Une révolution de l’IA redonne espoir aux hommes déclarés infertiles – Tchadinfos

Researchers at the Columbia University Fertility Center in the United States have developed an artificial intelligence-based technology designed to identify hidden sperm in men diagnosed with azoospermia. This condition is characterized by the apparent absence of sperm in the ejaculate, which often leaves couples facing significant challenges in achieving biological parenthood.

The new system, known as STAR (Sperm Track and Recovery), integrates high-performance imaging, microfluidic chips, and machine learning algorithms to locate rare sperm cells that typically evade traditional detection methods. According to reporting by Tchadinfos, which relayed information from BBC Afrique, the technology provides a new path for millions of couples dealing with male infertility.

The STAR system operates by analyzing samples at a high velocity of 300 images per second. This allows the AI to identify rare spermatozoa in real time, distinguishing them from surrounding cellular debris. Once a sperm cell is located, a robotic system extracts it within milliseconds to ensure its viability is preserved for fertility treatments.

The precision of the system marks a significant shift from manual laboratory techniques. Researchers state that the technology possesses a sensitivity of 100%, meaning it is capable of detecting a single sperm cell if one exists within the provided sample.

In clinical trials, the STAR system demonstrated a substantial increase in efficiency compared to human analysis. The system was able to identify up to 40 times more spermatozoa than a manual search conducted by an experienced technician.

Clinical Application and Patient Outcomes

The technology has already shown results for patients with complex genetic conditions. One case highlighted by the researchers involves a man identified by the pseudonym Samuel, who suffers from Klinefelter syndrome. This genetic anomaly is frequently associated with a total absence of sperm in the semen.

Prior to the use of the STAR system, medical professionals estimated that Samuel had only a 20% chance of becoming a biological father. However, the AI-driven recovery process allowed for the successful retrieval of sperm.

As a result of this intervention, Samuel’s partner, identified by the pseudonym Penelope, is currently pregnant. The couple expects their baby by the end of July 2026. If the pregnancy results in a boy, he could be the first male child conceived using this specific AI technology.

Medical Context of Azoospermia

Azoospermia is a primary cause of male infertility and can be categorized into different types based on whether the testes are producing sperm that cannot be transported (obstructive azoospermia) or whether the testes are failing to produce sperm entirely (non-obstructive azoospermia).

For men with non-obstructive azoospermia, such as those with Klinefelter syndrome, the production of sperm is often so minimal that traditional surgical retrieval methods may fail to find any viable cells. The ability of the STAR system to scan samples with 100% sensitivity and high speed reduces the risk of missing the few viable cells that may be present.

By automating the search and extraction process, the Columbia University Fertility Center aims to increase the success rates of assisted reproductive technologies for men who were previously told they were infertile.