A method that allows the recovery of sperm in samples from patients who produce very few of them is a new weapon against male infertility.
A method that uses AI to scan human seminal fluid for rare viable sperm has resulted in a pregnancy for the first time in a couple where the man was suffering from a severe form of infertility.
The system is called STAR (Sperm Tracking and Recovery) and is a microfluidic platform, i.e. a circuit that allows a fluid to be examined quickly and with very high precision: in this case, the sperm of men suffering from a condition that involves the presence of very rare specimens of spermatozoa suitable for fertilizing an egg cell.
STAR had already been presented months ago, but now it has allowed you to start a healthy pregnancy, as reported in an article published on Lancet.
A research at the limits of the impossible
Approximately 40% of couples’ infertility diagnoses are attributable to the male factor, i.e. an anomalous quality of semen. STAR is an automated and non-invasive platform developed by scientists at Columbia University (United States): it was designed to expand the possibilities of biological paternity in men diagnosed with male infertility due to cryptozoospermia, a very low concentration of sperm in the seminal fluid, less than 100,000 sperm per milliliter of fluid.
For reference, the WHO considers “healthy” a sperm concentration equal to or greater than 15 million specimens per milliliter. The number of male cells in patients with cryptozoospermia is so reduced that these patients are often mistakenly diagnosed with azoospermia, i.e. the total absence of spermatozoa.
The road so far: invasive and sometimes harmful tests
Until now, the search for rare spermatozoa that can be used for reproduction in patients with severe forms of male infertility has required invasive surgeries to remove seminal fluid directly from the testicles, a procedure that can cause inflammation and localized vascular problems. The analyzes of the samples must also be carried out in specialized centres, with slow, very expensive techniques and entrusted to the experience and manual skills of the technicians, which however can end up damaging the few healthy spermatozoa.
Help from AI
STAR consists of a high-speed imaging system that can acquire up to 8 million images in an hour, a microfluidic chip (i.e. a platform that guides the movement of seminal fluid through microchannels the thickness of a hair) and a sperm detection model based on machine learning. These components allow us to analyze sperm samples in real time and continuously, and isolate rare spermatozoa even in fluids initially classified as azoospermic, i.e. completely devoid of male sexual cells.
Once the chip has isolated a portion of liquid in which the presence of a sperm is recognized, within a few milliseconds a micro-robot removes that single cell, so that it can be used to create an embryo (or frozen for future use).
The first pregnancy with this method
The platform was tested on a patient who with his partner had faced infertility treatments for almost 20 years, including many attempts at assisted fertilization preceded by two surgical sperm extraction procedures and several manual analyzes of seminal fluid, all without success.
STAR analyzed 3.5 milliliters of the patient’s sperm, obtaining 2.5 million images in two hours, and identifying 2 usable sperm. The cells were used to create two embryos, one of which gave rise to a pregnancy (ongoing but now entrusted to obstetric care).
Clinical studies are underway to evaluate the feasibility of the approach on a larger patient population, but the hope is that the platform can offer patients who were denied the possibility of becoming biological fathers a chance to reproduce. Because, as Zev Williams, Director of the Fertility Center at Columbia University, explained: “It only takes one healthy sperm to create an embryo.”
