MindScan’s Classification Report Technology

MindScan’s technology accurately reveals brain structure abnormalities indicative of neuropsychiatric disorders.

MindScan’s scientific team developed machine learning derived algorithms that use an edited anatomical MRI brain scan to identify patients with a specific neuropsychiatric disorder. This research and development effort was funded with non-dilutive grants totaling $13.8 million.

As shown in a rigorous, published study, the technology’s sensitivity was 93.6% to 100% and the specificity was 88.5% to 100% versus gold-standard, consensus panels of experts using research-grade clinical data.

As illustrated below, this accuracy is achieved because brain structure in a patient with a given disorder differs in subtle but fundamentally important ways from brain structure in a healthy individual or in a patient with a different disorder.

ADHD.png
BD.jpg
SK.jpg
 
Key NEW 2.png
 

Published Results

Bradley S. Peterson, MD and Ravi Bansal, PhD, co-founders of MindScan, along with other authors, published “Anatomical Brain Images Alone Can Accurately Diagnose Chronic Neuropsychiatric Illnesses” in 2012. (Citation: Bansal R, Staib LH, Laine AF, Hao X, Xu D, Liu J, et al. (2012) Anatomical Brain Images Alone Can Accurately Diagnose Chronic Neuropsychiatric Illnesses. PLoS ONE 7(12): e50698.) An edited abstract of the publication can be found below. The full study (including the original abstract) can be found at  https://tinyurl.com/ya9w6g9l.

Objective

Classification of neuropsychiatric disorders using imaging-based measures of brain abnormalities offer the hope of improving the accuracy of clinical diagnosis, thereby reducing the costs associated with incorrect treatments. Previous attempts to use brain imaging to classify patient as having or not having a neuropsychiatric disorder, however, have had only limited success in patients who are independent of the samples used to derive the classification algorithms. We aimed to develop a classification algorithm that can accurately classify chronic, well-characterized neuropsychiatric illness in single individuals, given the availability of sufficiently precise delineations of brain regions across several neural systems in anatomical MR images of the brain.

Methods

We have developed an automated method to classify individuals as having one of various neuropsychiatric illnesses using only anatomical MRI brain scans. The method employs a semi-supervised machine learning algorithm that discovers natural groupings of brains based on the spatial patterns of variation in the morphology of the cerebral cortex and other subcortical brain regions. We used split-half and leave-one-out cross-validation analyses in large MRI datasets to assess the reproducibility and classification accuracy of those groupings.

Results

In MRI datasets from persons with Attention-deficit/hyperactivity disorder, Schizophrenia, Tourette Syndrome, Bipolar Disorder, or persons at high or low familial risk for Major Depressive Disorder, our method discriminated with high specificity and nearly perfect sensitivity the brains of persons who had one specific neuropsychiatric disorder from the brains of healthy participants and the brains of persons who had a different neuropsychiatric disorder.

Conclusions

Although the classification algorithm presupposes the availability of precisely delineated brain regions, our findings suggest that patterns of morphological variation across brain surfaces, extracted from MRI scans alone, can successfully indicate the presence of chronic neuropsychiatric disorders. Extensions of these methods are likely to provide biomarkers that will aid in identifying biological subtypes of those disorders, predicting disease course, and individualizing treatments for a wide range of neuropsychiatric illnesses. 

The Science Behind the Technology

The fundamental scientific axiom in system neuroscience is that the brain is ultimately the source of all thought, emotion, and behavior, and that persistent alterations in thought, emotion, and behavior are likely represented in structural abnormalities that are unique to each disorder, since structure ultimately determines function. MindScan’s technology exploits this basic axiom by detecting unique patterns of brain abnormalities in individual patients. For example, patients with ADHD have bilateral enlargement in the inferior frontal and temporal lobes of the brain. This pattern of abnormality is very different from the pattern of abnormality for all other disorders. In particular, patients with Bipolar Disorder have enlargement across the entire lateral aspect of the left and the right hemispheres of the brain. MindScan’s technology quantifies these patterns in the brains of individual patients and applies machine learning to match them with the patterns of abnormality that we have identified in a large, proprietary database of patients with differing neuropsychiatric disorders. That matching process allows MindScan to provide a report that classifies individual patients as having or not having a specific neuropsychiatric disorder.

Anatomical studies have consistently identified structural disturbances in children and adolescents with ADHD. Individual studies and meta-analyses (some including thousands of participants) have identified the brain biomarkers. The following brain biomarkers have considerable scientific evidence to support their link to ADHD: reduction in either volume or thickness or both in frontal cortex; reduced volumes of the basal ganglia, especially the caudate nucleus and putamen and thalmus; enlarged hippocampus; reduced size of the cerebellum; and enlargement in certain subregions of the amygdala.

As illustrated below, MindScan’s technology captures these patterns of structural abnormalities. In a well-controlled retrospective study, MindScan identified patterns of structural abnormalities in these regions to classify whether a new patient did or did not have ADHD with 93.6% sensitivity and 88.5% specificity versus a gold-standard consensus panel of three experts using research-grade clinical assessments.  

ADHD Neural “Fingerprint”

 The figure below illustrates brain structure abnormalities or brain biomarkers found in ADHD patients. Blue indicates a reduction and red indicates an increase in local volumes of specific brain regions.

Patient’s Neural “Fingerprint”

The figure below illustrates the structural features in a patient. You can compare the pattern of biomarkers versus the reference figure in section 5. The arrows indicate where this patient differs from the reference.

ADHD just image base.png
ADHD just image pt.png

Key: Amg = amygdala; HC = hippocampus; CN = caudate nucleus; GP = globus pallidus; Put = putamen; Thal = thalamus.

MindScan’s Drug Efficacy Measurement Technology

Scientific investigations have shown that neuropsychiatric drugs case neuroplastic changes indicative of efficacy. MindScan will partner with pharmaceutical companies to increase the speed and accuracy of clinical trial recruitment and supplement subjective evaluations of efficacy with objective criteria and analyses based on changes in brain morphology.