Not Only Luck.
An essay by John Melonakos

How My PhD Research Was Not Commercially Viable

This week I have posted things that connect my Georgia Tech research area to my progression as an entrepreneur. Today I will share a high-level overview of my specific area of research.

The area of neuroimaging I studied dealt with a particular type of MRI, called diffusion-weighted MRI. Diffusion-weighted MRI data is formed by biasing the MRI magnetic field in particular directions, which forces water molecules in the brain to move along the direction of the bias. If you do this for enough different directions, you can begin to understand how easy or hard it is for water to flow in various directions in the brain. This is especially interesting for white matter areas where all the axonal wires cross each other. The knowledge of the fluid flow in those areas helps researchers reconstruct an understanding of how the axonal fiber structures are arranged in the brain.

My specific research dealt with creating software to process diffusion-weighted MRI data. From this data, we generated imagery of fiber bundles like these:

Visualization of a DTI measurement of a human ...

Visualization of a DTI measurement of a human brain. Depicted are reconstructed fiber tracts that run through the mid-sagittal plane. Especially prominent are the U-shaped fibers that connect the two hemispheres through the corpus callosum (the fibers come out of the image plane and consequently bend towards the top) and the fiber tracts that descend toward the spine (blue, within the image plane) (Photo credit: Wikipedia)

B0009489 Nerve fibres in a healthy adult human...

B0009489 Nerve fibres in a healthy adult human brain (Photo credit: welcome images)

While the area of study made for good paper writing, there were several problems with the research that made it very unattractive for startup building:

  • It is a very niche modality of MRI, not commonly used in clinical practice due to its research nature
  • The research quality of information is dubious. The spatial resolution of the MRI machines only provides millimeter-level accuracy, which is not enough for most interesting problems
  • The time scale on any innovation in the healthcare space can be very long, due to the FDA approval processes

For these reasons, I have not attempted to do anything with my PhD research area in a commercial setting.

Have the things you studied in school been useful to your startup, or not?

 

Related articles

From the conversation

Comments preserved from the original post.

Paul Nichols

If I had to use the particulars of what I learned from coursework getting a Masters in Electrical & Computer Engineering at Georgia Tech in my startup, I’d be inclined to jump out of a window. I see my degrees as a crutch that I hope I never have to use again. What I did take away from school, though, other than some useful tidbits from psychology and philosophy, is learning how to learn… quickly and efficiently. That trumps coursework any day in my book.