A novel deep learning method for automatic assessment of human sperm images
Graphical abstract
Introduction
Infertility is the problem of almost 15% of couples. The lack of pregnancy after 12 months of intercourse without protection is defined as infertility. About 30 to 40 percent of infertile couples suffer from male factor abnormalities [20,43]. One of the problems in male factor infertility is spermatozoa morphology abnormalities, which may show teratozoospermia or oligoasthenoteratozoospermia.
The quality of spermatozoa is one of the most important parameters for oocyte fertilization and embryo quality. It is shown that abnormalities in sperm correlate with cleaving embryo morphology at later stages [7]. In other words, the shape of sperm is reflected by sperm development during spermatogenesis. As a result, problems in sperm maturation causes abnormalities in sperm morphology and the functionality of egg fertilization [2]. By assessment of sperm parameters and seminal plasma characteristics such as semen pH and sperm morphology, viscosity, concentration, and motility, male factor infertility can be determined [6].
The first successful live birth of a child using intra-cytoplasmic sperm injection (ICSI) method happened in 1992 [33]. These days, the ICSI method is widely used for the treatment of various couples who need assisted reproductive technologies. These couples can have normal, mildly, or severely abnormal semen parameters [5]. In several studies, the positive correlation between high ICSI outcomes and normal sperm morphology has been proven. In other words, serious abnormalities of the sperm head cause low fertilization, implantation, and pregnancy rates [30].
The normal sperm morphology is defined in previous studies by Menkveld et al. [28]. The length of a normal sperm head is between 3 and 5 micrometers. This range for the width of the sperm head is between 2 and 3 micrometers (two-thirds or three-fifths of head length). The length of midpiece is one and a half of the head length and it is axially presented in width. The normal tail should be also visible, uniform, uncoiled and thinner than the midpiece. Its length must be about .
In the intracytoplasmic morphologically selected sperm injection (IMSI) procedure, sperm selection is performed at high magnification (usually 6000) [30]. However, existing microscopes of laboratories commonly have a low magnification (400 and 600). These magnification levels are routinely utilized for sperm selection in the ICSI procedure. The visual assessment of sperms are also commonly performed manually and it is only based on the judgment of embryologists. This method is inexact, subjective, non-repeatable, and hard to teach. Another solution for the assessment of men fertility is computer-aided sperm analysis (CASA) using different staining procedures [6]. Due to the flaws of manual solutions, automatic techniques are essential for analyzing human sperm morphology. As a result, designing efficient and accurate algorithms for analyzing and classification of sperms and selecting the best one before ICSI is a challenging and trending task [38]. In other words, automatic methods for selection of the best sperm during the ICSI process and without staining will be more desirable for embryologists. This will lead to higher fertilization and pregnancy rates.
In this paper, we propose a novel deep learning algorithm for automatic extraction of sperm morphological features. This problem is a challenging task, due to the following reasons: 1) the number of sperm images is not enough for the training phase; 2) the normal and abnormal sperm classes are highly imbalanced, thus making the problem harder; 3) the pictures are taken using a low-magnification microscope and the details of these images are not clear; 4) the pictures are very noisy; 5) the sperms should not be stained; and 6) the analysis should be done in real-time to be useful for treatment purposes.
As the first step, we introduce a new dataset for our deep learning purpose. This dataset is based on the Human Sperm Morphology Analysis dataset (HSMA-DS) [15]. It is a unique dataset based on the number of images and its characteristics. For solving problems of training sample scarcity and class imbalance, we apply data augmentation and sampling techniques, respectively. Then, we propose a deep neural network architecture that can be trained to classify the normality of sperm acrosome, head, and vacuole. The proposed model has 24 convolution, two max pooling, one average pooling, and two fully-connected layers. Our proposed model, unlike a lot of existing methods, is able to work well with non-stained pictures. In other words, our method is applicable for treatment purposes. In addition, the time required for checking each sperm is very short (under 25 milliseconds) and our method can work in real-time. Our experimental results also show the effectiveness of our method, which can be regarded as the state-of-the-art for this dataset.
The rest of the paper is organized as follows: Some previous works are reviewed in Section 2. The prepared dataset is presented in Section 3. Our proposed algorithm, including the designed deep learning model, is depicted in Section 4. Sections 5 and 6 contain a comprehensive description of experimental setup, comparisons, and discussion. Finally, conclusions and future work are summarized in Section 7.
Section snippets
Previous work
There are lots of research on automatic selection of sperms. In one of these studies, the fraction of boar spermatozoa heads was calculated and a pattern of intracellular density distribution was recognized as normal [37,38]. In this method, a deviation model was defined and computed for each sperm's head. Then, for each sperm classification, an optimal value was considered. Afterward, by applying morphological closing, sperms tails were removed and the holes in the contours on the heads were
Dataset
We introduce a new dataset called the Modified Human Sperm Morphology Analysis dataset (MHSMA), which is based on the Human Sperm Morphology Analysis dataset (HSMA-DS) [15]. The version of HSMA-DS we used contains 1,540 RGB images with a size of 1280 1024 pixels. These images were taken at a magnification of either 400 (706 images) or 600 (834 images) using a microscope (IX70, Olympus, Japan) equipped with a CCD camera (DP71, Olympus, Japan) with chromatic infinity objective lenses [15].
Proposed method
We propose a machine learning-based approach for detecting abnormalities in sperm morphology. We train a convolutional neural network to classify sperm images into normal (positive) and abnormal (negative) classes for different morphological features.
As noted in Table 1, there are only 69 samples for abnormal tail and neck. Also, while detecting abnormalities in sperm head is complicated, detection of abnormal tail and neck is usually fairly easy for human experts. As a result, we will focus on
Results
We trained the model independently for classification of sperm acrosome, head, and vacuole. For each one, the model is trained for 10,000 iterations on the training set (1000 samples). After each iteration, loss value is calculated on the validation set (240 samples) and the checkpoint with the lowest validation loss is saved. Once training is done, we evaluate the saved checkpoint on the held-out test set (300 samples).
Fig. 8 shows training and validation loss for each label over iterations of
Discussion
Ability to work with non-stained images from low-magnification microscopes is one of the main features of our proposed deep learning method. Almost all existing algorithms work on stained images from high-magnification microscopes. Our method is also capable of processing each sperm image in real-time (i.e., less than a second). These characteristics are desirable features that are important for treatments applications.
As mentioned earlier, we marked the position of the sperm head in each image
Conclusion
In this paper, we have proposed a deep learning method for selecting the best sperms in the ICSI procedure. First of all, we have prepared a unique dataset that contains 1,540 images from different types of sperms. This dataset is freely available to the public. Our proposed deep learning model is a compact and accurate model for classification of sperms. This model extracts features of acrosome, head shape, and vacuole from sperm images. In almost all previous studies, sperms were fixed,
Conflict of interest statement
None declared.
Acknowledgement
The authors thank Dr. Fatemeh Ghasemian for her valuable efforts for dataset annotation and her scientific advises.
References (47)
- et al.
Automatic classification of the acrosome status of boar spermatozoa using digital image processing and LVQ
Comput. Biol. Med.
(2008) - et al.
Gold-standard for computer-assisted morphological sperm analysis
Comput. Biol. Med.
(2017) - et al.
Gold-standard and improved framework for sperm head segmentation
Comput. Methods Progr. Biomed.
(2014) - et al.
An efficient method for automatic morphological abnormality detection from human sperm images
Comput. Methods Progr. Biomed.
(2015) - et al.
A novel data acquisition and analyzing approach to spermiogram tests
Biomed. Signal Process. Control
(2018) - et al.
Treatment of male infertility
Contraception
(2005) - et al.
Pregnancies after intracytoplasmic injection of single spermatozoon into an oocyte
The Lancet
(1992) - et al.
A dictionary learning approach for human sperm heads classification
Comput. Biol. Med.
(2017) - et al.
Is there a role for the nuclear export factor 2 gene in male infertility?
Fertil. Steril.
(2008) - et al.
A comparative study of sperm morphometric subpopulations in cattle, goat, sheep and pigs using a computer-assisted fluorescence method (CASMA-F)
Anim. Reprod. Sci.
(2013)
Automatic evaluation of ram sperm morphometry
Theriogenology
Tensorflow: a System for Large-Scale Machine Learning
Spermatozoa segmentation and morphological parameter analysis based detection of teratozoospermia
Int. J. Comput. Appl.
Image processing with ImageJ
Biophot. Int.
Intracytoplasmic sperm injection indications: how rigorous?
Fully automatic identification and discrimination of sperm's parts in microscopic images of stained human semen smear
J. Biomed. Sci. Eng.
Eliminating the effect of pathomorphologically formed sperm on resulting gravidity using the intracytoplasmic sperm injection method
Exp. Ther. Med.
Automatic human sperm concentration in microscopic videos
Med. Technol. J.
A computer aided tool for the assessment of human sperm morphology
Keras
Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)
Automated non-invasive measurement of single sperms motility and morphology
IEEE Trans. Med. Imaging
Fast R-CNN
Cited by (70)
Multidimensional morphological analysis of live sperm based on multiple-target tracking
2024, Computational and Structural Biotechnology JournalDeep learning-based sperm motility and morphology estimation on stacked color-coded MotionFlow
2024, Informatics in Medicine UnlockedImage trinarization using a partial differential equation: A novel approach to automatic sperm image analysis
2024, Applied Mathematical ModellingEnsembling noisy segmentation masks of blurred sperm images
2023, Computers in Biology and MedicineImproving outcomes of assisted reproductive technologies using artificial intelligence for sperm selection
2023, Fertility and Sterility