Photobiomodulation involves changes in cellular metabolism through the transfer of energy from photons to its photoacceptors 
. The pivotal role of mitochondria in the translation of biophotonic energy to biochemical changes has been previously demonstrated 
. However, the rigour of the photobiomodulation clinical outcomes has been questioned. The device technology used for photobiomodulation delivery plays a critical role in improving treatment standardization 
. Indeed, our data point out that power density delivery is affected by the probes used for these treatments. In this study, we noted that despite both the fibre and standard hand-piece delivering the correct power density, only a surface area less than 50% of the laser spot size was effectively illuminated. Further, the distribution of the different power densities on the treatment spot size was more non-uniform for the fibre compared to a standard hand-piece. Strikingly, flat-top prototype delivered the most consistent power distribution over 90% of the treatment area. Furthermore, the flat-top hand-piece was able to maintain constant power from contact to a couple of centimetres away allowing for improved clinical consistency during PBM treatment delivery as previously reported by Hanna et al. 
. This evidence supports the notion that the effectiveness of photobiomodulation therapy could be significantly influenced by both the probes employed and the delivery technique of the operator. The significant improvement in mitochondrial activity with the flat-top hand-piece prototype demonstrated a homogenous treatment spot size in the centre and on either sides. This contrasted with the significant differences in the fibre and standard hand-piece groups indicating that the biological responses were most uniformly achieved with the current flat-top prototype. Improvements in the laser device, fibre and collimating procedure with the lens and probe design could further significantly improve the rigour and consistency of PBM clinical treatments.
As noted in our previous work, PBM responses involve complex events resulting from the absorption and scattering of the photons and the generation of an electromagnetic field 
. This work was based on three-dimensional modelling of photon waves interacting within the mitochondrial droplet generated between the laser to the coverslip. The differences in the beam power density distribution during treatment with the three probes may impact these interactions affecting the overall photobiomodulation responses. Additionally, recent work from our group observed a limited dose range of a 980 nm diode laser, which affects the complexes III and IV as well as ATP production and oxygen consumption of mitochondria 
. Slight variations (0.1 W) in the treatment power appeared to drastically modulate the photobiomodulation outcomes. Basically, 0.8–1.1 W kept mitochondria coupled and induced increments of ATP production by increments of complex III and IV activities. In contrast, 0.1–0.2 W uncoupled the mitochondria and had an inhibitory effect of ATP synthesis and increment of oxidative stress, while 0.3–0.7 and 1.2–1.4 W did not appear to affect these responses. These data suggest a major limitation in the reproducibility of photobiomodulation responses lie in a non-homogeneous distribution of the laser energy. Variations in an area of only 1 cm2
appeared to induce drastically different photobiomodulatory responses (positive, null or negative) in these in vitro studies. Therefore, we would expect significant differences in groups of neighbouring cells and overall tissues when this therapy is used in clinical in vivo scenarios.
This work has some strengths and limitations. A single spot size of 1.13 cm in diameter was used allowing for a limited drop volume of isolated mitochondria sufficient to assess ATP. Further reduction of the drop diameter or an increment of the laser spot-size as well as reducing treatment power could highlight further, perhaps more dramatic, differences. However, the use of three-dimensional dose modelling 
, the temperature monitoring during treatments 
, the standardised ATP synthesis evaluation of treated mitochondria 
and the careful assessment of the treatment power at the target surface with a power meter 
, enabled comparisons of the three discrete probe designs and were clear strengths.
To summarize, our data demonstrated that the fibre, standard hand-piece and the flat-top hand-piece prototype have different beam energy distribution features. These differences significantly affected our primary endpoint, which was mitochondria activity with respect to their position in the treatment spot size. Our second endpoint showed that even at a power of 1 W, there was no damage to mitochondrial function. These results provide evidence that flat-top hand-piece allows improved photobiomodulation treatment reproducibility, especially in clinical scenarios where the distance from the target surface may vary during treatments and the wide affected area needs uniform irradiation to better experience the photobiomodulation effects.