Clinical AI · Computational Psychiatry · Human-Machine Interaction · Affective Computing
Joanne Osuchukwu
MD · PhD Student, Biomedical Informatics · CCHMC/University of Cincinnati
Advisors: Dr. John Pestian, PhD MBA · Dr. Judith W. Dexheimer, PhD MBA FAMIA
How well do you recognize behavioral phenotypes?
Drag constructs to their DSM clusters.
Round Complete!
Graphite · Joanne Osuchukwu
The Human Condition
When I first made this piece, I intended for it to depict the secure attachment from attachment style theory. Sturdy geometric stability that feels like peace. But look at it long enough and it expands. It becomes the tale of the human condition: trying to exist in a universe that keeps widening, keeps complicating, keeps arriving faster than we can integrate it. Most of what governs our daily life is mental. Our minds, but also the inherited structures of other minds and circumstances that surround us.
Sun Tzu writes that he can predict the outcome of war before it begins by studying the inputs. The terrain, the supply lines, the preparation of soldiers, the atmosphere. According to him, given enough relevant variables, carefully read, the shape of what is coming may reveal itself. In a recursive world, there is so much pattern.
We are not mechanical, but we are still creatures of habit. Zoomed out longitudinal data, gathered carefully, surfaces patterns that repeat themselves too often to be random. The recurring trajectories, the inputs that tend toward certain outcomes, the early signals of a life going off track. If we can see the pattern, we have better intelligence to navigate. To know what tends to work, what tends to break, and at what point someone we love might be drifting somewhere they can't return from alone, so we can reach back in time.
Machine learning, utilized correctly, lets us do exactly this at scale. Ultimately the goal is prevention, or at the very least, more accurate and faster detection that don't further mystify what's already mysterious.