Motivation

Taking a step back, modern Western medicine can be thought of as the center of mass for all interactions and exchanges of data between researchers and clinicians. As progress is made, this center shifts and so does the paradigm. Clinicians gather data from patients with diseases, subjecting them to treatments synthesized by researchers. Researchers, motivated by the chorus of clinicians signaling the need for cures, or by the insufficient effectiveness of existing treatments, shift their focus and resources toward creating new, improving, or maintaining these protocols.

Looking through a magnifying mirror into this web of data collection and analysis and fixing our focus to a single instance of doctor-patient interaction we can see a certain trend. It seems that the contact between the provider of treatment and the recipient has become a metaphor for measurement. We extract blood or tissue samples; we measure temperature, and analyze the chemical composition of various bodily fluids. As each new iteration of treatment aspires to take action as early as possible, the ability to recognize signals that predict the imminent coming of a disease has become of greatest importance to modeling and executing treatment effectively.

However, humans have long hit their limits in their ability to probe large sets of data while simultaneously performing calculations on them. We have turned to machines that collect, analyze, and output data only some of which we may fully comprehend. It does not appear that this disconnection between the actual corpus of the patient and the hands and eyes of the doctor will slow down. On the contrary, even for medical fields like orthopedics, the physical touch of the doctor and the subsequent communication about the patient’s pains have been, to a large degree, shifted to non-invasive scans and visualizations. One effectively no longer needs to move a knee joint to determine which knee hurts. The need to see a tumor to diagnose a cancer slowly fades in the face of blood analysis.

This thesis builds upon a crucial subtlety within this trend. The acceleration of data processing does not translate to a obsolescence of human perception. While the tactile link may be fading, the visual link is staying and through technology is becoming stronger. Specific domains, namely surgery, such as craniofacial surgery, orthopedics, and a behemoth that is medical education, refuse to distance themselves from the act of looking.

In these fields, “looking with eyes” is not a metaphor, it is a mechanical necessity. A surgeon cannot cut based on a spreadsheet of probabilities. They must see and internalize the spatial relationship of the animal anatomy. An orthopedic specialist must see the fracture within the context of the muscle and tissue. A medical student cannot learn the complexities of the human body through raw data points alone. They require a visual model to bridge the gap between theory and reality.

And as such, as we distance ourselves from the physical patient, the medical scan becomes the proxy for the patient’s body. The burden of reality presentation shifts from the flesh to the pixel. If the doctor is to rely on the screen rather than the touch, the image on that screen must not just be a data visualization. It should aim to be a faithful optical reconstruction of reality.