Valid and Ethical? A Critical Analysis
The diagnostic use of sophisticated brain imaging techniques like Single Photon Emission Computed Tomography (SPECT) scans and functional Magnetic Resonance Imaging (fMRI) has advanced our understanding of brain function and activity, enabling researchers and clinicians to peer into the brain’s intricate operations, providing insights into neuroanatomy, and normal and abnormal brain function. Transferring these tools from the research lab to clinical diagnosis and treatment of individual patients sounds like a wonderful idea and some clinicians are doing so. But is the legitimate use of these tools? And is it ethical to use them if not?
Chief among the concerns about such a transfer are the basic limitations in the ability to accurately infer from SPECT scans, fMRI, and other techniques to the analysis of a single individual. In research protocols, dozens and even hundreds of scans of people with a given psychiatric syndrome such as depression, are combined and averaged. This group result is then compared to an equal-sized group of matched individuals without any such syndrome. If enough scans are combined in this way subtle differences in brain structure and function may be detected.
However, if you take any single scan from the depressed group there is a very high probability that it will look more like the averaged scan from the non-depressed group and if you take any random single scan from the non-depressed group there is a very high probability that it will look more like the depressed group average.
A recent high quality study looking at more than 3000 combined fMRI scans from depressed and non-depressed individuals resulted in a accurate detection/differentiation rate of only 60%–statistically valid but little better than a coin toss in any individual case. And in this study advanced machine learning techniques were used to analyze the differences, a task that machines are far better at than humans.
Consider a favorite analogy: the averaged differences between the heights of randomly selected men and women may yield valid statistical differences. However, differentiating between a tall man and a short woman using these same statistics — i.e., assuming literally that men are taller than women — would be both over-simplifying and misapplying these findings. The phrase “men are taller than women” properly used means “on average men are taller than women but a great many women are taller than men.” In other words you can no more “diagnose” a person as male versus female based on his/her height alone than can you diagnose depression from a brain scan.
The takeaway is that the averaged differences between dozens or hundreds of images aggregated in research protocols may yield valid differences between the brains of psychiatric populations and those without the disorder under study. However, the application of these same findings to the diagnosis or treatment of individual psychiatric patients would be inaccurate and misleading.
In the realm of psychiatric diagnosis, clinicians assess behavioral symptoms as the primary means of identifying and categorizing mental health conditions. Embedded within the diagnostic manuals that guide clinical practice, this traditional approach reflects a glaring absence of definitive biological markers within the field of psychiatry. The invaluable subjective insights clinicians provide are inconsistent at best; they also introduce unavoidable challenges, underscoring the potential for adjunct methods that could provide a more nuanced comprehension of mental health disorders. The groundwork for the integration of brain imaging techniques – notably, Single Photon Emission Computed Tomography (SPECT) scans and functional Magnetic Resonance Imaging (fMRI) – into the diagnostic process is established within this context. However, the ethical and validity concerns associated with using these imaging technologies in their current state cannot be overlooked.
SPECT Scans
Single Photon Emission Computed Tomography (SPECT) scans monitor blood flow in the brain using radioactive tracers to detect abnormal patterns that underlie specific symptoms. SPECT scans are not only endorsed by the Canadian Association of Nuclear Medicine for a range of neuropsychiatric disorders, they are also praised for offering valuable insight into the biological underpinnings of mental health conditions. SPECT scans are much more affordable than fMRI, with costs varying significantly.
fMRI
Functional MRI assesses how the brain uses oxygen using radio waves and magnets to provide precise, moment-to-moment depiction of brain activity.
Although the expense of fMRI scans is prohibitive, the impact of metal implants preventing a patient from being able to have the scan is significant. The capacity of fMRI to provide superior spatial resolution far outweighs these considerations when interrogating brain function.
The use of SPECT scans and fMRI in the clinical setting to diagnose psychiatric conditions has been met with skepticism, largely due to their limited ability to distinguish individuals with psychiatric conditions from healthy subjects. Indeed, most imaging studies of psychiatric disorders have demonstrated low sensitivity and low specificity, which translates to an inability to reliably identify the presence of psychiatric disorders and to accurately differentiate among disorders. The observation of common patterns of brain activation across the spectrum of psychiatric conditions further complicates the notion that these techniques might be unique biomarkers for specific diagnoses. This overlap suggests that there may be similar sets of biological underpinnings for different psychiatric disorders, which challenges the utility of SPECT scans and fMRI in making precise, individualized diagnoses in the clinical setting. Importantly, the use of these imaging modalities raises ethical concerns with respect to their potential for misdiagnosis or overdiagnosis, and for this reason, their findings ought to be incorporated within the context of an overall clinical assessment. At Sterling Institute, we recognize these challenges and support a well-rounded approach to mental health care, which marries scientific insights with empathetic, personalized care to deliver optimal clinical outcomes. To learn more about our comprehensive approach, please visit Sterling Institute.
Research Versus Clinical Applications of Brain Imaging Studies
The distinction between research and clinical applications of brain imaging studies, such as SPECT scans and fMRI, is crucial to understanding their role in psychiatric diagnosis. In research protocols, these imaging techniques have demonstrated their ability to identify trends based on brain differences between individuals with a given psychiatric condition and those without. These differences become statistically significant and therefore clinically relevant — though still with substantial inter-individual variation — once the data from a large number of scans (often more than a hundred) is aggregated. This mechanism depicts the average features of a group — like a composite sketch — rather than the individual unique features that would be necessary for clinical purposes. This approach is invaluable in research for identifying underlying mechanisms and broad trends of psychiatric disorders. However, it presents a significant barrier to applying these findings directly to the diagnosis and treatment of individual patients.
The challenge is further underscored by the inherent variability in brain structure and function from one person to the next. Just as no two people have the same fingerprint, each person’s brain demonstrates unique patterns of connectivity and activity. This individual variability is such that a finding that is statistically significant for a group may not be significant for any single individual in that group. For example, while a research study could show that a particular brain region is consistently less active in a large group of people with depression, the same is certainly not true for every single individual diagnosed with the condition. Thus, reliance on neuroimaging techniques for clinical diagnosis in psychiatry risks oversimplification and the potential for misdiagnosis or misapplied treatments. Comprehensive, personalized assessments remain a core characteristic of integrated neuropsychiatric services at institutions such as Sterling Institute. For more information about the comprehensive neuropsychiatric services provided by Sterling Institute, please visit sterlinginstitute.org
Future Implications for Psychiatry
The evolving world of neuroimaging offers promise as we navigate the increasingly complex terrain of mental health care. Neuroimaging holds great potential to change the future of psychiatric diagnosis and treatment. Though current SPECT scans and fMRIs are limited in their use for individual patient diagnosis and their ethical implications; research in the applications of these technologies is advancing rapidly. These advanced imaging techniques, with their ability to provide a quantitative biological marker, have the power to overcome the limitations of traditional diagnostic approaches, which rely on behavioral criteria alone[3].
At specialized clinics like Sterling Institute, this ground-breaking technology is already being incorporated into the diagnostic process for patients with ambiguous or complex symptoms. Using the detailed insights provided by SPECT scans and fMRI, these clinics remain on the cutting edge of the field’s progression; refining their approach to patient care in accordance with the unique neurological signatures of each patient’s condition. The precision with which these clinics are now able to dispense diagnoses and treatment is evidence of the technology’s transformative power in enhancing patient outcomes as well as shaping the future of psychiatry. As we progress through the next phase of research, we can only hope that these advancements will soon provide the foundation for behavioral analysis in mainstream psychiatric practice, providing a more scientifically-driven approach to mental health care at large. To find out more about how Sterling Institute is integrating these advanced technologies into its holistic approach to mental health care, visit Sterling Institute for additional information.
Sterling Institute’s Holistic Approach to Mental Health Care
At Sterling Institute, we take a full-spectrum approach to caring for our patients, combining the latest scientific advances with an intimate knowledge of the human heart. Treating an array of mental health issues — including anxiety, depression, mood disorders, and more — we are able to provide a broad suite of services, ranging from psychiatry and pharmacology to psychotherapy, counseling, ketamine treatment and transcranial magnetic stimulation, refinements that allow us to create a tailored, specific care plan for the unique needs of each patient. However, with our dedication to this personalized, compassionate care, we hope to facilitate care that not only cares for the biological, but for the emotional as well.
There are no blood tests, x-rays, Scans or MRIs that can diagnose these disorders. As in many areas of medicine, health professionals rely on a careful history and a compilation of observations of behavior over time together statistical information on populations, to reach a probable diagnosis.
Sensitivity and Specificity
SPECT scans and fMRI are helping us learn more and more about what goes on in the brain when someone has a particular emotional or psychological experience. However, these tools have yet to show sensitivity and specificity that makes them useful in diagnosing individual cases.
Specifically, the sensitivity and specificity of a test are two measures related to how well the test is able to identify people with or without a particular disease. If a test has high sensitivity, it is good at correctly identifying people who have a particular disease. If a test has high specificity, it is good at correctly identifying people who don’t have a particular disease. If a test is 100% sensitive and 100% non-specific it will diagnose everyone with condition. If a test is 100% specific and 100% insensitive it will diagnose no one with the condition.
The cost and potential side effects resulting from SPECT and fMRI scans, particularly those with injecting radioactive tracers, argues for caution in the application of these tools in the diagnosis of psychiatric conditions as they are as yet neither sensitive enough nor specific enough for clinical use. While we might expect that people with PTSD would show different patterns of brain activation when they are re-living, completely immersed in an experience, there are a lot of other states of mind that can lead to numbness, and we can’t draw a line on an fMRI scan and say, “At this point, these findings indicate that you have PTSD.”
Still, the future of neuroimaging in psychiatry remains bright. Advances in technology and methodology are allowing for ever-improved sensitivity and specificity of such tools. Ongoing research is helping to determine the ways in which these imaging techniques can identify biomarkers for psychiatric conditions, thereby supplementing the traditional diagnostic process with biological—and objective—information. As psychiatric diagnosis evolves, so too will the development of treatment plans that are more personalized and more effective. At Sterling Institute, we remain poised to lead the way on such advancements, seamlessly integrating compassionate care with evidence-based practices to offer our patients the most comprehensive array of treatment options. To learn more about how we are incorporating the latest research into our approach to mental health care, visit Sterling Institute.
Conclusion: Emphasizing the Limitations of Brain Imaging Techniques
Given the pitfalls of brain imaging techniques in psychiatric diagnosis, it’s crucial that we not lose sight of the bigger picture. While SPECT scans and fMRI are groundbreaking in their capacity to reveal important aspects of the brain at work, their application to individual psychiatric diagnoses is marked by significant limitations. The potential of such scans to inform psychiatric diagnoses is often overshadowed by the thorny ethical and validity concerns that emerge when attempts are made to translate group-based research findings to the unique neurobiological landscape of a single patient. This is further evidence of the crucial role that a holistic approach to mental health care—one that integrates the hard science of the day with the myriad, complex facets of human existence.