Deep Learning-Based Morphological Classification of Human Sperm Heads

Abstract

Human infertility is considered as a serious disease of the reproductive system that affects more than 10% of couples across the globe and over 30% of the reported cases are related to men. The crucial step in the assessment of male infertility and subfertility is semen analysis that strongly depends on the sperm head morphology, i.e., the shape and size of the head of a spermatozoon. However, in medical diagnosis, the morphology of the sperm head is determined manually, and heavily depends on the expertise of the clinician. Moreover, this assessment as well as the morphological classification of human sperm heads are laborious and non-repeatable, and there is also a high degree of inter and intra-laboratory variability in the results. In order to overcome these problems, we propose a specialized convolutional neural network (CNN) architecture to accurately classify human sperm heads based on sperm images. It is carefully designed with several layers, and multiple filter sizes, but fewer filters and parameters to improve efficiency and effectiveness. It is demonstrated that our proposed architecture outperforms state-of-the-art methods, exhibiting 88% recall on the SCIAN dataset in the total agreement setting and 95% recall on the HuSHeM dataset for the classification of human sperm heads. Our proposed method shows the potential of deep learning to surpass embryologists in terms of reliability, throughput, and accuracy.

Importance of real-time measurement of sperm head morphology in intracytoplasmic sperm injection

“…In addition, automatic classification using a gold standard for morphological sperm dataset (SCIAN-MorphoSpermGS; Chang et al, 2017a) and support vector machines (SVMs; Chang et al, 2017b), and a combination of interferometric phase microscopy and SVMs have been reported (Mirsky et al, 2017). Recently, Iqbal et al (2020) established a morphological classification of human sperm heads (MC-HSH) that has higher accuracy than previously reported methods that use the SCIAN or HuSHeM datasets. Unfortunately, most of these analytical methods are based on images captured at ultrahigh magnification (×6300-10,000) using fluorescence microscopy.…”

Bayesian Pixel Density Estimation Modeling to Detect Human Sperm Sample Image Based on Sperm Head Shape

“…The most visible part of a human sperm image sample is the head of the sperm (Figure 1). This sample is called HSH (Iqbal et al, 2020). A normal HSH image has the head as a point of interest position with a single best guess to detect the head position (Leung et al, 2011).…”

“…Human sperm is unique in shape, which is characterized by the shape of the head, body, and tail. Healthy sperm has a standard head shape, called human sperm heads (HSH) (Iqbal et al, 2020). A method to detect images containing normal HSH based on certain criteria and image processing techniques is needed by medical laboratories (Yoon et al, 2008).…”

Deep Learning Based Evaluation of Spermatozoid Motility for Artificial Insemination

“…In another paper, Ilhan et al [ 25 ] suggested using the Mobile-Net neural network, which has a relatively low number of parameters, for the classification of normal/ab-normal sperms, achieving 87% accuracy. Iqbal et al [ 26 ] developed a custom convolutional neural network (CNN) architecture to categorize human sperm heads, achieving 88% recall on the SCIAN dataset and 95% recall on the HuSHeM dataset. Javadi and Mirroshandel [ 27 ] trained a custom CNN to recognize morphological abnormalities of sperm heads and achieved over 83% accuracy in recognizing acrosome, head, and vacuole defects.…”