Analyzing human fertility using Artificial Intelligence
Infertility affects around 15% of couples globally. It is estimated that men contribute to 20-30% of infertility cases according to Agarwall et al. (2015), however, it is hard to determine an accurate percentage. The reason for this being that the amount of tests undergone on spermatozoa are very limited as most fertility tests tend to be undergone on women. This makes it difficult to fully understand male fertility. This issue has attracted a lot of attention with studies showing that, globally, there has been a decline in semen quality during the past few decades.
Currently, semen samples tend to be evaluated manually usually, using a microscope. These techniques are known to be lengthy and time-consuming while requiring professionals with substantial expertise in the specific field of male fertility. Hence, new technologies are needed in order to perfect and accelerate fertility analysis on men. Current improvements in artificial intelligence and machine learning, specifically deep learning, have proven that these technologies can be used to undertake fertility analyses on spermatozoa. Deep learning can be used to imitate the functions of the human brain in processing data and creating patterns that enable faster and more precise analyses.
Kantify is developing an Artificial Intelligence software technology that is able to analyse microscopical images of human semen in real time with high accuracy and speed using Artificial Intelligence.
Purpose
The Kantify spermatozoa analysis technology is being developed to help clinicians diagnose and characterize male fertility pathologies based on videos of human semen samples. The Kantify team has designed this technology with the following objectives:
-
Help with a faster and more accurate diagnosis of male fertility issues
-
Enable healthcare practitioners to perform tasks that are more suited to their expertise, such as interpreting the insights provided by the technology and defining a treatment
Key Features
In a nutshell, Kantify’s spermatozoa analysis technology:
-
Automatically analyzes video streams of semen observations in real time
-
Relies on state of the art Deep Learning techniques that reach high performance
-
Generates insights (metrics) based on the video streams
Metrics
The metrics generated through by the software technology are designed to help clinicians in their work. The following metrics can be assessed by the technology:
-
The total sperm count
-
The percentage of sperm vitality: percentage of live spermatozoa
-
The percentage of normal spermatozoa
-
The percentage of head defects: percentage of spermatozoa with abnormal head morphology
-
The percentage of tail defects: percentage of spermatozoa with tail abnormalities
-
The percentage of progressive mobility: spermatozoa with active movement, the forward or large circle movement
-
The percentage of non progressive sperm motility: spermatozoa with movement but with no forward progression
-
The percentage of immotile sperm: no movement at all
The insights can therefore be used to detect fertility issues in order to help design appropriate protocols and treatments.
Benefits
1.Scaling lab automation: It equips laboratories with a software that can efficiently analyse multiple samples
2.Decreasing operational costs: It Helps analysts focus only on their core job as it automates the rest
3.Providing better insights: It enables better early detection of more pathologies (immature granulocytes, parasitized cells, etc.)
4.Going beyond human analysis: It helps in eliminating human error by easily expanding the scope of what is analyzed, at a lower cost, with the same sample volume
If you are considering AI applications for human fertility analyses, don’t hesitate to contact us. This use case study focuses on male fertility and can be adapted to female fertility. Our experts are currently working on a variety of AI solutions in the healthcare industry and would be happy to help.