**3. Morphological imaging**

Coppola's group [33] investigated the possibility to better understand the sperm behavior by means of a quantitative analysis of the 3D spermatozoa's morphology. In particular, experiments on bovine sperm cells were performed. The recorded hologram is illustrated in **Figure 2**, whereas the intensity of the fringe pattern generated by the interference between the object and reference beam is highlighted into the inset.

In **Figure 3**, the reconstructed images of abnormal bovine spermatozoa are reported [33]. For the reported analysis, the spermatozoa were fixed and without the surrounding liquid. In particular, the retrieved image shown in **Figure 3(a)** is relative to a spermatozoon with a cytoplasmatic droplet along the tail. Cytoplasm surrounding the sperm cell is accumulated during maturation, and it is extruded from the cell in the last phases of this maturation. However, cytoplasmatic residues may persist in the cell and, in particular, are retained in the tail as a droplet [34]. Thus, the presence of drops along the tail is connected to the degree of cell maturation and may indicate an excessive utilization of a donor.

In **Figure 3(b)**, an image of a spermatozoon with a bent tail is shown. Generally, when this defect is present in the semen either before or after the freezing process, the donor may be afflicted with a reproductive problem. On the other hand, if this anomaly appears with high frequency only in frozen semen, it can indicate that the spermatozoa have been subjected to hypoosmotic stress possibly due to an improper use of freezing extender and to an extremely low concentration of solutes. Finally, the reconstructed image of a spermatozoon with broken acrosome is illustrated in **Figure 3(c)**. In particular, the loss of acrosomal substances indicates premature acrosome activation far from the site of fertilization. This defect is present with high percentage in frozen semen samples due to incorrect sperm handling during the freezing process. The great advantage of managing quantitative information allows carrying out different numerical analysis, such as estimation area/volume, profiles along particular

procedure allows retrieving a discrete version of the complex optical wavefront (amplitude and phase) present on the surface of the specimen under test. This optical wavefront is obtained multiplying the recorded hologram by a numerical replica of the reference beam and numerically backpropagating this product. Actually, this product generates three diffraction terms: zeroth order, real image, and conjugate image. In an on-axis configuration, unless a partially coherent light is adopted that allows reducing the speckle and multi-reflection interference noise [29], all terms are superimposed. Thus, in order to recover only the real image, i.e., an exact replica of the object wavefront, either a spatial or temporal phase-shifting methods has to be employed. However, in this way the complexity and capture time increase [13]. On the other hand, by introducing a small angle between the object and reference beam (off-axis configuration), a spatial separation between the three terms is obtained, at the expense of suboptimal use of camera sensor space-bandwidth product. Thus, this separation allows selecting and retrieving only the real image. The possibility offered by DH to numerically retrieve the phase distribution of the object wavefront allows not only the possibility to evaluate the object morphology but also to remove and/or compensate the unwanted wavefront variation (such as optical aberrations and slide deformations) [30–32].

Coppola's group [33] investigated the possibility to better understand the sperm behavior by means of a quantitative analysis of the 3D spermatozoa's morphology. In particular, experiments on bovine sperm cells were performed. The recorded hologram is illustrated in **Figure 2**, whereas the intensity of the fringe pattern generated by the interference between the

In **Figure 3**, the reconstructed images of abnormal bovine spermatozoa are reported [33]. For the reported analysis, the spermatozoa were fixed and without the surrounding liquid. In particular, the retrieved image shown in **Figure 3(a)** is relative to a spermatozoon with a cytoplasmatic droplet along the tail. Cytoplasm surrounding the sperm cell is accumulated during maturation, and it is extruded from the cell in the last phases of this maturation. However, cytoplasmatic residues may persist in the cell and, in particular, are retained in the tail as a droplet [34]. Thus, the presence of drops along the tail is connected to the degree of cell matu-

In **Figure 3(b)**, an image of a spermatozoon with a bent tail is shown. Generally, when this defect is present in the semen either before or after the freezing process, the donor may be afflicted with a reproductive problem. On the other hand, if this anomaly appears with high frequency only in frozen semen, it can indicate that the spermatozoa have been subjected to hypoosmotic stress possibly due to an improper use of freezing extender and to an extremely low concentration of solutes. Finally, the reconstructed image of a spermatozoon with broken acrosome is illustrated in **Figure 3(c)**. In particular, the loss of acrosomal substances indicates premature acrosome activation far from the site of fertilization. This defect is present with high percentage in frozen semen samples due to incorrect sperm handling during the freezing process. The great advantage of managing quantitative information allows carrying out different numerical analysis, such as estimation area/volume, profiles along particular

**3. Morphological imaging**

338 Holographic Materials and Optical Systems

object and reference beam is highlighted into the inset.

ration and may indicate an excessive utilization of a donor.

**Figure 2.** Acquired hologram, a region is enhanced in order to show the interference pattern (inset). Ref. [33] (by permission of IEEE Society).

**Figure 3.** Pseudo 3D representation of the thickness of a spermatozoon with: (a) a cytoplasmatic droplet along the tail; (b) a bent tail; (c) an acrosome broken. Ref. [33] (by permission of IEEE Society).

directions, and selection of different zones. The 3D analysis can add further information to the data provided by the traditional bi-dimensional optical techniques, allowing to better understand the relationship between the male infertility and the abnormal morphology. Due to this potentiality, the holographic approach has been also employed to analyze the human sperm characteristics [35, 36]. Crha et al. [35] tested about 3000 sperm cells to individuate a phase difference between spermatozoa in normo-zoospermia (NZ) and oligoasthenoteratozoospermia (OAT). **Table 1** summarizes the obtained results in terms of mean, median, standard deviation (SD), and confidence intervals (CI).

The Dale's group performed a comparison between the results of spermatozoa analyzed both by a semiautomated digitally enhanced Nomarski microscopy (DESA) and by the holographic imaging [36]. In **Table 2**, the values relative to five primary parameters (length, width, perimeter, area, and volume) for normal human spermatozoa are compared. Results shows that no significant differences were observed in the gross morphometric values of the sperm cells analyzed.


**Table 1.** Descriptive statistics for statistically significant (*ρ* < 0.001) phase shifts according to the NZ/OAT group [35].


**Table 2.** Mean morphometric values of normal human sperm heads obtained by DESA and holographic techniques [36].

It is important to note that the volume estimation could not be performed by the DESA technique. Due the great influence of nuclear vacuoles in the sperm head on the fertilization capacitance of sperm cell [9], the volumetric analysis was performed to analyze vacuolated human spermatozoa. In **Figure 4(a)**, a conventional differential interference contrast (DIC) image of a vacuolated human sperm is reported. The holographic 3D reconstruction of the vacuolated spermatozoon is illustrated in **Figure 4(b)**. From this figure, it is clear that spermatozoa with vacuole had a reduced volume, and this reduction could be probably be due to variation of the inner structure of the sperm head with a loss of material.

**Table 3** summarizes the volumetric analysis carried out for three different groups of spermatozoa defined by the length and width of the head. Mean values of the total volume of the spermatozoa minus the vacuoles volume are also reported.

Memmolo et al. [37] proposed to use DH in order to identify and measure specific region-ofinterest of spermatozoa. **Figure 5** shows some steps of the applied method that starts from a filtered version of the reconstructed spermatozoon image and is able to identify the head region.

**Figure 4.** Differential interference contrast image (a) and pseudo 3D holographic reconstruction (b) of a vacuolated sperm head. Ref [36] (by permission of Cambridge University Press).


**Table 3.** Mean volumetric values of vacuolated sperm clustered in three different subpopulations [36].

**Figure 5.** Detection of sperm head by the algorithm proposed by Memmolo et al. [37] (by permission of The Optical Society): (a) and (d) are the results of the denoising, (b) and (e) are the results of extraction algorithm, (c) and (f) are the best fit ellipses.

The proposed algorithm could be very useful to retrieve noisy holograms due to the impurity of the liquid surrounding the spermatozoa.
