Optimization of human semen analysis using CASA-Mot technology

Abstract

The purpose of this study is to investigate the optimal framerate (FR) and the use of different counting chambers for improving CASA-Mot technology use in Andrology. Images were captured at 500 fps, then segmented and analyzed in several ranges of FRs (from 25 to 250) to define the asymptotic point that as an optimal FR. This work was replicated using counting chambers based in capillarity (disposable) or drop displacement (reusable) to study their effects on the motility results and kinematic values of the samples under the different experimental conditions. The α value (asymptote corresponding to FR_o) of the exponential curve was 150.23 fps, corresponding to a VCL of 130.58 mm/s, far from the value of 98.89 mm/s corresponding to 50 fps (the highest FR used by most current CASA-Mot systems). Our results have shown that, when using reusable counting chambers, type and depth have influence. In addition, different results were obtained depending on the area of image captured inside the different counting chamber types. To have reliable results in human sperm kinematic studies, almost 150 fps should be used for capturing and analyzing and differences between chambers should be considered by sampling from different areas, to obtain a representative value of the whole sample.

Keywords:

• CASA-Mot technology
• frame rate
• counting chamber
• semen analysis

Introduction

Infertility is defined like as the inability of a sexually active, non-contracepting couple to achieve pregnancy in one year. The male partner can be evaluated for infertility or subfertility using a variety of clinical interventions, and also from a laboratory evaluation of semen (WHO Semen manual, 5th Edition³). It is estimated that about 25% of couples show infertility in developed countries, where more information is available (Mascarenhas et al. 2012). Although male reproductive function is impaired in about half of these infertile couples (Kumar and Singh 2015), it depends on the geographical area (Agarwal et al. 2015). Semen analysis is the best way to evaluate male fertility ability being centered on sperm concentration, motility, morphology and viability (Tomlinson et al. 1999; WHO 2010).

Two approaches to perform the seminal analysis ways are coexisting around the world, the subjective and the computer assisted semen analysis (CASA). The subjective method is being used from the middle of past century (MacLeod and Gold 1951), and was systematized in the successive WHO (WHO 2021), and other manuals (Björndahl et al. 2010). The main issue with this approach is the need for high training and the fact that subjectivity introduces a bias in the final results with a high CV amongst the repetitions in the same sample analysis (Cooper et al. 2002). Nevertheless, the number of nonconformities to international guidelines is close to 90%, indicating that nonstandard criteria are used when semen analyses are performed (Bailey et al. 2007; Palacios et al. 2012). This fact has claimed for the necessity to introduce accurate quality control programs to be sure that almost the technician is, at least, working inside some reasonable margins (Cooper et al. 2002; Mallidis et al. 2012; Filimberti et al. 2013; Sánchez-Pozo et al. 2013).

The vagaries of the eye-brain system in assessing semen quality can be eliminated by objective assessment where sperm characteristics are defined by a battery of quantitative parameters (Soler et al. 2016). This principle was the basis for the development in the 1980’s of Computer Assisted Semen analysis (CASA) technology (Bompart et al. 2018). These computer-aided sperm analysis (CASA-Mot) systems were developed to provide more reliable, repeatable, and objective measurements of sperm movement (Rurangwa et al. 2004), and to reduce the intertechnical variation in the estimation of sperm concentration, regardless of the type and depth of the counting chamber used. (Lu et al. 2007). In addition, CASA-Mot systems are used for obtaining high quantities of spermatozoon data, such as kinematic parameters. The sperm quality of each male could be assessed and bundles of spermatozoa with similar motility features could be defined, by applying subpopulation analysis to the mentioned data (Soler et al. 2014). Nevertheless, like any other technology CASA depends on the capabilities of both the hardware and the software used (Talarczyk-Desole et al. 2017; Yeste et al. 2018). Likewise, it is important to mention other limitations of the CASA systems, such as the absence of defined standard values for variables in the different species and lack of reliable standardization of settings from different equipment (Gill et al. 1988; Vantman et al. 1988; Hoogewijs et al. 2012). Despite these limitations, throughout the last two decades CASA systems have been implemented in a lot of human clinical laboratories showing a good predictive value of the fertility of the samples even considering all of their technological limitations (Holt et al. 2018; Tomlinson and Naeem 2018).

Particularly, the effect of frame rate and counting chamber use has been studied in a variety of species, showing that the sperm behavior is species-specific and the need for accurate evaluation in each species and condition (Bompart et al. 2019; Valverde 2019; Caldeira et al. 2019).

In addition, the effect of the counting chamber used and how it is manipulated has an important effect on what was observed a long time ago (Coetzee and Menkveld 2001) and recently analyzed in detail (Bompart et al. 2018).

The aim of the present work is to use the most advanced CASA technology to define the optimized protocol conditions for human semen analysis.

Results and discussion

Definition of optimal frame rate for kinematic analysis

As explained in more detail in the materials and methods section, the optimal frame rate was determined from the VCL of the sperm using point-to-point reconstructions of their trajectories at each tested FR. VCL was used to define the FR_o (α value of the exponential curve indication asymptotic value) because it is the most sensitive parameter to FR changes. When using disposables chamber, FR_o was quite similar despite its depth (132.2 fps for 10 µm depth and 133 fps for 20 µm depth). Nevertheless, when reusable chambers were used, they did show differences (126.9 and 150.2 fps for 10 and 20 µm depth, respectively) (Table 1, Figure 1).

Figure 1. Sperm curvilinear velocity (VCL. µm/s) obtained at different frame rates in the different counting chambers. Optimum frame rate (FR_o) was calculated for the different types of chamber: disposable chambers 10 and 20 µm depth (ISAS^®D4C10, ISAS^®D4C20) and reusable chambers 10 and 20 µm depth (Spermtrack^® 10, Spermtrack^® 20). VCL: curvilinear velocity (μm/s).

Table 1. Optimum frame rate calculated to have the threshold level for different chambers types and depths.

Chamber	N	α	SEMα	β	SEMβ	VCLα	VCL25	VCL50	VCL100	VCL150	VCL200	VCL250
D4C10	5762	132.17	1.20	18.04	0.67	115.3	64.2	92.1	111.4	117.2	120.8	123.0
D4C20	10861	132.97	0.86	19.49	0.49	114.8	61.0	90.1	109.4	116.8	120.6	123.0
SpK. 10	13906	150.23	0.81	20.91	0.40	130.7	65.1	98.9	121.9	130.7	135.3	138.2
SpK. 20	20494	126.88	0.72	15.60	0.39	112.2	68.0	92.9	108.6	114.4	117.4	119.2

N: number of spermatozoa analyzed; α: asymptote of curvilinear velocity; β: rate of increase; SEM: standard error of the mean; VCL: curvilinear velocity (µm/s). D4C10, D4C20: disposable chambers 10 and 20 µm depth. SpK. 10, SpK. 20: reusable chambers (Spermtrack^®) 10 and 20 µm depth.

Curvilineal velocity of human sperm for the optimal frame rate (VCLα) and with different frames rates (25–250) calculated based on α and β values.

These data indicate results are influenced by the counting chamber type and depth, the latter only in the case where a reusable chamber is used.

Effect of counting chamber type and area of analysis

Besides the differences in the FR_o observed in different counting chamber types, the highest sperm kinematic values were observed in the reusable Spermtack^® chamber and even higher with the 10 µm than with 20 µm depth. In the case of the disposable chambers, the depth influence observed was not significant. The movement in reusable chambers was more linear and faster, thus showing higher head lateral displacement (ALH) (Table 2).

Table 2. Motility parameters (mean ± SEM) of human spermatozoa determined by CASA system using different counting chambers, and using a frame rate capture of 250 fps.

	ISASD4C10	ISASD4C20	Spermtrack10	Spermtrack20
n	995	1895	2054	3474
VCL	116.1 ± 1.8^b	115.4 ± 1.3^b	132.9 ± 1.3^a	114.1 ± 1.0^b
VSL	37.8 ± 0.8^c	33.9 ± 0.6^c	50.3 ± 0.6^a	38.6 ± 0.5^b
VAP	84.0 ± 1.2^b	80.2 ± 0.9^b	92.2 ± 0.8^a	76.9 ± 0.6^c
LIN	28.0 ± 0.5^c	29.7 ± 0.4^c	38.3 ± 0.4^a	33.5 ± 0.3^b
STR	39.21 ± 0.6^d	42.5 ± 0.5^c	54.3 ± 0.5^a	49.1 ± 0.4^b
WOB	72.7 ± 0.3^a	70.42 ± 0.24^b	70.24 ± 0.23^b	67.90 ± 0.18^c
ALH	1.18 ± 0.01^b	1.21 ± 0.01^b	1.33 ± 0.01^a	1.31 ± 0.01^a
BCF	23.1 ± 0.5^b	22.9 ± 0.3^b	27.7 ± 0.3^a	22.7 ± 0.2^b

SEM: standard error of the mean; n: number of spermatozoa analyzed; VCL: curvilinear velocity (µm/s); VSL: straight line velocity (µm/s); VAP: average path velocity (µm/s); LIN: linearity (%); STR: straightness (%); WOB: wobble (%); ALH: amplitude of lateral head displacement (µm); BCF: beat-cross frequency (Hz). ISASD4C10, ISASD4C20: disposable chambers 10 and 20 µm depth; Spermtrack10, Spermtrack 20: reusable chambers 10 and 20 µm depth. ^a,b,c,dWithin columns, rates with different superscripts (p < 0.05).

Reusable chambers tended to decrease sperm kinematic values as we shift from the center to the periphery of the counting area but was only significant for some of the parameters in the most exterior region. This phenomenon was even more sensitive for 20 µm than for 10 µm depth (Table 3).

Table 3. CASA motility parameters (mean ± SEM) of human spermatozoa in different areas of the reusable counting chambers Spermtrack^®10 and 20 µm depth.

	VCL	VSL	VAP	LIN	STR	WOB	ALH	BCF
SpK.10
1	132.6 ± 3.0	52.4 ± 1.6^a	91.1 ± 2.0	40.5 ± 1.0^a	57.5 ± 1.3^a	69.8 ± 0.6	1.33 ± 0.02	28.5 ± 0.7
2	132.2 ± 1.7	51.9 ± 0.9^a	91.8 ± 1.1	39.5 ± 0.6^a	55.9 ± 0.7^a	70.3 ± 0.3	1.32 ± 0.01	28.0 ± 0.4
3	134.0 ± 1.9	47.3 ± 1.0^b	93.1 ± 1.3	35.8 ± 0.7^b	50.9 ± 0.8^b	70.3 ± 0.4	1.33 ± 0.01	27.0 ± 0.4
SpK.20
1	118.7 ± 2.8^a	40.6 ± 1.3^a	79.3 ± 1.9^a	34.6 ± 0.8^a	51.0 ± 1.0^a	67.3 ± 0.5	1.37 ± 0.02^a	24.1 ± 0.6^a
2	115.9 ± 1.5^a	40.0 ± 0.7^a	77.9 ± 1.0^a	34.0 ± 0.4^a	49.7 ± 0.5^a	67.7 ± 0.3	1.30 ± 0.01^b	22.9 ± 0.3^a
3	109.0 ± 1.7^b	36.0 ± 0.8^b	74.0 ± 1.2^b	32.4 ± 0.5^b	47.5 ± 0.6^b	68.0 ± 0.3	1.29 ± 0.01^b	21.6 ± 0.4^b

SEM: standard error of the mean; VCL (µm/s) curvilinear velocity, VSL (µm/s) straight line velocity, VAP (µm/s) average path velocity, LIN (%) linearity, STR (%) straightness, WOB (%) wobble, ALH (µm) amplitude of lateral head displacement, BCF (Hz) beat-cross frequency. Reusable chambers SpK. 10 µm and SpK 20 µm. Different letters of the alphabet within columns indicate significant differences between different areas (p < 0.05). Capture areas for semen analysis proximal (1), medium (2), distal (3) as described in Figure 3.

When using 20 µm depth disposable chambers, LIN and BCF were reduced in areas far from the charging point, while VCL and VAP remained invariable along the counting points (Table 4).

Table 4. CASA motility parameters (mean ± SEM) of human spermatozoa in different areas of disposable counting chambers ISAS^®D4C 10 and 20 µm depth.

	VCL	VSL	VAP	LIN	STR	WOB	ALH	BCF
D4C10
1	117.2 ± 4.2	35.2 ± 1.6^a	82.7 ± 3.0	29.9 ± 1.1^a	44.1 ± 1.6^a	70.0 ± 0.9^c	1.18 ± 0.03	22.8 ± 0.9
2	113.8 ± 4.4	32.7 ± 1.7^a.b	81.8 ± 3.2	28.4 ± 1.2^a.b	39.9 ± 1.7^a.b.c	72.5 ± 0.9^b.c	1.16 ± 0.03	21.7 ± 1.0
3	114.3 ± 4.4	33.8 ± 1.7^a.b	81.2 ± 3.2	29.9 ± 1.2^a	41.4 ± 1.7^a.b	71.9 ± 0.9^b.c	1.17 ± 0.03	22.9 ± 1.0
4	114.8 ± 4.4	32.1 ± 1.7^a.b	83.7 ± 3.2	28.7 ± 1.2^a.b	39.6 ± 1.7^b.c	73.5 ± 0.9^a.b	1.18 ± 0.03	23.5 ± 1.0
5	119.1 ± 4.2	29.2 ± 1.6^b	86.4 ± 3.0	25.8 ± 1.1^b	36.0 ± 1.6^c.d	72.6 ± 0.9^b	1.21 ± 0.03	23.8 ± 0.9
6	116.1 ± 4.6	29.1 ± 1.8^b	87.3 ± 3.4	25.9 ± 1.3^b	34.7 ± 1.8^d	75.5 ± 1.0^a	1.19 ± 0.03	23.8 ± 1.0
7	117.0 ± 5.3	29.4 ± 2.1^b	85.9 ± 3.9	26.7 ± 1.5^a.b	37.5 ± 2.0^b.c.d	73.5 ± 1.1^a.b	1.20 ± 0.03	23.7 ± 1.2
D4C20
1	119.3 ± 3.3	33.5 ± 1.4^a.b.c	83.1 ± 2.2	27.9 ± 0.9^c.d	40.5 ± 1.3^c.d	70.3 ± 0.7	1.24 ± 0.02	23.8 ± 0.8^a
2	116.6 ± 3.2	33.9 ± 1.3^a.b	81.5 ± 2.0	29.2 ± 0.9^b.c.d	41.2 ± 1.2^b.c.d	70.7 ± 0.7	1.22 ± 0.02	23.2 ± 0.7^a.b
3	117.9 ± 3.3	36.7 ± 1.4^a	81.3 ± 2.2	31.0 ± 0.9^a.b	45.1 ± 1.2^a	69.7 ± 0.7	1.23 ± 0.02	23.7 ± 0.7^a
4	114.5 ± 3.0	35.0 ± 1.3^a.b	79.0 ± 2.0	31.5 ± 0.8^a	45.7 ± 1.1^a	69.6 ± 0.6	1.22 ± 0.02	23.0 ± 0.7^a.b
5	115.0 ± 3.0	35.0 ± 1.2^a.b	79.5 ± 2.0	30.7 ± 0.8^a.b	44.3 ± 1.1^a.b	70.5 ± 0.6	1.20 ± 0.02	22.4 ± 0.7^a.b
6	113.0 ± 3.2	32.8 ± 1.3^b.c	79.7 ± 2.1	29.5 ± 0.9^a.b.c	41.6 ± 1.2^b.c	71.1 ± 0.7	1.20 ± 0.02	22.6 ± 0.7^a.b
7	111.6 ± 3.5	29.7 ± 1.4^c	77.8 ± 2.3	26.9 ± 1.0^d	37.9 ± 1.3^d	71.2 ± 0.7	1.19 ± 0.02	21.3 ± 0.8^b

SEM: standard error of the mean; VCL (µm/s) curvilinear velocity, VSL (µm/s) straight line velocity, VAP (µm/s) average path velocity, LIN (%) linearity, STR (%) straightness, WOB (%) wobble, ALH (µm) amplitude of lateral head displacement, BCF (Hz) beat-cross frequency. Disposable chambers D4C 10 and 20 µm. Different letters of the alphabet within columns indicate significant differences between different areas (p < 0.05). Capture areas for semen analysis (1–7) Figure 2.

Arguing the use of CASA systems based on our results

CASA technology is commercially available from since the 1980’s. However, its presence in human clinical practice is much lower than expected (Tomlinson et al. 2010), despite the fact that the relevance of analytical data obtained in different research fields involving spermatozoa has been demonstrated (Yániz et al. 2015, 2018).

One reason for this paradox could be the following: CASA systems have just been used to replace human observations, but follow the same subjective criteria (Yeung et al. 1997). This premise can be validated by looking at the different WHO manuals; in the fourth edition of the WHO manual (1999), class “a” of sperm motility, is defined as fast and progressive motility (i.e., ≥25µm/s at 37 °C and ≥20µm/s at 20 °C). But quantifying what "fast" means in these magnitudes is not possible for the human eye and, as a consequence, to be fast just means not to be slow, which is something completely subjective, even if a kinematic limit is proposed; for example, why 25 and not 26? This problem of having four motility classes was skipped in the fifth edition of the WHO manual (2010) by reducing those classes to three. Currently, we have the latest edition of the WHO manual (2021), where the four-speed classes classification have returned.

Another reason is related with the technological limitations of the systems (Bompart et al. 2018). It was shown long ago that the FR interferes with the quality of the sperm kinematic results (Mortimer et al. 1988; Castellini et al. 2011). The most commonly used video cameras could operate from 25 to 60 fps (Robinson et al. 2018); nowadays, some commercial systems are able to capture at 100 fps (García-Molina et al. 2021). But working with these FRs entails problems because the identification of spermatozoa and their trajectories is not as satisfactory as is required and therefore a manual correction of the results must be made, which then interferes with the final results (Robinson et al. 2018).

In the present study, we attempted to standardize the technique used for the motility assessment of human sperm using high-resolution cameras and specific chambers. The first step was to define the FR_o for sperm kinematic analysis. In this study, we found a direct positive relationship between capture frequency and VCL. The tracking of the sperm from one dot to the following is allowed by a proper capture frequency, which permit a correct reconstruction of sperm trajectory; thereafter, VCL is calculated from its trajectory (speed = trajectory/time). The maximum frame required is obtained by non-linear regressions, defined as FR_o, which represents the threshold (α) from which more captures will not imply a better definition of sperm path.

In recent years, similar studies have been published in different species to find the FR_o and to achieve a satisfactory sperm kinematic analysis. This FR_o was 300 fps in stallions (Gacem, Bompart, et al. 2020), 250 fps in bull (Bompart et al. 2019), donkies (Gacem, Catalán, et al. 2020) and salmon (Caldeira et al. 2019), 225 fps in boar (Valverde et al. 2019) and sturgeon (Caldeira et al. 2019) and 200 fps in eels (Caldeira et al. 2019). In our study, the value obtained for human semen is 150fps, quite lower than in other species, which indicates how slowly and regularly human spermatozoa moves, compared with other species.

The obtained value for the FR_o of 150 fps indicates that FRs below this value are offering suboptimal kinematic results. Obviously, the lower the FR, the less clinical and scientific value can be sustained from CASA-Mot analysis, and this is one of the reasons limiting its use and significance (Knuth et al. 1987; Knuth and Nieschlag 1988; Bompart et al. 2018). But it does not mean that all the former work devoted to human sperm kinematics has no value; throughout the history of Science, the improvements on measurement capabilities has overcome meaning of the old results but never meant to invalidate them.

The second step was to analyze the sperm behavior in different chamber types. Everyone knows that use of CASA systems is influenced by a variety of technical factors, such as the use of different counting chambers (Bailey et al. 2007; Soler et al. 2012; Bompart et al. 2018; Valverde et al. 2019). One traditional (but inconvenient) alternative is to use just a simple microscope slide with a coverslip but, in this case, it is not possible to standardize the depth of the preparation. Moreover, as the coverslip floats on the sample, it induces additional high drifting sperm movement (Del Gallego et al. 2017; Robinson et al. 2018). Two alternative counting chamber types have been developed by different manufacturers: drop displacement reusable (Makler^®, Spermtrack^® …) and capillary loaded disposable (Leja^®, ISAS^®D …). These chambers are based on different physics diffusion principles, so it is not surprising that the obtained results are also different (Del Gallego et al. 2017; Bompart et al. 2018; Valverde et al. 2019). In the present work, the highest velocities and the most linear motility were observed when Spermtrack^® (drop displacement type) of 10 µm depth was used. Therefore, the results of our study demonstrate that the depth of the counting chamber influence the kinematic values. In fact, when comparing between reusable chambers but at different depths, we can observe that the sperm presents higher values for all the kinematic parameters (VCL, VSL, VAP, LIN, STR, WOB, ALH and BCF) in the narrower 10 µm chamber than in the 20 µm chamber.

There are different approaches to evaluate the effectiveness of counting chambers such as the use of limits of agreement (Bailey et al. 2007), but the purpose of the present work was not to try to define the “real values” for evaluation of kinematics (which is not possible under our point of view) but just to point out how important is to consider the effect of counting chambers type on the final results and how it is not advisable to compare results among different works without considering it this.

Finally, in addition to the type and depth of the chambers, it is necessary to consider the location where sampling is produced inside the counting chamber. Most of the studies done throughout time have never indicated this fact and counts were commonly done in the center of the counting area after putting a volume inside that was not fitted to the actual space (Bompart et al. 2018). In a previous study, the same type of reusable counting chambers (Spermtrack^®) were tested but at 25 fps (with much lower sensitivity than the one used here, 150fps) and no differences were observed regarding the counting area considered (Soler et al. 2012). In this study, the movement was less linear with a significantly lower VSL only in the periphery of the drop displacement chambers, independent of the depth. When capillary loaded disposable chambers were used, the effect was just the opposite, the further away from the sample deposit point, the lower the VSL and the LIN of the movement, being significant from the fifth field. This fact was previously observed in other species such as foxes (Soler et al. 2014) or bulls (using Leja and CellVu chambers in these cases, Valverde et al. 2019). This may be due to the interaction of the spermatozoa during their passive progressing on the glass surface or due to the surfactants that are added to the glass in these types of chambers. But another not-tested-hypothesis is based on the possibility that a swim-up process could occur inside the tip of the micropipette: the sperm at the bottom of the tip will be pushed to the most advanced areas of the counting chamber, while the cells on the top of the tip (which are the ones with high movement) will remain close to the place of sample deposition.

Materials and methods

Patients

Thirteen semen samples were obtained from adult volunteers (mean age 25.85 ± 8.75, range 19–53). All the volunteers signed informed consent forms to participate in and provide semen samples for the study.

The samples were obtained by masturbation following sexual abstinence for 3-5 days, collected in a clean 60 mL sterile standard container and stored in an incubator at 37 °C for 30 min to allow liquefaction (García-Molina et al. 2021).

Frame rate effect

All semen samples were recorded at 500 fps frame rate (FR) for 1 s. This video was segmented into 25, 50, 100, 150, 200 and 250 FR videos. The command used for video segmentation was: [echo off: set fps= 25, 50, 75, 100, 150: for %%i in (.Ä*.avi) do (set fname=%%∼ni) & call: encodeVideo; goto eof: encodeVideo: ffmpeg.exe -i %fname%.avi -r %fps% -clibx264 -preset slow -qp 0 "%fname%_(%fps%fps).avi"; goto eof].

Motility evaluation by CASA-Mot

The assessment of sperm motility and kinematics was carried out with the ISAS^®v1 CASA-Mot system (Proiser R + D, Paterna, Spain), connected to an ISAS^®-UOP200i negative phase contrast microscope equipped with a heating stage at 37 °C and using a 10× objective (0.25 NA) (Proiser R + D). A high-speed camera (500 fps; M03-CM, Proiser R + D) was used for image capture. The array size of the video frame grabber was 648 × 488 × 8 bits and 256 grey levels. Image resolution was 0.70 µm per pixel in both the horizontal and vertical axes. The tail detection facility of the system was activated to ignore non-sperm particles, with a particle area between 5 and 80 µm² and a connectivity value of 10 µm. Track recognition mistakes were deleted, when needed, to avoid the introduction of distortions in the final results.

The sperm kinematic parameters analyzed were curvilinear velocity (VCL, µm/s), straight-line velocity (VSL, µm/s), average path velocity (VAP, µm/s), linearity (LIN, %), straightness (STR, %), wobble (WOB, %), amplitude of lateral head displacement (ALH, µm) and beat/cross frequency (BCF, Hz).

Among kinematic parameters, the VCL is the most sensible to FR changes (Bompart et al. 2018; Valverde et al. 2019) and, therefore, this was the parameter used for the calculation of the optimum FR.

Counting chambers

In general, the design of this study was based on the recommendations made by Björndahl et al. (2016). The effects of capillary-loaded, disposable, 10 and 20 µm depth chambers (ISAS^® D4C10 and D4C20, respectively, Proiser R + D), and drop-displacement, reusable, 10 and 20 µm depth (Spermtrack^®, Proiser R + D) counting chambers were tested. All the chambers were prewarmed to 37 °C in a heating plate before use. Disposable chambers were loaded by capillarity with a volume of 2 and 3 µL (for 10 and 20 µm depth, respectively) maintaining the tip in contact with the coverslip until the fluid reached the end of the counting track. In these chambers, seven microscopic fields (one in each chamber square) were analyzed (Figure 2). The reusable chamber was loaded placing a 2 and 4 µl drop (depending on the depth) in the center of the base and immediately placing the cover, with no further movement of the cover. In this chamber, nine fields were captured, one in the center (white area), 4 in the mid-circle (light-grey area), and 4 in the outside-circle (dark grey area) (Figure 3). An average of 500 sperm was captured per field. To avoid possible errors related to biases in examining samples in the different chambers, the counting chambers were assessed randomly. In both cases (Figures 2 and 3), they were the same independent of the depth of the used chamber. Comparison amongst the counting chamber types and counting areas inside each chamber were done at 150 fps.

Figure 2. Disposable counting chambers (ISAS D4C^®) and forms used and the fields analyzed. ISAS D4C^® is a commercial disposable counting chamber, available in two different depths: 10 µm (D4C10) and 20 µm depth (D4C20). These are counting chambers based in capillarity. The arrow demonstrates the place of drop deposition. Numbers (1–7) inside the chambers demonstrate different capture fields.

Figure 3. Reusable counting chambers (Spermtrack^®) and forms used and the fields analyzed. Spermtrack^® is a commercial reusable counting chamber, available in two different depths: 10 µm (SpK 10) and 20 µm depth (SpK 20). These are counting chambers based in drop displacement. The arrow demonstrates the place of drop deposition. Drop deposition is done in the center of the circle (white area); the covers were rapidly but gently put in place to achieve a homogenous distribution of the sample. There are 3 different colors (grayscale) that mark the different capture fields and the symbol indicates the specific point of capture.

Statistical analysis

The data obtained from the evaluations of all ejaculates were analyzed by descriptive statistics. Distribution properties for all variables were also explored using histograms and probability plots to check for a normal distribution. The analysis of variance was further applied to evaluate statistical differences between counting chambers on seminal parameters. Furthermore, the effect of the field within the counting chambers was analyzed, also by analysis of the variance, for all kinematic variables.

The statistical model used was: $$X_{\mathit{ijk}}=\mu +C_i+F_j+C{F}_{(ij)}+{\epsilon }_{\mathit{ijk}}$$

Where X_ijk = Measured sperm kinematic variable; µ = Overall mean of variable x; A_i = Effect of counting chamber B_j = Effect of field; AB _(ij) = Effect of interaction between counting chamber *field; ε_ijk = Residual.

Calculating the optimum frame rate

For regression analyses, the effects of FR were tested in an exponential model, in the form: $$\mathrm{y}=\beta *\alpha \hspace{0.17em}\exp \hspace{0.17em}\left(-\beta /x\right)$$

Where y is VCL and x is FR, α is the asymptotic level, β is the rate of increase to the asymptote, and exp is the base of natural logarithms. The biological significance of the equation is that the asymptotic values α represent the maximum achievable when the FR is above the threshold level. The threshold level was calculated as the FR needed to obtain 95% of the maximum value. The rate of the approach to the asymptote represents the dependence of the curve on the FR; that is, a high value of β indicates high growth of VCL as FR increases and vice versa.

Following Mortimer et al. (2015) only one decimal was expressed for most of the parameters. Statistical significance was considered at p < 0.05. Furthermore, pairwise comparison between means effects of counting chambers and fields were performed by the Tukey–Kramer test. The results are presented as the mean ± standard error of the mean (SEM). All data were analyzed with IBM SPSS package, version 23.0 for Windows (SPSS Inc., Chicago, IL, USA).

Ethics approval

All work with semen samples was carried out with patients approved by the Ethic Committee of the IVI- University of València (1705-UV-035-AG). All signed informed consent forms to participate in and provide semen samples for the study. These forms indicate the purpose of the study and the fact that the samples will not be used for any other purpose as they would be destroyed after their use for the study.

Authors’ contributions

Conceptualization: AGM, NG, CS; methodology and investigation: AGM, CC, NN; validation: AGM, SS; formal and statistical analysis: AV; writing: AGM, CS. All authors have read and agreed to the published version of the manuscript.

Abbreviations
WHO	=	World Health Organization
CASA	=	computer assisted semen analysis
CV	=	coefficient of variation
VCL	=	curvilinear velocity
FR	=	frame rate
ALH	=	amplitude of lateral head displacement
BCF	=	beat/cross frequency
VAP	=	average path velocity
CASA-Mot	=	CASA-Motility
VSL	=	straight-line velocity
NA	=	numerical aperture
LIN	=	linearity
STR	=	straightness
WOB	=	wobble
fps	=	frames per second
SEM	=	standard error of the mean
SpK	=	reusable counting chamber (Spermtrack®)
D4C	=	disposable counting chamber

Acknowledgment

This study did not receive any specific funding.

Disclosure statement

No potential conflict of interest was reported by the author(s).