Simulation-based dosimetry of transcranial and intranasal photobiomodulation of the human brain: the roles of wavelength, power density, and skin tone

Abstract

Photobiomodulation (PBM) using near-infrared (NIR) light is a novel neuromodulation technique. However, despite the many in vivo studies, the stimulation protocols for PBM vary across studies, and the current understanding of the physiological effects of PBM, as well as the dose dependence, is limited. Specifically, although NIR light can be absorbed by melanin in the skin, the understanding of how skin tones compare and how their influence interacts with other dose parameters remains limited. This study investigates the effect of melanin, optical power density, and wavelength on light penetration and energy accumulation via forehead and intranasal PBM. We use Monte Carlo simulations of a single laser source for transcranial (tPBM, forehead position) and intranasal (iPBM, nostril position) irradiation on a healthy human brain model. We investigate wavelengths of 670, 810, and 1064 nm at various power densities in combination with light (“Caucasian”), medium (“Asian”), and dark (“African”) skin tone categories as defined in the literature. Our simulations show that a maximum of 15% of the incidental energy for tPBM and 1% for iPBM reaches the cortex from the light source. The rostral dorsal prefrontal cortex and the ventromedial prefrontal cortex accumulate the highest light energy in tPBM and iPBM, respectively, for both wavelengths. Notably, we show that nominally “Caucasian” skin allows the highest energy accumulation of all three skin tones. Moreover, the 810 nm wavelength for tPBM and the 1064 nm wavelength for iPBM produced the highest cortical energy accumulation, which was linearly correlated with optical power density, but these variations could be overridden by a difference in skin tone in the tPBM case.The simulations serve as a starting point for enabling hypothesis generation for in vivo PBM investigations. This study is the first to account for skin tone as a tPBM dosing consideration. For the future of PBM research, it is important to evaluate combinations of stimulation parameters (wavelength, optical power density, pulsation frequency, duration, light source) when working to determine an optimal dosage for PBM-based therapy.

1. Introduction

Photobiomodulation (PBM) is the application of low levels of red or near-infrared light (NIR) to stimulate neural tissue [1]. Originally referred to as low-level laser (light) therapy (LLLT), it was used for wound healing and pain reduction. NIR light can penetrate human skin and tissue to various depths [1] and give rise to intracellular processes in vivo [2]. The act of shining light on the head to modulate brain function has been utilized in the field of neuroscience since 1967 [2]. Red or NIR light stimulates cytochrome c oxidase (CCO) inside mitochondria, which leads to increased enhanced CCO metabolism, nitric oxide (NO) release and elevated ATP production [3–6]. While ATP production can in theory enhance mitochondrial function, NO release can lead to vasodilation and increase cerebral blood flow. Despite its promise, the effectiveness of PBM therapy is hindered by inconsistent stimulation parameters across various clinical studies, leading to difficulties in protocol optimization and cross-study comparison. Moreover, the role of skin tone, has been largely overlooked in PBM research. The human skin tone is influenced by the presence of melanin pigment. The amount and type of melanin in the skin determine its tone, largely driven by genetic factors but may also fluctuate with sun exposure. Those with higher melanin levels tend to have darker skin, while those with lower melanin levels have lighter skin. Melanin is a natural skin pigment that absorbs light energy upon interaction, with the extent of absorption corresponding to the amount of melanin present. Simultaneously, the skin's surface can scatter light that remains unabsorbed. This scattering process alters the direction of light as it interacts with specific tissues or mediums. Both scattering and absorption are influenced by melanin content, making it an important consideration in tPBM. This is the first simulation study aimed at addressing this gap by incorporating skin tone into tPBM dosing, offering a way to develop personalized, clinically effective PBM therapy.

To more effectively penetrate the human head's multi-layered tissue, several parameters can be optimized: (1) wavelength (λ), (2) optical power density (OPd), (3) light pulsation frequency (f), (4) irradiation duration (T), (5) light source positioning. Each parameter can be adjusted to impact the overall energy accumulation in tissue. The range of parameter values commonly used in past studies is summarized in Table 1. Previous research, based on cultured cortical neurons, suggests that the peak PBM response happens when the energy induced by the light source (i.e. the fluence) reaches 0.3-3 J/cm² in the specified brain region [7].

Table 1. Past research parameter values [8–10].

Parameters of Interest	UNIT	common research values
Wavelength	λ	532 - 1070 nm
Optical Power Density	OPd	0.1 - 1000 mW/cm2
Pulsation Frequency	f	1 - 10,000Hz
Duration	T	1 - 20 minutes

As shown in Table 1, the choice of these values in the literature is highly variable. While the applications varied, a recent review on the pro-cognitive effects of PBM in healthy young adults alone reported large variations in selected stimulation parameters across studies: OPd range = 44.4-285 mW/cm², λ = 633 nm, 850 nm and 1064 nm [9]. While previous PBM simulation studies have shown large variations in stimulation parameters, none have accounted for skin tone as a contributing factor in light coefficients of absorption and scattering. Moreover, while the majority of PBM studies are transcranial (tPBM), more recently, additional areas of light source positioning have been studied, including intranasal stimulation (iPBM) [11]. It is hypothesized that iPBM can penetrate deeper into the brain than tPBM due to the access of thin and porous cribriform plate [12], and project to the hippocampal cortex [13]. iPBM has shown promise in traumatic brain injury (TBI), sleep disorders and dementia [12,14]. In the context of this work, we will use ‘tPBM’ to refer to scenarios in which the light needs to penetrate the cranial bone, and ‘iPBM’ to refer to scenarios in which the light travels through the porous cribriform plate. While full experimental validation of these simulation results are forthcoming, our interest in the effect of skin tone was also motivated by preliminary experimental findings. We have shown that tPBM delivery resulted in weaker and more inconsistent BOLD signal changes in individuals with higher melanin levels (darker skin), mirroring the reduced cortical energy delivery predicted by our models [15]. Additionally, tPBM was found to increase BOLD activity in the bilateral sensory cortex regions [16], in which individuals with darker skin tones exhibited an opposing response (suppression) to those with lighter or intermediate tones (activation). Together, these findings underscore the physiological relevance of melanin concentration in shaping the brain's hemodynamic response to tPBM.

While many studies have reported tPBM induced cognitive improvement, outcomes vary widely due to inconsistencies in stimulation parameters [6,9,17]. Studies on Alzheimer’s disease (AD) have shown varying positive effects with different wavelengths; 810 nm (increased cerebral perfusion and connectivity [18,19]), 660 nm (alleviated cognitive dysfunction and reduced microgliosis [20]), 1070 nm (microglia responses and angiogenesis promotion [21]). Similarly, study results on depression vary by wavelength and power density; 823 nm LED, with 36 mW/ cm² (antidepressant properties [22]), 945 nm (improved brain activity [23]), and contradicted by 830 nm showing no significant differences [24].

Despite the well-established understanding of the biological mechanisms and the large number of ongoing studies, the factors influencing the outcome of PBM remain incompletely understood [25]. Moreover, despite the large variations in outcome, there are approximately 300 registered PBM clinical trials on the effects of PBM on various brain diseases, including AD, Parkinson’s disease, depression, and TBI among others [26]. The current practice is to rely on existing literature for choosing stimulation parameters, with little consideration for cohort or subject-specific dose dependence. This highlights the urgent need for research aimed at systematically optimizing parameters.

This study reports a novel approach to understanding tPBM dose dependence by incorporating skin tone as a key parameter in light propagation through neural tissue. By modelling this penetration using Monte Carlo simulations on an MRI-based head model, we aim to show the differences in light delivery, based on stimulation parameters, to enhance treatment outcomes for individuals with varying skin tones. While previous research used a similar head model to characterize light penetration and dose dependence [13,27] our work details the spatial profile of light penetration, accounting for the role of melanin.

2. Methods

2.1. Monte Carlo method

The Monte Carlo technique is suited for simulating light propagation through the multi-layer tissues and has been widely used to simulate the effect of PBM in both human and rat brains [13,27–29], as well as validated by optical dosimetry probes [30]. When a photon strikes an object, its trajectory is determined by three main properties: (1) absorption, i.e. when the neural tissue absorbs the incoming photon, (2) scattering, i.e. when the incoming photon is ricocheted and diverted from its original path, and (3) transmission, i.e. when light passes through one layer of tissue into another. These properties determine the patterns of light propagation and dispersion, based on which the Monte Carlo results can be used to determine the total energy accumulation in different neural tissue regions.

For each specified power density, wavelength, and skin tone the Monte Carlo algorithm was iterated 10 times. We modelled three wavelengths 670 nm, 810 nm, and 1064 nm, which have been prominent in the PBM literature as discussed earlier. Additionally, we modelled common optical power values, which range from 100 mW/cm² to 300 mW/cm². Importantly, nominal “Caucasian”, “African” and “Asian” skin tone categories were modelled with all three wavelengths and optical densities. An overview of the experimental simulation process is shown in Fig. 1. To our knowledge, this is the first study to consider the effects of skin tone in PBM.

Fig. 1.

Overview of the simulation process experimental design (a) inputted simulation parameters, selected modelling software and atlas. b) optode source positioning indication and direction through the multi-layer tissue. c) monte carlo energy deposition map output and subsequent energy (J) percentages.

2.2. Simulation software

We used Monte Carlo Extreme (MCX) [28] in conjunction with the Colin27 head model (the Montreal Neurological Institute (MNI) in 1998)). The Colin27 atlas was generated as a stereotaxic average of 27 T1-weighted MRI scans of a single healthy individual [31], and contains approximately 56.9 million voxels, with a voxel size of 0.5 × 0.5 × 0.5 mm³, with each voxel assigned to a specific tissue type. The MCX simulated normalized energy density of 1 J, modelled with 5e⁸ photons transported through multiple layers of media as detailed in Table 2. The 5 × 10⁸ incidental photons are propelled from the source stochastically, interacting with the various media through absorption, scattering and transmission. As the photon travels, energy is lost according to the Beer-Lambert law, and deposited in the voxel from which it departs, generating a three-dimensional energy deposition map.

Table 2. Summary of coefficients related to the optical properties of different head tissues. (a) Coefficients for each wavelength, without considering the contribution of skin melanin. (b) Coefficients for each skin tone category. CSF = cerebrospinal fluid, µ_s = scattering coefficient, µ_a = absorption coefficient, n = refractive index coefficient and g = anisotropy coefficient.

a)		Tissue Type
Wavelength	Parameter	Scalp	Skull	CSF	Grey Matter	White Matter
	µs	22.65	10.82	0.091	8.40	40.10
	µa	0.056	0.021	0.0004	0.02	0.07
670nm	n	1.37	1.37	1.37	1.37	1.37
	g	0.89	0.89	0.89	0.9	0.85
	µs	17.82	17.82	0.091	7.30	38
	µa	0.017	0.011	0.0026	0.028	0.092
810nm	n	1.37	1.37	1.37	1.37	1.37
	g	0.89	0.89	0.89	0.89	0.87
	µs	14.63	14.63	0.091	5.90	30
	µa	0.019	0.019	0.0144	0.053	0.88
1064 nm	n	1.37	1.37	1.37	1.37	1.37
	g	0.89	0.89	0.89	0.91	0.88

b)		Skin Type
Wavelength	Parameter	“Caucasian”	“Asian”	“African”
	µs	21.21	21.96	14.93
	µa	0.018	0.021	0.058
670nm	n	1.37	1.37	1.37
	g	0.89	0.89	0.89
	µs	10.45	12.73	7.95
	µa	0.019	0.023	0.029
810nm	n	1.37	1.37	1.37
	g	0.89	0.89	0.89
	µs	15.26	17.05	11.08
	µa	0.014	0.017	0.028
1064 nm	n	1.37	1.37	1.37
	g	0.89	0.89	0.89

2.3. Simulation approach

2.3.1. tPBM versus iPBM

This study simulates a single source laser between the left and right hemispheres. The default MCX single source optodes sizing, the laser irritation site is the size of one voxel, i.e. 1 mm². The optode source was positioned for tPBM and iPBM simulation and applied over 5 ns. The optode location positions are shown in Fig. 2(a) and 2(b). In the tPBM setup, for the light to reach the cortex, it needs to first penetrate the scalp, the skull bone, and the subarachnoid space (containing cerebrospinal fluid (CSF)). In the iPBM set up, instead of the skull, the light penetrates the cribriform plate, which is a part of the ethmoid bone, an unpaired and porous cranial bone that contains many olfactory foramina through which olfactory nerves penetrate the nasal cavity. Whereas the frontal skull is 5-6 mm thick, the cribriform plate is only about 1 mm thick.

Fig. 2.

Sagittal slice optode positioning for (a) tPBM and (b) iPBM settings overlayed onto the Colin27 head atlas. The arrow indicates the position and direction of the flow of photons. Figure 2(c) highlights the anatomical regions of the brain that both tPBM and iPBM incoming photons must penetrate, as well as the tissue types included in the simulations.

2.3.2. Effect of optical power density

The optical power is measured in watts (W), when accounting for the area of the applied light, this parameter is restructured as optical power density (OPd), in units of mW/cm² (Table 3) [32]. Existing literature cites OPd values ranging from 100 to 300 mW/cm², with 250-300 mW/cm² being most common in in-vivo human research [9]. The skin-exposure safety limit of near-infrared lasers ranges up to 500 mW/cm² for Class 3B and 4 lasers, depending on the wavelength and type of laser [33]. A 2019 publication demonstrated that lower OPds can be effective therapeutically without brain heating or damaging effects [33].

Table 3. Parameters of Interest.

Parameters of Interest	Symbol	Units
Wavelength	λ	nm
Optical Power Density	OPd	mW/cm2
Skin Tone	-	Melanin μg/mL
Energy	E	Joules

As mentioned earlier, the MCX software assumes a normalized incidental total light energy of 1 J. However, to translate the results to more realistic OPd values, including 100, 200 and 300 mW/cm², we calculated power scaling factors for MCX. Specifically, the total incidental optical energy is equivalent to the optical power density multiplied by application time. Thus, the default 1 J incidental energy was divided by a circular area of 1 cm², and then scaled to an application duration of 60 seconds. Thus, the default power density was 1J/60s/1 cm² = 16.7 mW/cm². A scaling factor was then generated and multiplied to the default optical power density and energy values, to scale the MCX results to reflect OPd values of 5, 7, 9 mW/cm² for iPBM, and 100, 200 and 300 mW/cm² for tPBM. The energy deposition (E) over the 60 s window based on each power density level is then normalized by the volume of a given ROI to produce the energy deposition in terms of J/cm³.

2.3.3. Effect of wavelength

In general, a shorter wavelength, usually in the visible range between 635-700 nm (red light), has shorter penetration than a longer wavelength in the near-infrared range (810-1070 nm) [33]. Many clinical studies utilize an 808 nm laser or 810 nm LED, reported by some studies to produce the deepest penetration and highest cognitive impact in subjects with impaired cognition [19]. However, longer and shorter wavelengths have also been used.

The effect of light wavelength is modelled using known scattering, absorption, anisotropy and refractive index coefficients, summarized in Table 2 for 670 nm, 810 nm and 1064 nm, which are commonly used wavelengths in the PBM literature, and which have previously been investigated in simulations [34]. When comparing across wavelengths, a default power density of 100 mW/cm² and 5 mW/cm² were assumed for tPBM and iPBM, respectively. A power scaling factor was calculated as described earlier.

2.3.4. Effect of skin tone

Skin tone was only considered relevant for tPBM and is driven mainly by melanin production. Melanin, specifically eumelanin, is a substrate produced in the innermost layer of the epidermis, and the main variable in characterizing skin tone. Broad inter-racial differences can be associated with diversity in skin tone, and the concentration of epidermal melanin is double in darker skin types compared to light-pigmented skin types [35]. When light strikes melanin, the pigment molecules within melanin absorb the photons, and skin with higher melanin content absorbs more light, preventing deeper penetration, which is an important consideration for tPBM.

The scalp consists of the skin (i.e. epidermis, dermis, the subcutaneous tissue) and an inner layer (i.e. galeal, subgaleal and periosteal layers). To accommodate variations in melanin concentrations, we adopted three distinct skin tone categories: “Caucasian”, “Asian”, and “African”, the optical properties of which were measured using diffusion optical spectroscopy in prior work [34]. The coefficients related to the optical property of each category were extracted from previous Monte Carlo simulation publications [28,34] where the scalp absorption values were extrapolated using the following equation. The values are summarized in Table 2b.

$${\mu }_{a(scalp)}=0.3\ast {\mu }_{a(skin)}$$ (1)

For each parameter setting in each of the categories, we computed the energy deposition profile over a 4 cm axial slab directly adjacent to the position of the light source. We additionally computed the percent energy deposition for each parameter setting as a fraction of the incidental laser energy. In-house Matlab (Mathworks, Natick, USA) scripts [36] were used for compiling across multiple iterations of the simulations as well as for visualizing the energy deposition contours and penetration profiles. Finally, based on these profiles, we identified the brain regions of the Colin27 atlas into which the most energy was deposited.

3. Results

3.1. Transcranial photobiomodulation (tPBM)

The modelled tPBM penetration profile on the Colin27 brain atlas is shown in Fig. 3. The distance between the forehead skin surface and neural tissue is approximately 1-2 cm [8]. Based on Monte Carlo simulations, Fig. 3 displays the modelled energy-deposition profile of the laser, which is further condensed into a penetration profile plot (Fig. 4). According to a sample axial profile derived from the simulations (wavelength = 810 nm), the light energy declines rapidly as it enters the brain (Fig. 4). In this case of tPBM, the peak light energy declines sharply over the 5-8 mm that accounts for the skull and sub-cranial tissue layers; light fails to penetrate beyond a depth of 15 mm.

Fig. 3.

Transcranial PBM irradiation configuration and key regions of interest (ROIs). The yellow arrow represents the laser location, and the white line indicates the axial slice location. The sagittal and axial views of the cortical parcellations are shown in (a) and (b), respectively. The regions receiving the highest energy deposition according to the profile are labelled. The codes correspond to the colour codes used to identify these brain regions in the atlas. Moreover, a sample axial view of the energy deposition profile is shown in the right bottom corner, in which the axes represent the voxel dimensions, and the colour scale represents the log of energy levels.

Fig. 4.

Energy depth penetration profiles for transcranial PBM (tPBM) as a function of power density. In this plot, a wavelength of 810 nm and the MCX software’s default scalp parameters are assumed. a) The light penetration profile is shown as a function of increasing distance from the light source. Energy (Joules) shown on the y axis for each of the three optical power densities, and light penetration depth on the x axis, where a distance of zero is defined as the position of the tail end of the arrow shown in Fig. 2. b) Percent incidental energy accumulated at the specified penetration depth. Error bars are representative of the standard deviation across the 10 Monte Carlo simulations.

Empirically, recent studies applying a 5 W laser, with a 30 mm diameter beam size (708 mW/m²) at 808 nm on cadaver heads showed that the light can penetrate up to 4 cm from the skin surface (penetrating through the scalp, skull, meninges, and brain) [37]. The increased penetration distance in this cadaver model compared to our simulation (Fig. 4) could be influenced by the difference in beam (disc) size and number of beams (1 vs 9). The cadaver study’s optical power density (OPd) of 708 mW/cm² exceeds our simulated range (100–300 mW/cm²). Energy deposition scales linearly with OPd, hence the increased depth of 4 cm vs 1.5 cm.

The age of the individual also plays a role in the effect of photobiomodulation, the modelled brain template, Colin27 was 27 averaged images of a 28-year-old, in contrast to the age of the cadaver brains - 66-97 years of age. Aging can cause immense changes in the brain, especially in regard to cortical thickness, skull thickness [37] and dermal thickness, but this is secondary to the difference in optical power densities. Lastly, significant changes occur immediately after death, especially in the absorption bands, primarily due to blood deoxygenation and alterations in blood concentration within the tissues [38]. Additionally, the energy depth penetration profile shown in Fig. 4 displays how far individual photons can travel and deliver energy.

3.1.1. Skin tone dependence

Our results show that those with “Caucasian” skin tone category accommodate a higher tPBM energy accumulation to brain ROIs, followed by those of the “Asian” skin tone and of the “African” skin tone. Our results also show a wavelength-skin-tone interaction in energy accumulation (Fig. 5). The skin-tone dependence in the fractional energy deposited is highest for 810 nm wavelength, and lowest for 1064 nm. Furthermore, 810 nm is associated with the highest peak percent energy deposition, up to just under 16% for the “Caucasian” skin category (Fig. 5(c)).

Fig. 5.

Energy accumulation in key ROIs for tPBM as a function of wavelength and skin tone type, assuming wavelengths of 670 nm, 810 nm, and 1064 nm, and an optical power density of 100 mW/cm². (a-c) The percentage of incidental energy accumulated in each region of interest. The total energy in each region of interest after 1 minute of stimulation (plotted on log scale). (d-f) The total energy in each region of interest after 1 minute of stimulation (plotted on log scale) Note: 1-ROI: Rostral Dorsal Prefrontal, 2-ROI: Rostromedial Prefrontal, 3- ROI: Rostral Dorsolateral Superior Prefrontal, 4-ROI: Anterior Cingulate, 5 -ROI: Rostral Dorsolateral Inferior Prefrontal.

3.1.2. Optical power dependence

Our results show a linear response to an increase in OPd, as shown in Fig. 6.

Fig. 6.

Optical power Plot energy accumulation of transcranial PBM (tPBM) as a function of power density (OPd) assuming a wavelength of 810 nm wavelength and the MCX software’s default scalp parameters. a) The total energy in each region of interest (ROI) after 1 minute of irradiation (plotted on log scale). b) The percent of incidental energy accumulated in each ROI, where all three power levels coincide.

Note: 1-ROI: Rostral Dorsal Prefrontal, 2-ROI: Rostromedial Prefrontal, 3- ROI: Rostral Dorsolateral Superior Prefrontal, 4-ROI: Anterior Cingulate, 5 -ROI: Rostral Dorsolateral Inferior Prefrontal.

3.1.3. Wavelength dependence

As shown in Fig. 7, the 810 nm light produced the deepest transcranial penetration.

Fig. 7.

Energy accumulation over key ROIs for transcranial PBM (tPBM) as a function of wavelength, assuming a power density of 100 mW/cm² and MCX software’s default scalp parameters. a) total energy in each region of interest after 1 minute of stimulation (plotted in log scale). b) The percentage of incidental energy accumulated in each region of interest. In both (a) and (b) 1064 nm percent incidental energy is shown in light blue, which is overlapped by the purple 670 nm, in regions of interest (a) 3 and 5 and (b) 1 and 3.

Note: 1-ROI: Rostral Dorsal Prefrontal, 2-ROI: Rostromedial Prefrontal, 3- ROI: Rostral Dorsolateral Superior Prefrontal, 4-ROI: Anterior Cingulate, 5 -ROI: Rostral Dorsolateral Inferior Prefrontal.

3.2. Intranasal photobiomodulation (iPBM)

The modelled iPBM penetration profile on the Colin27 brain atlas is shown in Fig. 8. Our modelling of the propagation of light through the nasal cavity shows that the photon dispersion can lead to a near whole brain response, including the hippocampus.

Fig. 8.

Nasal irradiation configuration and key ROIs. The yellow arrow represents the laser location, and the white line indicates the axial slice location. The sagittal and axial views of the cortical parcellations are shown in (a) and (b), respectively. The regions receiving the highest energy deposition according to the profile are labelled. The codes correspond to the colour codes used to identify these brain regions in the atlas. Moreover, a sample axial view of the energy deposition profile is shown in the right bottom corner, in which the axes represent the voxel dimensions, and the colour scale represents the log of energy levels.

According to a sample axial profile derived from the simulations (λ = 1064 nm), the light energy declines rapidly as it enters the nose (Fig. 9). In the case of iPBM, the deposited light energy climbs steadily as the light enters the nasal cavity, being deposited in surrounding non-brain tissue. Subsequently, light deposition is reduced at the cribriform plate and in the frontal sinus (about 20 mm from the light source). However, the light deposition subsequently penetrates the thin cribriform plate, reaching about 35 mm from the light source before declining.

Fig. 9.

Energy depth penetration profiles for intranasal PBM (iPBM) as a function of power density. Varied optical power densities at 810 nm wavelength. a) Light penetration profile as a function of distance from the light source. b) Percent incidental energy accumulated at the specified penetration depth. Error bars represent 1 standard deviation across the 10 Monte Carlo iterations.

3.2.1. Optical power density dependence

Similar to the effect of tPBM on optical power density, our results show a linear response to an increase in OPd, as shown in Fig. 10. The 9 mW/cm² power elicited the largest energy accumulation in comparison to the two smaller optical power density values.

Fig. 10.

Energy accumulation over key ROIs for transcranial PBM (tPBM) as a function of power density, assuming an 810 nm wavelength and MCX software’s default scalp parameters. a) total energy in each region of interest after 1 minute of stimulation (plotted in log scale). b) percent incidental energy accumulated in each region of interest, where all three power levels coincide.

Note: 1 - ROI: Ventromedial Prefrontal, 2 - ROI: Ventromedial Orbitofrontal and 3 - ROI: Ventral Orbitofrontal.

3.2.2. Wavelength dependence

Unlike in tPBM, light transported through the nasal cavity interacts with a significant portion of air when it passes through the frontal sinus. As shown in Fig. 9, the energy deposition drops significantly around 15 mm of depth penetration; this drop is hypothesized to be the area of the frontal sinus. Our results (Fig. 11) show that the 1064 nm light produced the highest percent energy deposition in the brain.

Fig. 11.

Energy accumulation over key ROIs for transcranial PBM (tPBM) as a function of wavelength., assuming a 5 mW/cm² optical power density and MCX software’s default scalp parameters. a) Total energy in each region of interest after 1 minute of stimulation (plotted on a log scale). b) The percent of incidental energy accumulated in each region of interest.

Note: 1 - ROI: Ventromedial Prefrontal, 2 - ROI: Ventromedial Orbitofrontal and 3 - ROI: Ventral Orbitofrontal.

4. Discussion

The Monte Carlo simulation method was first developed to improve decision-making under uncertain conditions [39]. To this day, Monte Carlo simulations have a variety of applications in many mathematical and scientific fields. Assigning a sporadic random value to an uncertain variable in a problem and calculating the result numerous times is the basis behind this mathematical computation. For many linear and complex problems, this simulation technique is considered the gold standard when accounting for unknown variables. However, additional consideration is required when using this technique to model computations related to the human body, especially to that of the brain.

The brain is the most complex organ in the human body [40], consisting of neural tissue, neurons (nerve cells), and glial cells, the brain works as the control system to the human body. Over the last decade, many publications have worked to show the impact of light penetration into the brain through Monte Carlo simulation. To properly model this complex light delivery, several parameters are required, including scattering, absorption anisotropy and refractive index coefficients. These four measurements are provided for each tissue type of the brain: scalp, skull, cerebral spinal fluid, grey matter and white matter. This simulation technique models the linear path of light through each tissue, based on the calculated coefficients. The results of this paper show the distinctions between transcranial and intranasal placements, showing the trajectory of light penetration into neural tissue from forehead and nostril laser placements. While expanding on previous simulation results [13,27], by presenting regional energy accumulation, we also highlight the importance of skin tone as a key determinant of light penetration in tPBM.

4.1. Skin tone

Melanin, the substrate in the human skin that produces skin pigmentation, is the main variable in characterizing skin tone. Human skin can have a vast range of melanin levels, ranging from very low in light complexion; “Caucasian” skin (type I), to very high in black “African” skin (type VI) [41]; that is, “Caucasian” skin is generally known to contain less pigmentation than “African” and “Asian” skin types. This study takes the first important step of accounting for the effects of skin tone, and skin pigmentation with PBM, only relevant for tPBM. In this study, we show that although 810 nm deposited the highest percentage of optical power in the cortex for all skin tones, “Caucasian” skin tone was associated with a higher overall energy accumulation than other modelled skin tones. However, these skin tone categories are only names that derive from very limited literature and are not meant to be fully generalizable to races or geographical regions. It is important to recognize that in each of the skin tone categories, there is in reality a wide range of melanin levels. Moreover, not only do skin melanin levels vary across races and ethnicities, they also vary with sun exposure, thus forming an important consideration in personalized PBM.

Our simulations demonstrate that lighter skin allows for greater energy accumulation, with “Caucasian” skin tones achieving up to 16% of incidental energy reaching the cortex. As shown in Fig. 5 for 810 nm light, in the rostral dorsal prefrontal cortex, the lightest skin tone was associated with 16% energy accumulation, while the darkest skin tone was associated with only 6%, which implies that to achieve a similar level of energy deposition as for lighter skin, more than twice the incidental light energy would be required. This enhanced light penetration of lighter skin tones could be attributed to lower levels of eumelanin, consistent with studies showing that the reduced eumelanin leads to less absorption of light in the epidermis [42]. Overall, these findings highlight the importance of considering skin tone when optimizing PBM therapies, as lighter skin can enhance treatment efficacy by improving energy delivery to targeted areas.

4.2. Wavelength

Wavelength determines the distance of which light can travel through the brain. A shorter wavelength, usually in the visible range between 635 - 700 nm (red light), is said to have shorter penetration than a longer wavelength in the near-infrared range (810 - 1070 nm). As shown in our simulations, for tPBM, the 810 nm wavelength produced significantly higher fractional energy deposition than all other wavelengths investigated herein, consistent with previous findings by Cassano et al. and Li et al. using Monte Carlo simulations [13,43]. The penetration profile resembles that observed for rodent brain [44], and the peak energy deposition of ∼16% (for “Caucasian” skin) is also similar to reported by Cassano et al. [13], although in our case we are simulating a single light source instead of an array. For iPBM, on the other hand, the 1064 nm wavelength produced the highest energy accumulation, although Cassano et al. previously reported 810 nm depositing the highest energy for the mid-nose position [13]. The difference between the modelled penetration profiles for tPBM and iPBM derive from the difference in optode source positioning and the tissue types that they penetrate. With a forehead tPBM the light must first penetrate through the scalp, then skull, followed by CSF, before reaching the neural tissue. The advantage of the 810 nm wavelength in tPBM may be due to it being more distal to the CSF absorption peaks of ∼450 nm [45]. However, for iPBM, the light passes into the nasal cavity and through the porous and thin cribriform plate and the frontal sinus. Not only is the cribriform plate much thinner than the skull (as mentioned earlier), there is also minimal CSF in the path based on the head model, which may be the reason that iPBM is optimized by the longer 1064 nm wavelength. Moreover, as shown in the light penetration profiles in Fig. 4 and Fig. 9, the energy drop-off with penetration distance is significantly steeper for tPBM compared to iPBM. However, when averaged over specific ROIs, tPBM deposited higher energy in the rostral dorsal prefrontal (15% - 810 nm, 14% - 1064 nm) and rostromedial prefrontal (10% - 810 nm, 6% - 1064 nm) compared to the maximum energy deposition of 1% of the incidental energy for iPBM in the ventromedial prefrontal ROI (Fig. 6 and 11, respectively). However, we are aware of limitations in the accuracy of tissue optical properties measured for wavelengths near 1000 nm. Research at this specific wavelength is less common compared to shorter wavelengths such as 810 nm and 830 nm [46].

In comparison to experimental studies, researchers have examined the penetration of near-infrared (NIR) light through human skin and underlying tissues, revealing that different wavelengths have varying degrees of effectiveness. For instance, 633 nm light showed 78% transmittance through 0.4 mm of epidermis, decreasing to approximately 5% at 2 mm, with no penetration beyond 3 mm [47]. This is consistent with our results in energy drop-off at approximately 5-8 mm (tPBM) and 12-15 mm (iPBM) of depth [47].

4.3. Optical power density

Optical power density is the energy of light, per unit time, delivered to a specific area. The laser power density has a direct impact on the overall energy accumulation in targeted areas of the brain. To optimize energy accumulation, additional Monte Carlo simulations were run to review the impact of optical power and determine the peak value. The energy accumulation results of the Monte Carlo simulation show a linear response to an increase in optical power density. This linear relationship aligns with the review findings of Hamblin, Selting, and Zein, who present a comparable graph of optical power density versus energy accumulation [10]. Their analysis indicates a linear correlation between the two variables at lower power levels (1-200 mW/cm²) [10].

By modelling the light propagation through the human head and scaling to prominent optical power density values, it dictates the total energy accumulated in specific brain regions and its relationship to the OPd parameter. Although this knowledge is crucial for targeting brain areas, it does not consider the metabolic or functional connectivity effects of the incoming light. Previous literature has suggested that the brain’s response to photobiomodulation is biphasic, such that the CCO response is maximized at an intermediate OPd. However, this observation may hinge on physiological processes that cannot be modelled using Monte Carlo simulation.

The current mechanisms of action of PBM are unclear. Recent literature states that incoming red or NIR light interacts with specific cell photoreceptors which can affect cellular pathways [48]. This relationship has been demonstrated in both in vitro cell cultures and human studies, when investigating the ATP and mitochondrial membrane potential upregulation in response to photobiomodulation. Many studies point to the oxidation of CCO, stating that the incoming light photon dissociates an inhibitory (NO molecule allowing the CCO to be oxidized and form ATP). Additionally, the dissociated nitric oxide molecule is a known vasodilator, therefore causing an increase in blood flow due to dilation of blood vessels. This mechanism of action has been discussed in many publications over the past several decades [49,50]. However, more recently, in-vitro studies have shown that CCO has no impact on ATP production, that cells without CCO present still have an upregulation of ATP following NIR light administration [51]. Other areas of interest include Ca²+ channels, K + channels and reactive oxygen species (ROS). With these uncertainties in PBM mechanisms, the Monte Carlo method is beneficial for modelling the light penetration through the multi-layered tissue of the human head. However, the metabolic and hemodynamic effects remain indistinguishable in this model.

4.4. Implications for human PBM

For the future of photobiomodulation research, it is important to evaluate a multitude of stimulation parameters (wavelength, optical power density, frequency, duration, light source) when working to determine an optimal dosage for PBM therapy. This process is not only crucial for optimizing cognitive outcomes but also allows for comprehension of the wide variations observed in PBM outcomes across diverse studies.

In PBM research, it is apparent that many studies use stimulation parameters that vary significantly. This heterogeneity leads to inconsistent results, making it challenging to replicate findings. Therefore, there is a need to standardize the photobiomodulation parameters to compare across studies, improve the reliability of outcomes, and ultimately advance the understanding of the therapeutic potential of PBM.

Additionally, it is important to understand that the Monte Carlo technique inputs the brain as a static object of several tissue types, therefore disregarding the on-going metabolic or hemodynamic processes. To gain a more comprehensive understanding of how PBM influences the living and dynamic brain, in vivo experiments are warranted. Utilizing imaging techniques like functional magnetic resonance imaging (fMRI) could aid in assessing the accuracy of the Monte Carlo method in reflecting real human subjects.

4.5. Limitations

Our study presents several limitations. First, we only probed the forehead tPBM and nostril iPBM positions for the light source. In practice, the positioning of both transcranial and intranasal approaches can vary across studies. An additional limitation of our study revolves around the limited availability of optical coefficient parameters beyond 1000 nm. Within the 650 to 850 nm range, melanin absorption dominates, potentially overshadowing weaker absorption traits at higher wavelengths. Additionally, as we delve into the near-infrared (NIR) spectrum past 1000 nm, water starts exhibiting robust absorption characteristics. This added absorption from water presents a competing factor against skin absorption, thereby complicating result interpretation [13,34].

Secondly, our approach did not allow more comprehensive consideration of inter-subject variations. First, we were only able to identify one previous publication that provides documentation on skin attenuation and scattering for various skin types. These values do not define each of the three races (“Caucasian”, “Asian” and “African”), but are rather included as representative values. Skin tone varies widely across each racial category and throughout different seasons, emphasizing the importance of precise measurements rather than relying on broad racial categorizations. Moreover, we did not include sex-related or age-related differences, such as that in cranial thickness [42], into our modelling consideration.

Moreover, there are inherent limitations in the modelling approach, as modelling is almost never truly realistic. For example, the tissue types we modelled were limited by the atlas available in the MCX software. While the main cranial tissue types are included, it was not possible to model tissues such as the mucosae were not included. Likewise, while lasers have conventionally been used for PBM, LEDs are more easily accessible and cheaper and have become more dominant commercially. However, although more convenient, LEDs produce a broader spectrum of light, generating a less focused beam and scattering more when contacting the multi-layer tissue of the human head. The behaviour of LEDs are not addressed by this study, although the findings could be extrapolated to LEDs to some degree.

Furthermore, our simulation methodology does not enable us to investigate the effects of laser pulsation, blood flow, and other hemodynamic processes. PBM's impact on dynamic physiological processes is a limitation in PBM Monte Carlo simulations in general. In particular, the flow of neurofluids such as blood and cerebrospinal fluid can alter photon absorption and scattering patterns. The interactions between PBM and cerebral blood flow emphasizes the need for comprehensive simulations that accurately capture the hemodynamic process, thereby enhancing our understanding of the extent to which light can penetrate through the brain.

4.6. Uncertainties and future directions

Despite the accumulation of publications supporting the mechanisms of PBM, uncertainties remain in understanding the underlying biological effects of its clinical benefits. While our Monte Carlo approach can simulate photon propagation, which is the fundamental element in PBM having any effect at all, the interactions between the photons and the biological system remain incompletely understood.

First, based on our results, in vivo experiments will be needed to confirm the effect of different skin tones, textures, thicknesses and melanin content on the tPBM response. For instance, NIR light penetration through different tissue layers can also be sex- and age-dependent, as the tissue-layer composition and the thicknesses of different layers can vary by sex and age [52].

Beyond tissue composition, factors such as hormonal balances and age-related changes to general cellular processes [53] can also influence PBM efficacy. Breaking down the individual level of target cells, such as how CCO activity varies between individuals, and how its influences light absorption and key cellular responses, could be an important step toward personalized PBM treatment.

Potential interactions between wavelength and power density: Moreover, our current Monte Carlo simulations model the different wavelengths propagation through tissue, however, predicting power density changes that vary with wavelengths is a limitation. The power density and wavelength interaction is most likely non-linear, where specific combinations can trigger different biological mechanisms in the mitochondrial or vascular response [15]. Additionally, current Monte Carlo simulations tend to assume continuous wave light sources, however pulsed photobiomodulation is commonly used in experimental and clinical settings. Pulsed light can induce unique several biological effects which are not yet fully understood [11].

The existence of neuronal inhibition (which runs counter the current understanding of the primary mechanisms that focus on neuronal excitation and vasodilation): Additionally, while the primary known mechanisms of PBM focus on pro-metabolic changes through the CCO, studies have also looked at PBM-induced neuronal inhibition, with findings that can complicate the current understanding [54]. While our simulations track the propagation of photons through tissue, the interactions between pro-metabolic excitatory activity and concurrent inhibitory activity cannot be explained through the simulations, and remain to be understood.

The effect of PBM beyond neurons and vasculature (e.g. CSF): Furthermore, simulation-based prediction focuses on tissue penetration as the primary predictor, but the brain is a dynamic system, especially the neurofluids. There is growing evidence regarding the effect of PBM on the dynamics of the cerebrospinal fluid (CSF) and glymphatic glial cells, myelin and immune responses [55]. The brain is the most complex organ in the human body, each element plays an essential role and PBM can potentially modulate their effects. CSF, in particular, is increased during sleep within the brain's glymphatic system, clearing metabolic waste. Recent work has shown that PBM can influence CSF, and therefore the sleep state of the individual could play a role in the effect of this neurofluid modulation [55]. PBM timing could interact with circadian rhythms and glymphatic activity, which could provide another avenue for optimizing the therapy.

The effect of sleep and brain state on PBM response: Finally, the physiological response to PBM may also be brain state and vigilance dependent [55]. The cellular processes underlying this dependence cannot be modelled with simulations of this kind, and will need in vivo investigations.

5. Conclusion

Our research represents an important addition to the existing knowledge of light penetration and propagation through the multi-layer tissues of the human head to reach the cerebral cortex. The simulations indicate that even in the cortical region closest to the light, no more than 15% of the incidental energy is deposited into brain tissue. This is higher than previously reported using cadaver heads and skull fragments [19], possibly due to major differences in hydration and other tissue properties between living and cadaver tissues. These physical simulations do not account for dynamic physiological processes that may modulate PBM response in vivo. Thus, these results serve as a starting point to establish the physical baseline of light penetration and factors that modulate it. This study accelerates the PBM field by introducing a novel approach to understanding how skin tone influences tPBM. Existing research has focused on generalized models of light absorption and tissue penetration, this study emphasizes the crucial role of skin tone in modulating the effectiveness of PBM. By directly addressing how varying skin tones impact the ability of light to penetrate the cranial tissues, our findings pave a way for personalized PBM protocols based on individual skin characteristics.

Funding

Natural Sciences and Engineering Research Council of Canada 10.13039/501100000038 ( NSERC ALLRP 576161-22).

Disclosures

Lew Lim is the CRO and the shareholder of Vielight Inc. Nazanin Hosseinkhah is the director of special projects of Vielight Inc. Paolo Cassano is the Cofounder, chair of scientific advisory board, board member and stock owner of Niraxx Inc.

Data Availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.