Kris T. Huang, MD, PhD, CTO

Innovation & Legacy

Global vs Local optimization


During a recent, socially distanced Pymedix corporate walking meeting outside in early January 2021, we discussed which company/companies were the most innovative personally. Perhaps predictably, companies like SpaceX, Tesla, and Apple were at the top of the list.

Besides innovation, what do these companies have in common? They all operate in safety-critical industries. SpaceX is aerospace, Tesla is automotive, and Apple entered the medical device industry in 2018 with the single-lead EKG feature in its Series 4 Apple Watch.

These industries come with a lot of technical legacy, and yet here we are with 3 wildly successful companies showing the old guard how innovation is done. SpaceX and Tesla in particular didn’t achieve their current success immediately, but they both began with a fresh approach that flew in the face of accepted legacy practice but proved to be correct.

The difference between leaps of innovation and incremental legacy improvement is the difference between global and local optimization. While global optimization still involves local optimization at some point, the difference is the goal, known in math, physics, and AI as the objective function.

Continue reading “Innovation & Legacy”


Kris T. Huang, MD, PhD, CTO


All models are wrong.

But, some are useful.


The idea that “All models are wrong” is generally attributed to statistician George Box [1], but it certainly isn’t new. Of course, even though all models are wrong, it’s clear that some models are quite useful, and good models (should) encapsulate some insight that provides meaning to observed data and an explanation for predictions. A good model should provide insight into a system’s mechanism.

This more traditional scientific notion of the model concept is in fairly stark contrast to today’s growing reliance on so-called “black box” deep learning solutions for increasingly important decision processes. It’s not that deep learning methods are not understood, but the models produced by deep learning typically are, at least for the time being, not interpretable, i.e. they cannot provide explanation for the calculations that led to their predictions.

The need for insight doesn’t just apply to results, it also applies to model formation.

Continue reading “All models are wrong.”


Kris T. Huang, MD, PhD, CTO


Mapping image registration confidence

Automated TG-132 patient-specific QA

  Pre- (green) and post-treatment                 (magenta) CT fusion
   Registration confidence heat map
         (Red: low, blue/green: high)


Kris T. Huang, MD, PhD, CTO


Advanced interactive visualization in Autofuse

Real-time volumetric rendering–
See and explore the human body like never before



MRI or CT? Tell us what you think!

Continue reading “Can you tell between an MRI and a CT?”