Photobiomodulation (PBM) is a therapeutic technique wherein light, delivered by laser or LED, is used for treating wounds and pain. The mechanism of action varies depending on the nature of the therapy which can call for different protocols regarding wavelength, intensity and application.
Many studies have explored the opportunity to leverage PBM for treating traumatic brain injury and dementia with extremely positive results. While most light doesn’t penetrate the skull, near infrared light is an exception which, depending on the region, has the potential to deliver light as deep at 50mm into the brain. Studies have tested a variety of protocols that I have reviewed; My analysis is that a band around a wavelength of 810nm is ideal, pulsed delivery is superior to constant application although the ideal frequency and duty cycle of the pulses remains unclear and surprisingly there does not seem to be a strong correlation between therapeutic effect and the intensity of the light applied (mW/cm2).
My wife, a doctor who works daily with high-end lasers in a medical setting and has first hand experience in PBM, piqued my interest on this topic. Commercial devices for brain PBM are many tens of thousands of dollars, and are also pretty inferior as far as I can tell. She asked me if I could build one for us for less, so the game was afoot!
I sourced a supply of 810nm high power (3W) LEDs and designed and fabricated an articulating head piece in which to mount them.
The geometry of each light cone was formulated based on the LED specs on power and beam angle in order to optimize for the largest surface area, minimal depth, and a power density in line with study material:
Assembled and wire run:
As you can see, there are four colors, Red is a common anode, while Blue, Green and Yellow are for each of three series of six LEDs that will run in parallel.
Now, we need to build a control box to properly power this unit!
The unit consists of an Arduino microprocessor, a high power mosfet trigger, a couple plugs and switches, an LCD display and two rotary encoders. The unit allows the user to power on/off the device and head piece as well as control the Frequency and Duty Cycle of the pulse width modulation (PWM) that will power the LEDs. Think of the Frequency as the number of pulses of light per second and Duty Cycle as how long those pulses of light are on for each pulse. If the frequency is 10hz, it means the lights flicker 10 times per second, or once every 10th of a second, and if the duty cycle is 50%, then the light is on for half that 10th of second and off the the other half. Using the super sexy dials (rotary encoders) the user can select a frequency between 2hz and 30khz and a duty cycle between 2% and 100%.
Funny story: I remember taking a walk with my wife a year or two ago and her complaining about up/down buttons and wishing everything had dials to turn things up and down, so I made sure to invest the extra time and energy to make sure this used rotary encoders to allow for super snazzy control with satisfying little pulses as you turn over the infinite range of détentes. I made a point of showcasing this to her and how I lovingly remembered her hardware interface preferences (clearly seeking well-earned praise), and it turns out she has no recollection of this conversation or any preference on controls! …on a related note, I expect that brain PBM is good for memory too! :p
With the unit together I assembled and switched it on…. I could quickly smell my limiting resistors we frying; no bueno! I did not consider the power ratings on my resistors nor the amount of heat these LEDs would throw; I really should have, but I’ve never used such high power LEDs and I always just think of LEDs as cool components. This device needed some serious heat-consciense re-engineering to not catch on fire!
I picked up some extra beefy 5W resistors (that only need to carry 770mW), aluminum backed PCBs designed specifically for high power LEDs and absolutely ridiculous heatsinks that attach to them (plus some thermal glue). The resistors I mounted inside a separate unit inline along the wire to the head piece on an aluminum heat spreader along with a 3A fuse and red indicator LED. This allows me a long disconnectable ‘extension’ cable from the control unit to the head piece as well which is nice.
As for the head unit, I needed to fabricate a special part that would augment my ‘light cones’ for the heat sink to mount to and support their enormous finned aluminum craziness. I opted to not glue the LED and heatsinks to the head unit in case I need to maintain/replace anything, so I designed my add-on piece to have tabs to use rubber bands to hold the heatsinks in place. With the light cones upgrades, LEDs cemented and soldered down to the PCBs, all wires rerun, and the heat sinks locked down, we were ready to test again!
And so here we have it, the completed device!
What’s pretty neat is that you can see it glowing purple in the picture with it on my head because the camera on my cellphone can pickup the infrared light that you can’t see with your eyes.
So I have used it a couple times and everything seems to be holding up well. I do intend to line the inside with foam around the edges to make it more comfortable otherwise I feel no affects for better or worse. While not a publicly reported benefit of brain PBM I could find anywhere in the published literature, I have to admit I did test to see if I could move objects with my mind while it was on, sadly no. Perhaps some mind reading or astral project testing is in order though, c’mon ya never know!
