AI startup makes significant breakthrough in speeding up the diagnosis of male fertility

According to recent research, infertility affects one in seven couples in the UK, with an estimated seven per cent of men struggling with infertility. Furthermore, research has also found that the rate of male fertility has been increasingly declining over the past 40 years by more than 50 per cent.

The various factors affecting male fertility as outlined by Fertility Family include sperm DNA fragmentation, a process whereby sperm DNA is damaged through oxidative stress due to environmental and lifestyle factors. Another factor that may affect male fertility is sperm motility, which is the ability of sperm to swim through the female reproductive tract, with low levels of testosterone and sperm counts vastly affecting the ability to conceive.

The AI company Bayezian founded in 2020 by CEO Ed Dixon, who is behind the team, integrates "academic excellence" and "industry expertise" to use data that inspires solutions to real-world challenges. Bayezian recognises that "artificial intelligence has the potential to improve humanity for the greater good", and with a team of scientists, engineers, ethicists and more, they are committed to the dedicated application of artificial intelligence and the advancement of science to benefit the medical industry.

The diagnosis of fertility can be a slow and demanding process. Bayezian was first approached after a direct meeting with an individual who struggling with the challenges of fertility diagnosis. The AI company then became involved in the project starting 18 months ago, with CEO Ed Dixon commenting: "This project is the perfect example of the tech for a good approach that the team is undertaking. We see accurate diagnosis as a critical tool in helping address male fertility."

Following this, a team led by Bayezian have seen success, developing a data science algorithm to determine male fertility quicker and more accurately. The team used the Modified Human Sperm Morphology Analysis (MHSMA) dataset to build deep learning frameworks that can see a sperm's morphology.

The dataset consisted of a group of human sperm images from 235 surveyed patients with male factor infertility. Each of the images is expertly labelled for normal or abnormal sperm acrosome, head, vacuole and tail, giving researchers a better understanding of the causes and drawbacks in identifying male fertility and what can be done to improve this on a larger scale.

How was AI used as a tool in developing the data science algorithm by Bayezian?

Speaking with Dixon in an exclusive interview, he explained how "neural networks, specifically convolutional neural networks, have been extensively utilised for their exceptional capabilities in image and pattern recognition. In this case, it was applying them to differentiate between a classically fertile and infertile sperm cell".

Therefore, using the latest advancements in data science and AI, they can now identify sperm fertility at a faster rate, with a 96 per cent accuracy rate, which is two per cent higher than existing scientific approaches, improving the chances of diagnosis fundamentally.

According to Dixon, the significance of this breakthrough is that "our increased performance metrics enhance the precision of identifying fertility issues in males, leading to more targeted treatments and interventions".

He further stressed: "This serves as a critical tool by enabling healthcare professionals to provide personalised and effective solutions, ultimately improving male fertility outcomes and reproductive health."

As to whether this improved data science algorithm is going to be implemented by healthcare professionals in the public domain for a speedier diagnosis of male fertility is yet to be seen, with the team at Bayezian still hopeful that "in the near future" they will "aim to get this into the hands of researchers and clinicians".