QuantaBrain’s Manifesto
Today we are officially establishing QuantaBrain, an Italian startup that develops artificial intelligence systems for the diagnosis of psychiatric disorders. The essence of our undertaking is crystallized in three foundational pillars that govern our research and activity.
1. Bridging the Gap between Psychiatry and Neurology
Historically, the brain has been the domain of two separate medical specialties: neurology, that focuses on brain abnormalities; and psychiatry, that deals with behavioral variations lacking clear biological markers. Since the beginning of the century1, there have been efforts to integrate these fields, targeting a reevaluation of mental illnesses as biological in nature. However, the division persists due to challenges such as the impracticality of studying human behavior with animal experiments, and the complexity of accessing and analyzing a living brain.
The limitations have directed researchers towards non-invasive methods like electroencephalography (EEG) and functional magnetic resonance imaging (fMRI), the latter providing the detailed spatial resolution crucial for brain functioning analysis. However, traditional statistical methods are insufficient to interpret the rich, dynamic data generated by fMRI.
At QuantaBrain, we confront these challenges by leveraging AI to examine fMRI data, preserving the detailed insights necessary to understand the heterogeneity of psychiatric conditions. Our mission, therefore, is to understand the complex biological nature of behavioral disorders, merging the descriptive focus of psychiatry with the quantitative rigor of neurology.
2. Personalized approach
Taxonomy simplifies complexity by subdividing phenomena into more homogeneous groups for easier analysis. However, to correctly define a taxonomy, a hierarchy of features’ importance must be defined. For instance, while frogs and humans both lack tails, higher order features align humans more closely with monkeys. However, ranking the importance of behavioral traits poses challenges; it is difficult for example to deem whether stability of mood is more important than empathy. This difficulty is also exacerbated by the fact that behavioral traits exist on a spectrum rather than as binary attributes (i.e. tail presence/absence).
At QuantaBrain, we realize diagnostic labels serve as practical simplifications for actions like treatment reimbursement yet fail to capture the intricacy of behavior and its neurobiological basis. We adhere to the “Diagnostic and Statistical Manual of Mental Disorders (DSM)” for diagnostic guidelines while providing a data-driven, individualized profile to assist clinicians in tailoring precise treatments for patients.
3. Empowerment Through Knowledge
We are deeply convinced that a more scientific understanding of psychiatric disorders and their description in biological and measurable terms would allow to:
(1) eliminate the stigma towards these disorders;
(2) measure incremental improvements that are difficult to notice in behavior, providing evidence of the effectiveness of the treatment;
(3) increase self-understanding and acceptance.
For many people, points (2) and (3) are in contradiction. How is it possible to want to undergo treatments to modify our own behavior and at the same time accept ourselves?
We draw a parallel between behavior and appearance—it is natural to seek improvement, as we would with grooming or attire, while also embracing our authentic self. QuantaBrain’s diagnostic reports aim to be enlightening, providing clear insights into the biological basis of behavior to empower individuals in their therapeutic journey and promote a deeper self-understanding.
QuantaBrain is not just bridging the gap between neurology and psychiatry—it is adapting the lexicon and precision of neurology to the qualitative dimensions of psychiatry, using AI not just to categorize, but to comprehend and celebrate the full spectrum of human behavior.
- Price BH, Adams RD, Coyle JT. Neurology and psychiatry: closing the great divide. Neurology. 2000 Jan 11;54(1):8-14. doi: 10.1212/wnl.54.1.8. PMID: 10636118. ↩︎