What if you could observe the return back to normal function?

Warning: technical jargon ahead


Aging takes a long time. It’s like fragmention of a hard disc. The return to normal is like “defragmenting” the drive. It just works better.

How many ways might chronic pain, with its impairment of innate functions, accelerate fragmentation?

Could the back to normal functions reverse some of the effects of aging? Sensations resume their normal functions. Cold hands become normal, as in warm. It takes about as long as defragging a disc.


PhotoMed’s team of nerdy software and hardware engineers uncomfortably backed into these questions. We still don’t have answers after 20 years and self-funding of $20M. Machine learning (ML) techniques uncovered a few interesting features of anti-aging. It may be that ML works more efficiently with events in real time rather than it works with slow processes.


Unexpected outcomes

The team’s task was to develop systems to test PhotoMed’s therapy with patients having “nothing worked” types of pain. Surgeries, medications, opioids, more surgeries, you get the picture.  The patient’s disorders were not expected to improve. (You can read about the therapy elsewhere on this website.)


Reversing diabetic numbness to avoid wounds and amputations.


Ending persistent post-surgical pain and “phantom” pain.


The engineers hooked up sensors and cameras to their computers. The tools recorded events in real time without the bias of a hypothesis. Most of the time nothing happened. At least there were no side effects beside disappointment.

The events were unexpected by the patient, like hands warming after 30-years of coldness.

The engineers were happy because it was so easy to detect when the therapy “works”.

Find a partial-list here.


The anesthesiologist and neurologist advisors that they consulted were puzzled by the results. The outcomes appeared to be incompatible with current medical understanding and theories.


Could machine learning help you get lucky?

Silicon Valley thinking directed the engineer’s attention to things that happened quickly. The software engineers looked for binary outcomes. “0” or “1”. Did the patient respond, or not?

The engineers built tools for Intel before starting this adventure. Andy Grove taught us that only the paranoid survive. Just seeing it with our own eyes meant nothing for our paranoid minds. So we built tools to document in real time what was going on and when.


The founder funded PhotoMed Technologies through his exit from building tools to help reduce computer chip defect rates from 1-in-a million to 1-in-a billion. Self-funding removed the pressure to come up with a short-term fix. Which drug company can you think of that would fund research for such a cheap, fast, and effective solution? At the time, drug discoveries were made through programmed learning that aims to confirm a hypothesis. Sometimes it tries a bit too hard.

The team applied machine-learning logic to looking at chronic pain. They didn’t know that others expected their machine-learning style efforts to be futile. They quickly learned that the observed outcomes conflicted with the notion that pain will remain a long time.

Machine learning recognized that the responses were not just masking pain. It seemed to be that a switch was being flipped. The improvements appeared to be deep in default systems. It was like that a hard disc was “defragged” and the operating systems went back-to-normal.

The team’s Silicon Valley roots demanded that the systems design consider the delivery cost and complexity for the user and the patient.


Without a hypothesis, it was time to be quiet. This wasn’t a science project. The task was to make the therapeutic algorithm more efficient.

Pain is the biggest contributor to premature aging. (1)

Many people have permanent “nothing worked” pain. They may have tried everything available. The attempts for relief may have introduced new sources of pain. 1-in-12 knee replacement patients wish that they did not have the operation. (2)

The “nothing worked” types of pain present a low probability of resolution, close to zero, using conventional interventions. The yield from the Triple 2 Algorithm is about 1-in-3.


The engineers were surprised when machining learning (ML) found what patients often remarked. They said, “I feel normal again”. ML also found that the “back to normal” function appears to be independent of life’s variables.


Less pain. Live longer. Live happier.




(1) https://www.apa.org/pubs/journals/releases/amp-a0035794.pdf Molton IR,Terrill AL, Overview of Persistent Pain in Older Adults, American Psychologist, Vol 69(2), Feb-Mar 2014, 197-207, DOI: 10.1037/a0035794

(2) https://www.ncbi.nlm.nih.gov/pubmed/14659522 , Harden RN, et al, Prospective examination of pain-related and psychological predictors of CRPS-like phenomena following total knee arthroplasty: a preliminary study, Pain. 2003 Dec;106(3):393-400. DOI: https://insights.ovid.com/article/00006396-200312000-00020