Transcutaneous Transmission of Light of Photobiomodulation Therapy Wavelengths at 808 nm, 915 nm, 975 nm, and 1064 nm to the Spinal Canal of Cadaver Dogs

Featured Application

Featured Application: The results presented inform a parametric wavelength at which photobiomodulation therapy of spinal cord injury could be most viable for clinical testing.

Abstract

This study compared the transcutaneous target level irradiances from the thoracic to lumbar segments of the interior spinal canal in three cadaver dogs, measured for light at four wavelengths (808 nm, 915 nm, 975 nm, and 1064 nm), common in photobiomodulation therapy (PBMT). Intra-spinal irradiances at nine sites spanning approximately 8 cm in length were measured using a flexible intra-spinal probe under surface application of continuous-wave (CW) light with powers ranging from 0.5 W to 2 W. Surface illumination was applied using an acupuncture treatment head in three modes: non-contact with skin removed, non-contact with skin intact, and contact with skin intact. During surface application, the treatment head was positioned over the spinal canal near the 13th vertebrae (T13, surface site 1), and approximately 4 cm (surface site 5) and 8 cm (surface site 9) caudal to T13. At each position of the treatment head, the light was multiplexed among the four wavelengths at the same power setting. In all three modes of surface application, the target level irradiance at the 1064 nm wavelength was significantly greater than that at the other three wavelengths (p ≤ 0.0017). At a surface irradiance of ~157 mW/cm², corresponding to 0.5 W light applied with the treatment head directly in contact with the skin, the intra-spinal irradiance at 1064 nm reached 0.137 ± 0.095 mW/cm². Obtaining a dosage of PBMT-associative wavelengths of this magnitude at the level of the spinal canal may guide focused research into the transcutaneous applicability of PBMT for spinal cord injuries.

1. Introduction

Traumatic spinal cord injuries (SCIs), degenerative spinal cord diseases, and chronic neuropathic pain are neurological lesions that negatively impact the function, wellness, and overall quality of life of both humans and animals. The management of spinal cord disease and injury generally involves a combination of medical and surgical interventions, along with intense rehabilitation to optimize functional outcomes [1]. The goal of spinal cord intervention is to “rescue, reactivate, and rewire” the neuronal network. Tissue rescue aims to prevent damage beyond the primary site of injury and is achieved through surgical methods or modalities targeting inflammation [2]. Tissue reactivation involves using spared systems to stimulate spinal networks or re-myelinate denuded axons [3]. Tissue rewiring focuses on treatments that promote the regrowth of damaged axons or the adaptation of spared ones [3]. Ongoing trials are exploring alternative treatment strategies to mitigate the effects of spinal cord injuries and disease. These strategies include anesthetics, neurosurgical procedures, psychotherapy, and physiotherapeutic resources [2]. For acute SCI, where delays can cause additional trauma, there are few effective, simple, and low-risk or risk-free interventions suitable for pre-hospital care or field-care settings. The specific challenges of SCI necessitate novel approaches to treatment.

Photobiomodulation therapy (PBMT) is a non-invasive treatment which has been widely investigated in both human and veterinary medicine. It shows promise in improving outcomes in SCI. PBMT has been demonstrated to reduce inflammation by lowering fibrinogen levels, edema, and the presence of inflammatory cells [3,4]. In transcranial applications, PBMT has been reported to improve cerebral neurological function by ameliorating mitochondrial dysfunction and modulating apoptosis [5]. It also maintains mitochondrial survival by enhancing the antioxidant defense system [6]. Extrapolating these physiological effects, PBMT studies in neuronal regeneration have shown stimulation of axonal regrowth in the spinal cord following acute structural damage in rodent models [7]. Although the exact cellular mechanism for axonal regeneration is not fully understood, it is theorized to result from the inhibition of inflammatory cells. The inhibition alters the extra-cellular environment, creating a local milieu that may encourage axonal regrowth [1]. By reducing inflammation, PBMT also helps to reduce pain, potentially providing additional analgesia through increased synthesis of endorphins [6]. Several other studies have also suggested additional mechanisms that may govern PBMT’s promotion of analgesia [8,9].

To treat the spinal cord and surrounding tissues effectively, PBMT treatment dose must penetrate deeply without causing collateral damage to the surface or target tissues. The anatomical safeguards, such as heavy muscle cover and tall vertebral processes, make delivery of PBMT dose to the spinal cord exceedingly challenging. Significant attenuation of the treatment dose occurs as PBMT light passes through tissue and bone. The skin-to-spine dose attenuation necessitates a higher entrance power to deliver an effective irradiance dose to the spinal cord. High-intensity PBMT has been advocated [10]; however, its effectiveness must be carefully assessed [11] due to concerns about collateral thermal side effects from the higher intensity of light [12]. Class IV lasers, which provide longer wavelengths (up to 1000 nm) with higher power outputs, can activate potentially therapeutic cellular metabolic changes in deep tissue [11]. Nevertheless, they have proven ineffective if the terminal irradiance (or dose) is below a therapeutic threshold.

Multiple wavelengths of PBMT have demonstrated varying therapeutic effects in SCI in small animal models. These wavelengths include 655 nm [13], 660 nm [14], 780 nm [15], 810 nm [16], 808 nm, and 905 nm [17]. A limited number of studies have reported the wavelength dependency of the PBMT light reaching the target level of bone. In 1994, the wavelength dependence of light transmission through the skull bone was reported for incident radiation [18] at 476.5 nm, 488.0 nm, 501.7 nm, 514.5 nm, and 632.8 nm, respectively. It appears that light transmission increases with wavelength, suggesting that a greater target level irradiance can be achieved with longer wavelengths.

Although this information, combined with computer simulations, could help determine the best candidate wavelength for greater PBMT spinal cord transmission, the complexities of anatomical structures and the variability of tissue optical properties limit its usefulness. These complexities are further confounded by the lack of an adequate library of tissue properties and the likely changes in optical properties induced by light via photothermal effect at the level of power necessary for deep transmission.

The light transmittance through tissue is dictated by two collective properties: scattering and absorption. In the visible to near-infrared (NIR) spectrum, tissue scattering decreases as wavelength increases, leading to greater line-of-sight penetration at longer wavelengths. However, absorption of light by tissue constituents, such as water and hemoglobin, has complex spectral profiles, with increased absorption occurs as wavelength increases from visible to NIR. Identifying the wavelength that provides maximal target level transmission through deep tissue at a tolerable surface dosage should be experimentally examined. Such an examination requires dosimetry investigations relevant to the clinical setting. Measurements of light transmission from the skin to the spinal cord level conducted with small animals are not easily translatable to human use. Therefore, quantification of the intra-spinal irradiance should be conducted in an anatomic structure of an animal much closer in size to humans.

Towards the goal of intra-spinal target-level dosimetry of PBMT for SCI, our group developed a flexible nine-channel photodetector probe [19], measuring over an 8 cm length, designed for intra-spinal dosimetry. Using this intra-spinal probe, we measured the transcutaneous transmission of 980 nm PBMT light to the level of the spinal canal along the thoracic to lumbar segment in six cadaver dogs [20]. Our measurements indicated that non-contact transmission of 980 nm with skin intact was as low as 12% of the non-contact transmission without skin. Additionally, contact application increased transmission by up to 67% compared to non-contact application. Under a maximal surface irradiance of approximately 3.14 W/cm², resulting from 10 W of laser power applied via an acupuncture treatment head, the mode of contact application transmitted a maximum irradiance of 85.4 ± 139.1 μW/cm² to the spinal canal. This information highlights the need to clinically consider the impact of skin transmission and contact application techniques when attempting to treat spinal cord disease with PBMT.

In this study, we address another aspect of the transcutaneous transmission of PBMT light: the wavelength dependency of the transmission. We used the same intra-spinal dosimetry probe in cadaver dogs to measure whether PBMT light of different wavelengths transmitted different amounts at the spinal cord level under the same conditions. The experiments were carried out on three cadaver dogs and are reported in detail in the following sections.

2. Materials and Methods

2.1. The Flexible 9-Channel Dosimetry Probe at Four PBMT Wavelengths

The fabrication of the flexible nine channel probe, its interfacing configuration, and the calibration method for assessing the responsiveness of its nine channels over five orders of magnitude of irradiance at a single wavelength (850 nm) using a thermoelectrically cooled laser with precise power control are detailed in [19]. Briefly, the flexible probe measures intra-spinal irradiance at four different wavelengths: 808 nm, 915 nm, 975 nm, and 1064 nm. It operates under a surface illumination power of up to 10 W, available from two combined research-laser units. To accurately compare the irradiances across these four wavelengths, spanning nearly 260 nm, the probe’s spectral responses must be compensated. Here, we reiterate some fabrication parameters of the 9-channel probe. This facilitates the subsequent discussion on the methods necessary for examining the inter-channel comparability of the measured irradiances and for compensating the irradiance responses of the dosimetry probe across the four wavelengths from 808 nm to 1064 nm.

2.1.1. The Flexible 9-Channel Probe and the Inter-Channel Consistency Assessment

The flexible multi-channel dosimetry probe integrates nine miniature surface-mounting silicon positive-intrinsic-negative (PIN)-type photodiodes (PDs) with a spectral response range of 750–1100 nm (SFH2400FA-Z, OSRAM Opto Semiconductors GmbH, Regensburg, Germany). These nine miniature PDs are epoxied at longitudinal intervals of 10 mm into pockets machined on the shell of a flexible tubular substrate with a diameter of 6.325 mm, resulting in a total longitudinal span of approximately 8 cm. Each PD is individually wired to an interfacing module, which conditions the photocurrent of each PD through a scaling resistor (22 MΩ for measurements on dogs) into a voltage signal to read the absolute irradiance.

Adjacent to each PD on the flexible probe is a cylindrical fiber-optical diffuser (FOD) with a diameter of 200 µm (Pioneer Optics, Bloomfield, CT, USA). The nine FOD channels are sequentially routed by a fiber switch to a spectrophotometric device for illumination readings. The FODs, co-localized with the PDs, provide an additional relative measure of local irradiances near the PDs to cross-check the inter channel responses.

The inter-channel responses of the 9-channel PDs were examined in two configurations under illumination. These configurations facilitated comparing the irradiances measured at the 9 PDs and cross-checking them against the illuminations picked up by the 9 FODs, with theoretical predictions applicable to the medium–probe geometry.

In configuration 1, as illustrated in Figure 1, the 9-chanel probe was positioned in a large volume (~5 L) of 1% bulk intralipid solution. This solution was illuminated with an 850 nm laser using a spherical light diffuser SD200 (Medlight) near PD #5. The inter-channel irradiance profile measured by the 9 PDs agreed well with the theoretical predictions. These predictions were based on the unstressed and convexly curved positions of the 9 PDs, modeled as if placed in an infinite medium geometry and irradiated by an isotropic source near PD #5.

In configuration 2, as shown in Figure 2, the 9-channel probe was embedded in preserved porcine lung tissue (BioQuest, Fort Atkinson, WI, USA). This tissue was pressed by two stagged pieces of 2.5 cm thick solid tissue phantom (F0399, INO, Quebec, QC, Canada), creating a 5.0 cm thick solid medium buffer between the lung tissue and the acupuncture treatment head. The treatment head, placed in contact with the solid phantom, was illuminated at 980 nm, delivered by a companion laser unit (LiteCure, Carlsbad, CA, USA). The placement of the treatment head on the solid phantom atop the lung tissue caused deformation of the flexible probe containing the 9 PDs. Consequently, the probe was positioned along a slightly convexly curved line with PD #5 closest to the treatment head.

The inter-channel irradiances measured by the 9 PDs in the tissue showed reasonable agreement with theoretical predictions. These predictions assumed a semi-infinite medium geometry, a top-hat illumination equal to the aperture of the treatment head, and the 9 sites evenly spaced at 1 cm, distributed along a line slightly curved downward with respect to the medium–air interface. Tests with the solid phantom revealed that the irradiances received by the PD #9 in the lung tissue, under a surface power ranging from 0 W to 10 W in increments of 0.5 W and powered by a 980 nm laser, showed excellent linear responses to the surface irradiance. These measurements confirmed the acceptable quality of the in-house development of the 9 PDs for reading local irradiances. Significant inter-channel inconsistency of the PDs would have complicated, if not invalidated, the inter-position comparison of the irradiances measured in the anatomically and optically complex tissue.

2.1.2. Examination of the Linear Adjustability of the Laser Power at the Four Wavelengths

The two research-laser units used in this study operated in continuous mode but differed in the wavelengths of their laser outputs. One unit operated at 808 nm and 975 nm, while the other operated at 915 nm and 1064 nm. Both lasers had their output coupled into SMA-905 terminated fiber cables. Each unit could set the laser power from 0.5 W to 10 W in increments of 0.5 W.

It is important to note that robust inter-spectral comparisons of transcutaneous transmission of the four PBMT wavelengths using the intra-spinal probe required irradiance levels that surpass the baseline irradiance of the probe when in tissue to be measured, without causing saturation of the photodiode. The significant variation in tissue transmission across different animals and multiple wavelengths necessitated an arbitrary setting of the laser power at each specific wavelength to accommodate the range of irradiance resolvable by the intra-spinal probe. Therefore, it was imperative to examine how the 9-channel probe and the four laser modules compared in terms of irradiance at the four wavelengths, according to the power settings of the individual lasers. For the irradiance measured by the probe to be comparable across the four wavelengths, and at the same surface irradiance, the probe must respond linearly to irradiance, and the power setting of the laser must accurately represent the power output. The linear response of PD to irradiance at 850 nm has been previously verified over an irradiance range spanning five orders of magnitude [19]. It is reasonable to anticipate that the linearity of the PD responses at other wavelengths will be comparable to that at 850 nm.

The inter-wavelength comparability of the irradiation detection then relies on the linearity of the laser power as indicated by the panel setting. A test bench previously used for testing the linearity of the PD irradiance response was employed to verify if the power settings of the research-unit lasers at the four wavelengths accurately represented the power output when increased incrementally by 0.5 W.

The test bench was configured as follows. The output of one research-laser unit was coupled via a 1 mm diameter fiber patch cord (FT1000EMT, 0.37NA, low OH, Thorlabs, Trenton, NJ, USA) to a beam projection module (BPM). The BPM collimated the fiber output using an achromatic lens system and homogenized the beam using a 10° holographic diffuser window (Edmund Optics, Barrington, NJ, USA). The homogenized beam was then projected into a 115 mm long, longitudinally taped aperture with a terminal aperture of 1.6 mm in diameter. The beam exiting the tunnel aperture was intercepted by a fixated silicon sensor of a commercial power meter (PM200, Thorlabs, Trenton, NJ, USA) with an active reception window of 1 cm in diameter.

The wavelength setting of the PM200 power meter was adjusted according to the vendor-specified spectral response of the silicon sensor. To read the irradiance at 808 nm, the power meter was set to 830 nm. To read the irradiance at both 915 nm and 975 nm, the power meter was set to 980 nm. To read the irradiance at 1064 nm, the power meter was set to 1064 nm.

With the test bench coupled to the laser unit at a specific wavelength, the laser power was set from 0.5 W to 10 W in increments of 0.5 W. At each power setting, the free space coupled and diffusely attenuated light emitting from the tapered aperture was measured by the power meter using the silicon sensor. Due to the attenuation of the light delivery path, including the diffuser and the tapered aperture, the power meter readings at each of the four wavelengths were in the mW range, while the laser machine parameters were set in the watt range. It was anticipated that as the laser power was scaled up according to the panel setting of the laser unit, the measured power would scale up by the same amount.

After completing the measurements at one wavelength, the light was switched to another wavelength by one of two methods as follows: (1) If the wavelength to be switched to was available from the same dual-wavelength research laser, the wavelength was changed directly on the control panel of the laser; (2) If the wavelength to be switched to was available from the other dual-wavelength research laser, the fiber patch-cable coupling to the treatment head was manually switched to the other dual-wavelength research laser, on which the wavelength of choice was set by the control panel.

The diffusely attenuated irradiance measured by the commercial silicon sensor in response to the power settings of the research-laser units across four wavelengths (808 nm, 915 nm, 985 nm, and 1064 nm) is depicted in Figure 3A. All four traces had a linear regression better than 0.999, indicating that the panel setting of each research-laser unit accurately represented the true scaling of the power emitted by the laser unit. This confirmation of the laser output scaling according to the panel settings was critical for assessing inter-subject variations of the inter-channel and inter-spectral irradiances measured in tissue.

The absolute irradiance responses of the PD at the four laser wavelengths (808 nm, 915 nm, 975 nm, 1064 nm) were estimated based on the vendor-specified photocurrent responses, as shown in Figure 3B. Compared to the peak response at 900 nm, the photo-current responsivities of the PD at 808 nm, 915 nm, 975 nm, and 1064 nm were approximately 80%, 90%, 85%, and 25%, respectively. The photo-responsivity of the PD at 980 nm with a scaling resistor of 22 MΩ was previously quantified as producing 1 V per 7.58 µW/cm² continuous-wave (CW) irradiance. Using a 22 MΩ sensing resistor for current–voltage conversion, the photo-responsivities of the PD at the four wavelengths were thus estimated as producing 1V per 12.32 µW/cm², 22.34 µW/cm², 7.58 µW/cm², and 36.3 µW/cm² continuous-wave (CW) irradiance at 808 nm, 915 nm, 975 nm, and 1064 nm, respectively.

2.2. Protocol of Intra-Spinal Dosimetry at the Four PBMT Wavelengths with Cadaver Dogs

Cadaver dogs were used in this study to assess the feasibility of using the flexible probe for intra-spinal deployment and to measure transcutaneous transmission, with the aim of translating these findings to companion care and potentially to clinical applications. The use of cadaver dogs was exempted by the Institutional Animal Care and Use Committee of Oklahoma State University. Three cadaver dogs (mixed breed) were obtained from a regional animal shelter.

Table 1. Characteristics of the dogs involved in the experiment.

Parameter	Dog 1	Dog 2	Dog 3
Breed	German Shepherd Dog (GSD)	Pit Bull	Pit Bull
Gender	Male, intact	Female, intact	Male, neutered
Weight	20.30 kg	34.55 kg	26.10 kg
Skin-Spinal Muscle Thickness	1 cm	3 cm	2.5 cm
Acquisition Details	Obtained from regional animal shelter
Handling	Delivered frozen, thawed, and used within 48 h

describes the characteristics of the dogs involved in the experiment. Dog 1 was a male, intact, German Shepherd dog (GSD), weighing 20.30 kg, with a skin-paraspinal muscle thickness of 1 cm. Dog 2 was a female, intact pit bull, weighing 34.55 kg, with a skin-paraspinal muscle thickness of 3 cm. Dog 3 was a male, neutered pit bull, weighing 26.10 kg, with a skin-paraspinal muscle thickness of 2.5 cm. The dogs were euthanized due to terminal conditions that did not affect the dermis, musculature, bones, or nervous system.

The three dogs were acquired simultaneously for use over two consecutive days. The cadavers were delivered frozen, thawed, and used within 48 h. The hair on the dorsum of each cadaver dog was clipped (Oster A5 Cordless Clipper/#40 clipper blade/Valley Vet Supply, Marysville, KS, USA) from caudal to the spinous process of the 9th thoracic vertebrae (T9) to the lumbo-sacral junction to allow for surgical access. Following a lateral hemi-laminectomy, the spinal cord was removed to expose the spinal canal extending from approximately T12 to the 6th lumbar vertebrae (L6).

The flexible 9-channel probe was inserted to the spinal canal, with the wiring (~4 m long, sorted from channel 1 to channel 9) directed caudally to the interfacing device. The most cranially placed sensor, PD #1 of the probe, was positioned at T13 by marking the position on the tissue surface using a 22-gauge hypodermic needle. The 9 PDs, embedded in epoxy pockets and in direct contact with the dorsal wall of the spinal canal lumen, were positioned to directly face the treatment beam.

After placing the probe within the spinal canal, peripheral muscle and fascia tissues were opposed and closed at each level with full-thickness simple interrupted sutures (3-0 PDS, Ethicon) to simulate tissue planes between the probe surface and spinal canal. The tissues were moistened with tap water in the area between tissue layers and within the spinal canal to help remove air.

An acupuncture treatment head detached from a companion laser unit (Model CTS, LiteCure LLC, Newark, DE, USA) [20] was positioned by an articulated arm mounted to the surgical table to form the following probe–tissue configuration: (1) placing the treatment head 1 cm above (non-contact) the tissue with the skin removed; (2) placing the treatment head 1 cm above (non-contact) the tissue with the skin intact; and (3) placing the treatment head in contact with the skin intact.

Table 2. Laser device experimental configuration.

Parameter	Laser Unit 1	Laser Unit 2
Model	LiteCure LLC, Newark, DE (Modified Model LTS)	LiteCure LLC, Newark, DE (Modified Model LTS)
Wavelengths	808 nm, 975 nm	915 nm, 1064 nm
Max Power Output	10 W	10 W
Power Increment Steps	0.5 W	0.5 W
Treatment Head Connection	Accupuncture treatment head via fiber-patch cord
Surface Irradiance Positions	T13 (site 1); 4 cm caudal to T13 (site 5); 8 cm caudal to T13 (site 9)
Measurement Configurations	Non-contact with skin removed; Non-contact with skin intact; Contact with skin intact
Measurement Procedure	1. Measure at 808 nm 2. Switch to 975 nm 3. Measure at 975 nm	1. Measure at 915 nm 2. Switch to 1064 nm 3. Measure at 1064 nm
Measured Irradiance at 0.5 W Power	Registered at 808 nm and 975 nm	Registered at 915 nm and 1064 nm
Max Irradiance (Contact, 2.0 W)	Dependent on wavelength and position
Measurement Sites (Channels)	Channel 1 (T13); Channel 5 (4 cm caudal to T13); Channel 9 (8 cm caudal to T13)
Positioning	Articulated arm mounted to the surgical table
Safety Compliance	ANSI standards, in a dedicated laser surgery laboratory approved by the Research Compliances Office of the University

presents the configuration details of the laser device. In each configuration, the beam was directed to three positions along the spinal column: T13 (site 1), 4 cm caudal to T13 (site 5), and 8 cm caudal to T13 (site 9). These three sites roughly corresponded to channel 1 (cranial), channel 5 (middle), and channel 9 (caudal) of the intra-spinal probe, which contained 9 PDs stretching over an 8 cm length of the spinal column. The configuration of the measurement on the cadaver dogs is schematized in Figure 4. Laser operations followed ANSI standards in a dedicated laser surgery laboratory approved by the Research Compliances Office of the University.

Two dual-wavelength research-laser units on loan from LiteCure LLC. were used for the study. Each laser unit produced up to 10 W of power in increments of 0.5 W at each of two wavelengths, which could be switched via the user interface panel. Research-laser unit #1 operated at 808 nm or 975 nm, while the research-laser unit #2 operated at 915 nm or 1064 nm. The acupuncture treatment head was connected via its attached fiber-patch cord to one research-laser unit.

At a fixed position of the treatment head with respect to the tissue (e.g., site 1, non-contact with skin removed), the irradiances of the 9 channels embedded in the spinal canal were recorded at 808 nm using research-laser unit #1, with the laser power increasing from 0 W in increments of 0.5 W. After completing the measurements at 808 nm with research-laser unit #1, the laser wavelength was switched to 975 nm using the control panel. After completing the measurements at 975 nm with research-laser unit #1, the fiber-patch cord of the treatment head was manually switched to research-laser unit #2, and measurements were taken at 915 nm. Following these measurements, the laser wavelength was switched to 1064 nm using the control panel.

This cycle completed the measurements of four wavelengths at one position of the treatment head (e.g., site 1, non-contact with skin removed). The treatment head was then moved to another position (e.g., site 5, non-contact with skin removed), and the cycle of measurements at the four wavelengths was repeated.

Based on preliminary testing, it was determined to apply no more than 2.0 W of laser power to the tissue to prevent tissue burning during the timeframe of operation, which included setting the power, foot-activating the laser, triggering the 9-channel data registration, foot-deactivating the laser, and saving the data. Therefore, the intra-spinal irradiance was measured at laser power settings of 0 W, 0.5 W, 1.0 W, 1.5 W, and 2.0 W.

2.3. Statistical Analysis

The objective of the study was to compare the intra-spinal irradiances among four wavelengths to determine which wavelength had the highest transcutaneous transmission when delivered under the same surface irradiance. For an otherwise identical probe–tissue configuration, the measurements differed only by wavelength. The data groups for evaluating wavelength dependence had the same data structures and sample sizes. Therefore, these data groups were subjected to one-way analysis of variance (ANOVA) using GraphPad Prism 6 (GraphPad Software, GraphPad Prism V6.0, La Jolla, CA, USA). For comparisons of more than two group means, one-way ANOVA has been suggested as the appropriate method instead of the t test [21]. The choice of one-way ANOVA over t test for comparing more than two group means is justifiable as ANOVA is based on the same assumption as the t test, and the interest of ANOVA is on the locations of the distributions represented by means, as is the t test [21]. Additionally, no post-hoc tests were conducted to identify specific differences between wavelengths.

The data groups were loaded into GraphPad Prism after a priori calculation of the mean value and the standard deviation (SD) corresponding to each PD at each treatment head–tissue configuration at each wavelength. A p value of <0.05 was considered to indicate statistically significant differences.

3. Results

The irradiances at each of the four wavelengths that reached the 9 PDs embedded in the spinal canal were measured at the baseline (no laser illumination) and at laser powers of 0.5 W, 1.0 W, 1.5 W, and 2.0 W. For each laser power setting and treatment head–tissue configuration, the irradiance measured by each of the nine channels was averaged over three dogs for measurements following the same protocol. As a result, each irradiance value was represented by a mean and a SD. However, the values were noisy due to the limited number of measurements (three) for each mean ± SD.

The irradiance measurements are presented in three sections. Section one corresponds to the non-contact application with the skin removed. Section two corresponds to non-contact application with the skin intact. Section three corresponds to the contact application with the skin intact. The results include the irradiances measured at the baseline, 0.5 W, and 2 W laser power settings. Measurements at laser powers of 1.0 W and 1.5 W were omitted to avoid overcrowding the lines.

Each marker on any line profile shown in the following three figures represents the mean value, and the error bar associated with that marker represents the SD.

3.1. Intra-Spinal Irradiance under Non-Contact Irradiation of 0.5 W and 2 W with Skin Removed

Figure 5 presents the intra-spinal irradiances measured at nine sites spaced approximately 1 cm apart under surface irradiation of 0.5 W up to 2 W by the acupuncture treatment head in non-contact mode with the skin removed. The four columns, counted from the left, correspond to wavelengths of 808 nm, 915 nm, 975 nm, and 1064 nm, respectively. Within each column, the top panel corresponds to the treatment head aligned with T13, or approximately the intra-spinal site of channel 1. The middle panel corresponds to the treatment head aligned 4 cm caudal to T13, or approximately with the intra-spinal site of channel 5. The bottom panel corresponds to the treatment head aligned 8 cm caudal to T13, or approximately with the intra-spinal site of channel 9.

Where the intra-spinal irradiance peaked at a specific probe–tissue configuration was mostly straightforwardly identifiable from the profiles, which had discrete marker overlapped with large range error bars. At a surface power of 0.5 W, corresponding to 157 mW/cm², the 975 nm light transmitted the least to the nine intra-spinal sites. The 808 nm light transmitted more than the 915 nm, the 975 nm light transmitted more than the 915 nm, and the 1064 nm light transmitted the most.

The two discrete points corresponding to the irradiances measured at channels 1 and 2 when 0.5 W of 1064 nm light was applied to site 1 were leveled and associated with an unremarkable error bar. The unremarkable error bar at that high level indicated saturation of the PD reading due to local irradiance exceeding the peak value that could be accounted for at the current–voltage scaling used. This saturation became more identifiable at 2 W of surface power application, corresponding to ~0.5 W/cm² surface irradiance.

The intra-spinal irradiance across the four wavelengths picked up by the 9 PDs at a surface power application of 0.5 W, causing less saturation to the PDs near site 1, showed a statistical difference of p = 0.0007 for the small sample size of three per site, confirming the wavelength difference in transmitted irradiance. Similar analyses of the irradiances for surface application at sites 5 and 9 resulted in a statistical power of p < 0.0001 in both cases. Clearly, the 1064 nm wavelength had the greatest transcutaneous transmission from an acupuncture treatment head positioned dorsal to the spinal column at a 1 cm distance from the tissue surface with the skin removed.

3.2. Intra-Spinal Irradiance under Non-Contact Irradiation of 0.5 W and 2 W with Skin Intact

Figure 6 presents the intra-spinal irradiances measured at 9 sites spaced approximately 1 cm apart under surface irradiation of 0.5 W up to 2 W using the acupuncture treatment head in non-contact mode with the skin intact. The organization of the figure in terms of the wavelength dependency and the site of surface application is the same as in Figure 5. The inter-channel profiles in Figure 6 lacked the pattern of having a few leveled discrete markers overlapped with small error bars, indicating no apparent saturation of the PDs by the irradiances delivered at that probe–tissue geometry for laser power up to 2 W, regardless of the wavelength.

Similar to Figure 5, at a surface power of 0.5 W (corresponding to 157 mW/cm²), the 975 nm light transmitted the least to the nine intra-spinal sites. The 808 nm light transmitted more than the 915 nm, the 975 nm light transmitted more than the 915 nm, and the 1064 nm light transmitted the most.

The intra-spinal irradiance across the four wavelengths picked up by the nine PDs at a surface power application of 0.5 W near site 1 showed a statistical difference of p = 0.0017 for the small sample size of three per site, confirming the wavelength difference in transmitted irradiance. Similar analyses of the irradiances for the surface application at site 5 and 9 resulted in a statistical difference of p < 0.0001 in both cases. Clearly, the 1064 nm wavelength had the greatest transcutaneous transmission from an acupuncture treatment head positioned dorsal to the spinal column at a 1 cm distance (non-contact) above the skin-covered tissue.

3.3. Intra-Spinal Irradiance under Contact Irradiation of 0.5 W and 2 W with Skin Intact

Figure 7 presents the intra-spinal irradiances measured at 9 sites spaced approximately 1 cm apart under surface irradiation of 0.5 W to 2 W by the acupuncture treatment head in contact mode with the skin intact. The organization of the figure in terms of the wavelength dependency and the site of surface application is the same as in Figure 5 and Figure 6. Unlike Figure 5 and like Figure 6, the inter-channel profiles presented in Figure 7 lacked the pattern of having a few leveled discrete markers overlapped with small error bars. This indicated no apparent saturation of the PDs by the irradiances delivered at that probe–tissue geometry for laser power up to 2 W, regardless of the wavelength.

Similar to Figure 5 and Figure 6, at a surface power of 0.5 W (corresponding to slightly above 157 mW/cm² due to the smaller spot size of the beam as the probe was in contact with the skin when compared to non-contact mode of a 1 cm probe–tissue distance), the 975 nm light transmitted the least to the nine intra-spinal sites. The 808 nm light transmitted more than the 915 nm, the 975 nm light transmitted more than the 915 nm, and the 1064 nm light transmitted the most. The intra-spinal irradiance across the four wavelengths picked up by the 9 PDs at a surface power application of 0.5 W near site 1 showed a statistical difference of p = 0.0013 for the small sample size of three per site, confirming the wavelength difference in transmitted irradiance. Similar analyses of the irradiances for surface application at site 5 and 9 resulted in a statistical difference of p < 0.0001 in both cases. Clearly, the 1064 nm wavelength had the greatest transcutaneous transmission from an acupuncture treatment head positioned dorsal to the spinal column in contact with the skin-covered tissue.

3.4. Peak Intra-Spinal Irradiance at the Four Wavelengths, Measured at Sites 1, 5, and 9 in Three Different Probe–Tissue Geometries

The peak irradiance registered by the 9 channels of the intra-spinal probe under surface irradiation of 0.5 W laser power (corresponding to 157 mW/cm² surface irradiance at 1 cm distance from the skin) at three sites of surface application is tabulated in

Table 3. Peak irradiance (µW/cm²) among the 9 channels of the intra-spinal probe under the surface irradiation of 0.5 W with the treatment probe in the mode of non-contact with skin removed.

Beam Site\Wavelength	808 nm	915 nm	975 nm	1064 nm
Site 1 (beam @ T13)	51.10 ± 26.04	85.20 ± 41.46	21.46 ± 17.91	193.43 ± 4.47
Site 5 (beam @ 4 cm caudal to T13)	48.99 ± 29.43	88.13 ± 54.76	21.04 ± 18.53	193.66 ± 2.00
Site 9 (beam @ 8 cm caudal to T13	39.62 ± 28.37	70.85 ± 51.65	17.84 ± 19.96	192.76 ± 2.18

for the mode of application in non-contact with skin removed. Similarly,

Table 4. Peak irradiance (µW/cm²) among the 9 channels of the intra-spinal probe under the surface irradiation of 0.5 W with the treatment probe in the mode of non-contact with skin intact.

Beam Site\Wavelength	808 nm	915 nm	975 nm	1064 nm
Site 1 (beam @ T13)	24.18 ± 36.68	44.14 ± 66.43	14.72 ± 22.59	79.63 ± 101.82
Site 5 (beam @ 4 cm caudal to T13)	24.10 ± 34.58	43.68 ± 63.83	9.50 ± 13.78	87.55 ± 93.97
Site 9 (beam @ 8 cm caudal to T13	23.55 ± 36.28	42.68 ± 65.67	7.99 ± 10.89	77.20 ± 100.59

is for the mode of application in non-contact with skin intact, and

Table 5. Peak irradiance (µW/cm²) among the 9 channels of the intra-spinal probe under the surface irradiation of 0.5 W with the treatment probe in the mode of contact with skin intact.

Beam Site\Wavelength	808 nm	915 nm	975 nm	1064 nm
Site 1 (beam @ T13)	24.61 ± 36.38	44.73 ± 65.99	14.85 ± 22.49	80.38 ± 100.80
Site 5 (beam @ 4 cm caudal to T13)	25.78 ± 34.80	48.19 ± 61.65	15.00 ± 21.90	126.68 ± 94.03
Site 9 (beam @ 8 cm caudal to T13	24.42 ± 35.42	43.99 ± 64.03	10.23 ± 14.16	87.00 ± 93.83

is for the mode of application in contact with skin intact. Within each table, the three rows descending from the top correspond to site 1, site 5, and site 9, respectively, as the locations of surface application of the laser beam.

The maximal intra-spinal irradiance at 1064 nm under a surface irradiance of ~157 mW/cm² in contact with skin-covered tissue was approximately 127 µW/cm². Comparatively, the maximal intra-spinal irradiance at 915 nm under the same surface irradiance was approximately 49 µW/cm², which is less than 39% of that at 1064 nm. The maximal transmissions at 808 nm and 975 nm were than 21% and 12%, respectively, of that at 1064 nm.

4. Discussions

This study assessed the differences in transcutaneous transmission among the four wavelengths (808 nm, 915 nm, 985 nm, and 1064 nm) commonly used in PBMT. The assessments were based on measurements obtained from only three cadaver dogs, which varied considerably in weight and skin-spinal tissue thickness. These differences contributed to a large variability in the measurements of transcutaneous transmission, despite being conducted under the same controllable conditions. The small sample size and variability among the cadaver dogs thus limit the generalizability of the findings.

Despite these limitations, it was consistently observed that transcutaneous transmission at 1064 nm was greater than at the other three wavelengths. Specifically, the maximal transmission at the spinal cord level, with a contact mode of the treatment head application under a surface irradiance of 0.5 W power (delivering approximately 157 mW/cm²), reached 127 µW/cm². If the surface irradiance was scaled up by 20 times to 10 W, as in a previous examination at 980 nm on six cadaver dogs [20], the maximal intra-spinal irradiance could be scaled up similarly by 20 times to reach 2.54 mW/cm². However, this level may still be less than the terminal or target irradiance of 3.2 mW/cm², at which the effectiveness of PBM at 670 nm for rodent models of SCI was demonstrated [22].

Further increasing the surface entrance power or irradiance could potentially bring the target irradiance to the known therapeutic level of 3.2 mW/cm². However, this approach might not be practical, as the required duration of surface irradiation could cause significant tissue heating and pose a risk of collateral thermal injury.

A more practical approach to achieving greater penetration of the same total surface application of laser power to the spinal cord level could involve optimizing the beam profile. Shaping the beam of surface application to spatially distribute the laser power, rather than concentrating it in a localized spot, could create a more favorable starting phase for light propagation in tissue, thus transmitting more energy to deeper tissue layers. This concept has been well established through computational evaluations of wavelength and beam width on light–tissue penetration [23], including skin penetration [24].

The comparison of transcutaneous light transmission at the four wavelengths to the spinal cord level was conducted with irradiance measured at only 0.5 W, the initial operable output setting of the two research-laser units. Although measurements at the same site, the same mode of laser application, and the same animal were repeated at power settings of 1.0 W, 1.5 W, and 2.0 W, cross-wavelength comparisons at these higher power settings were deemed potentially misleading. This was due to the observation of non-scaled transmission to the spinal canal as the power increased from 0.5 W to 2.0 W, despite the tissue surface remaining visually intact and free of burns during laser activation.

Under ideally consistent conditions, if tissue optical properties remained unchanged, doubling the power applied to the surface would result in a doubling of the irradiance measured at the target level. This response should hold for any scaling of the surface power. Figure 8 displays the distribution of scaling of the irradiances measured at power settings greater than 0.5 W with respect to 0.5 W of all sites and modes of laser application at all four wavelengths. Ideally, a scalable response of irradiance to surface power application would result in all the data points aligning along the diagonal line, indicating that the irradiance increased X times if the surface irradiance increased X times. However, the observed patterns showed that while the target irradiance did increase as the surface irradiance scaled up, it did not increase proportionally. The increase in the target level irradiance was noticeably less than the proportion corresponding to the scaling of the laser power, and the under-proportionate increase became more severe for greater scaling of the laser power compared to that of 0.5 W. The combined distribution of data points suggests a saturation of the target irradiance as the surface power increased to approximately 2 W.

These findings indicate that while increasing surface power does enhance target irradiance, the relationship is not perfectly scalable. This saturation effect could be due to several factors, including the inherent optical properties of the tissue, and potential photo-thermal effects that alter tissue properties [25] during high-power applications.

We observed that at any given site and mode of laser application, there was an extended time required for sample preparation. Consequently, the application of the 0.5 W laser power was performed on tissue maintained at baseline temperature for an extended period. Due to the time demands of the challenging protocol, measurements at 1.0 W of surface power were conducted immediately after the completion of the 0.5 W measurement. The laser was foot-activated OFF once the computer interface registered readings from eight of the nine channels of the probe. Counting the manual recording of the reading on channel 9 from the digital multimeter and the subsequent data-logging operation on the computer, there was a gap of less than 30 s between the application of a laser power at the lower level (0.5 W) and the next incremental level (1.0 W).

This brief gap might not have been sufficient for the surface tissue to cool down to baseline conditions. Sustained thermal responses could have altered the tissue’s surface scattering or absorption properties [26,27] compared to baseline, significantly affecting the diffusive transmission of light in the skin and, consequently, the amount of light reaching the target level. It can be expected that the extent of optical property changes would increase as the laser power increases, due to the greater energy deposited as heat [12,24]. This is consistent with our observations of tissue burning at a higher laser power, which led to the decision to limit the laser power to 2 W for measurements. These findings suggest that thermal effects play a significant role in altering the tissue optical properties during laser application, thereby affecting the accuracy of transcutaneous transmission measurements.

The observed greater penetration at 1064 nm in this study corroborates earlier reports indicating higher soft tissue transmission at this wavelength compared to others in the visible and NIR band used in PBMT [28]. Measurements conducted through human skin tissues (e.g., ear lobes and between the fingers) in vivo at six wavelengths (532 nm, 632 nm, 675 nm, 810 nm, 911 nm, and 1064 nm) have shown the following: (1) the photons at 1064 nm penetrate deeper than those at other wavelengths studied for a given incident beam diameter; and (2) the transmittance at a particular wavelength increases asymptotically with incident beam diameter.

Since PBMT effects are dose-dependent, it is crucial to select the applied dose that minimizes collateral thermal damage while ensuring that it is adequate to produce the desired bio-modulatory therapeutic response. Clinical transcutaneous administration must consider both safe irradiance levels and manageable exposure times. For continuous-wave laser applications, the dose (or fluence) is the product of the irradiance and application time. For pulsed laser application, the dose is the product of irradiance per pulse, duty cycle, and pulse repetition rate [29,30]. Therapeutic responses and thermal effects can differ significantly for pulsed lasers, even with the same total dose, due to varying pulse parameters [31].

In both continuous and pulsed modes, the same treatment dose (energy/area) can be delivered with either higher target-level irradiance over a shorter exposure time or lower target-level irradiance over a longer exposure time. The physiological changes induced by PBMT in cranial nervous tissue have motivated its use for spinal cord nervous tissue, as both target tissues are similar in neuronal composition. However, the anatomic differences between the cranium and the spinal cord are significant. Thus, the clinical challenge is determining which wavelength allows the less attenuation of light transmission through the more substantial tissue and bone cover to deliver a therapeutic target dose to the spinal cord.

Understanding the wavelength dependency of dose delivery as PBMT light transmits through skin, underlying tissue, and bone to reach the target site is essential. Our limited study provides useful information for a more rigorous examination of machine parameters. Specifically, it suggests that it may be necessary to investigate if a medical 1064 nm laser, controlled at a low power, is applicable for PBMT of SCI. The investigation must carefully assess the collateral thermal effects on surface tissue in delivering therapeutic target dose to the spinal cord level. The clinical potential thus may be dictated by the control of the collateral thermal concern, over the readiness of delivering the therapeutic target dose at the spinal cord level. Delivering a therapeutic dose to the spinal cord level for effective PBMT while controlling the surface temperature required consideration of not only the wavelength that provides the strongest transmission, but also the optimization of surface beam profiling. This optimization should aim to maximize transmission to the spina cord level for the same amount of total laser energy delivered to the tissue surface.

5. Conclusions

This study aimed to assess whether one of four wavelengths commonly used for PBMT could transmit a greater amount of light to the spinal cord under the same surface irradiance. The target level irradiances of PBMT light at 808 nm, 915 nm, 975 nm, and 1064 nm were measured along the thoracic to lumbar segments of the interior spinal canal in three cadaver dogs, all subjected to the same surface irradiance values. Through lateral hemi-laminectomy, a flexible intra-spinal probe was embedded in the spinal canal to measure intra-spinal irradiances spanning approximately 8 cm at 9 sites of photodiodes sites, under CW surface application of the PBMT light at powers ranging from 0.5 W up to 2 W.

Surface illumination was applied using an acupuncture treatment head in three modes: non-contact with skin-removed, non-contact with skin intact and contact with skin intact. In each mode, the treatment head was positioned near T13 aligned at the cranial aspect of the probe, and at approximately 4 cm and 8 cm caudal to T13, roughly at the causal end of the probe. At each position, the laser was multiplexed among the four wavelengths at the same power setting, starting at 0.5 W. The intra-spinal irradiances at the four wavelengths were compared after compensating for the spectral sensitivities of the photodiodes.

In all three modes of surface application, the target level irradiance at the 1064 nm was significantly greater than at the other three wavelengths (p ≤ 0.0017). At a surface irradiance of ~157 mW/cm² corresponding to 0.5 W laser power applied with the treatment head in contact with the skin intact, the intra-spinal irradiance at 1064 nm reached 0.137 ± 0.095 mW/cm². These findings suggest that 1064 nm may be a more effective wavelength for transcutaneous PBMT for spinal cord injury. Additionally, this study provides a foundation for further investigations into the effects of machine parameters, such as continuous wave versus pulsed modes, on treatment efficacy, at 1064 nm as a candidate wavelength.