brain signals and targeting circuits directly
More recent work has gone beyond prediction to ask whether directly changing brain signaling patterns can relieve depression. At Stanford, researchers used a detailed analysis of functional MRI data to examine the timing of activity between brain regions, looking at minute differences in when one area activates relative to another. The new analysis used minute differences in timing between the activation of different areas to also reveal the direction of communication, which allowed the team to identify brain signals that seemed to flow in the “wrong” direction in people with depression, as described in a report on how researchers reverse brain signals.
Armed with that map, the group then used targeted neuromodulation to try to flip those signals back to a healthier pattern, essentially nudging information flow in the opposite direction. By aligning stimulation with the timing and direction of specific circuits, they reported rapid improvements in mood for some participants, suggesting that understanding the fine-grained dynamics of brain networks is not just an academic exercise but a practical guide for intervention. For me, this kind of work illustrates how brain scans can move from static pictures to dynamic roadmaps that show where and how to intervene when standard treatments fall short.
Brain-based markers and the promise of personalized medication
While circuit-level interventions are one frontier, another is using imaging to refine medication choices. A study described as Published in JAMA Network Open examined how brain-based markers could help personalize depression treatment by predicting which patients would respond to specific antidepressants. In that work, investigators combined functional imaging with clinical data to identify patterns that forecasted response to different medications, and the report noted that the study reveals promising progress toward predicting how patients with major depressive disorder will respond to antidepressant medications using brain imaging and clinical data, as detailed in coverage of brain-based markers.
These kinds of models could eventually sit inside electronic health records, offering clinicians a probability estimate that a given patient will respond to a particular drug based on their scan and symptom profile. Instead of cycling through medications in a fixed order, a psychiatrist could start with the option that the model suggests has the highest likelihood of success, potentially shortening the time to remission. I see this as analogous to how oncologists use tumor markers to choose targeted therapies, with the crucial difference that here the “tumor” is a dysfunctional network of brain regions that can be visualized and quantified.
AI, mobile data, and the next generation of decision tools
Artificial intelligence is rapidly amplifying what brain scans alone can do by integrating imaging with data from smartphones, wearables, and digital assessments. Dec has been used to frame this trend, with reports noting that AI, neuroscience, and data are fueling personalized mental health care by combining mobile device data, brain imaging, and clinical records into unified models. In that context, New technologies integrate mobile device data with brain scans and symptom reports to track mood and behavior in real time, creating feedback loops that can refine treatment recommendations as a person’s life and brain change, as described in an overview of how AI, neuroscience, and data are reshaping care.
In practical terms, that might mean an app that passively tracks sleep, movement, and social communication, then flags when a person’s patterns diverge from their baseline in ways that past data link to relapse. Combined with a baseline scan that identifies their depression biotype, such a system could alert a clinician that a patient with a cognitive biotype is slipping and may need a medication adjustment or cognitive training before symptoms fully return. I find this convergence of brain imaging and digital phenotyping both promising and fraught, since it raises questions about privacy, consent, and how to ensure that algorithmic recommendations remain transparent and clinically grounded.
Traditional medicine, Yueju Pill, and culturally specific brain signatures
Personalization is not limited to Western pharmaceuticals. Jan has been used to highlight work on Yueju Pill, a traditional Chinese herbal formula, in which researchers compared its effects with those of a standard antidepressant while tracking brain network changes. In that study, However, a key biological difference emerged between the groups: Only patients in the Yueju Pill group experienced a significant increase in connectivity within certain brain networks associated with emotion regulation, suggesting that the herbal treatment may engage distinct neural pathways compared with conventional drugs, as reported in an analysis of how Yueju Pill affects the brain.
Jan has also been used to frame the broader conclusion from that work, which was summarized under the heading Toward Personalized Treatment for Major Depression Taken together, the findings suggest that brain network patterns identified before treatment could help determine which patients are most suitable for Yueju Pill treatment. The report noted that Toward Personalized Treatment for Major Depression Taken together, the findings suggest that brain network patterns identified before treatment could help determine which patients are most suitable for Yueju Pill treatment, underscoring that even within traditional medicine, brain imaging can reveal who is likely to benefit and who might need a different approach, as described in coverage of personalized treatment for major depression. For me, this illustrates how neuroscience can bridge cultural divides in mental health care by providing a common biological language for evaluating diverse treatments.
How close are we to scan-guided care in everyday clinics?
Despite the excitement, there is still a gap between research findings and routine practice. Most of the studies on depression biotypes, insula predictors, and circuit-level interventions have been conducted in specialized centers with advanced imaging equipment and expert analysts. Scaling that work will require standardizing scan protocols, validating markers across diverse populations, and building user-friendly software that can translate complex brain data into clear, actionable recommendations for busy clinicians, rather than dense statistical outputs that only a handful of specialists can interpret.
At the same time, the trajectory is unmistakable. Jun and Williams are refining their biotype models, Aug and other Researchers are testing decision support tools that blend imaging with clinical data, and teams working with Yueju Pill and other treatments are showing that brain network patterns can identify who is most likely to benefit. When I look across these efforts, I see a field moving steadily toward a future in which a person’s brain scan, combined with their symptoms and daily-life data, will help determine whether they start with medication, psychotherapy, neuromodulation, or even a traditional herbal formula, making depression care more precise, more humane, and far less dependent on trial and error.
More from Morning Overview