STAR (Sperm Tracking and Recovery) System
Overview
The STAR (Sperm Tracking and Recovery) system is an AI-guided, microfluidic platform developed at the Columbia University Fertility Center to identify and isolate rare sperm cells in men diagnosed with azoospermia — a condition in which the ejaculate contains little or no sperm. The technology integrates high-speed imaging, artificial intelligence, microfluidics, and robotics to perform non-invasive sperm detection and retrieval.
The Clinical Problem
Male factor infertility accounts for up to 40% of infertility cases. Azoospermia and cryptozoospermia — characterized by absent or extremely rare sperm in the ejaculate — comprise approximately 10–15% of male infertility cases. For affected couples, diagnosis and treatment often involve years of repeated failed interventions, invasive procedures, and significant emotional distress. Historically, management options have included testicular sperm extraction (TESE) or prolonged manual sperm searches by skilled embryologists, followed by intracytoplasmic sperm injection (ICSI). These approaches are invasive, time-intensive, and frequently unsuccessful, with many couples ultimately advised to consider donor sperm or adoption (Suryawanshi et al., 2025).
Development & Research Team
The STAR system was developed over five years at the Columbia University Fertility Center, Department of Obstetrics and Gynecology, Columbia University Medical Center, New York, NY.
Principal Investigators:
- Hemant Suryawanshi, PhD — Assistant Professor of Reproductive Sciences, Columbia University Vagelos College of Physicians and Surgeons; Project Leader
- Zev Williams, MD, PhD — Wendy D. Havens Associate Professor of Women’s Health; Director, Columbia University Fertility Center; Senior Author
Additional authors (all Columbia University): Laura Gemmell, Stephanie Morgan, George Koustas, Robert W. Prosser, Ryan Fu, and Eric Forman.
The interdisciplinary team included specialists in advanced imaging, microfluidics, machine learning, robotics, and reproductive endocrinology (Columbia University Irving Medical Center, 2025).
For the best experience, view on a desktop or rotate your mobile device to landscape. Select the button to review more information
How the STAR System Works
The STAR system integrates three core technological components (Suryawanshi et al., 2025):
1. High-Speed Imaging The system employs high-powered imaging technology to scan an entire semen sample, capturing over 8 million images in under one hour.
2. AI Detection — YOLOv8 Architecture Based on the You Only Look Once version 8 (YOLOv8) object detection architecture, the AI model divides each image frame into a grid and predicts bounding boxes and confidence scores for sperm candidates in a single pass. A temporal consistency filter tracks objects across approximately ten consecutive frames to confirm detection, enhancing robustness and minimizing false positives.
- Precision: 0.89
- Recall: 0.90
- Mean Average Precision (mAP at IoU 0.5): 0.95
3. Microfluidic Chip & Robotic Retrieval A custom-designed Fusion DTx microfluidics chip — engraved with channels as thin as a human hair — isolates the portion of the semen sample containing the identified sperm cell. A robotic system then removes the individual sperm cell within milliseconds. Critically, this process avoids centrifugation, lasers, dyes, or other agents that can damage sperm, preserving viability for use in IVF or cryopreservation (ColumbiaDoctors, 2026).
Clinical Performance
- STAR detects approximately 40 times more sperm compared to manual searches by trained laboratory technicians.
- In cases where standard laboratory testing found no sperm, STAR found viable sperm in approximately 30% of azoospermia patients.
- In one documented pre-clinical case, skilled technicians searched a semen sample for two full days and found no sperm; STAR analyzed the same sample in one hour and identified 44 sperm cells (Columbia University Vagelos College of Physicians and Surgeons, 2025).
First Clinical Pregnancy
The first successful clinical pregnancy using the STAR method was reported in a peer-reviewed research letter published October 31, 2025 in The Lancet (Suryawanshi et al., 2025). The patient, a man in his late 30s diagnosed with non-obstructive azoospermia, had undergone nearly 20 years of infertility treatment — including multiple IVF cycles, several manual sperm searches, and two surgical sperm extraction procedures — all without success.
From a 3.5 mL semen sample, STAR scanned 2.5 million images in approximately two hours and identified 2 viable sperm cells. These were used to create two embryos. A successful embryo transfer resulted in the first clinical pregnancy achieved using the STAR method (Columbia University Irving Medical Center, 2025).
Weill Cornell Medicine & Male Factor Infertility: Context and Related Research
While no direct public statement from Weill Cornell specialists specifically addressing the STAR system has been identified in the peer-reviewed literature to date, Weill Cornell Medicine’s Center for Male Reproductive Medicine and Microsurgery — founded by Dr. Marc Goldstein in 1982 — is among the world’s foremost authorities on azoospermia management and directly relevant to the clinical landscape in which STAR operates.
Key Weill Cornell contributions to this field in 2025:
Dr. Marc Goldstein (Weill Cornell Medicine) co-authored a peer-reviewed book chapter published in April 2025 specifically examining the current and future role of AI in diagnosing and treating male infertility. The chapter, co-written with Dr. Jessica Marinaro, is published in Molecular Male Reproductive Medicine (Advances in Experimental Medicine and Biology, Vol. 1469) and addresses opportunities for AI-based computational tools in reproductive urology — the precise clinical space STAR occupies (Marinaro & Goldstein, 2025).
Dr. Peter N. Schlegel (Weill Cornell Medicine), Professor of Urology and Reproductive Medicine, is a senior author of the joint AUA/ASRM clinical guideline on male infertility — the standard of care document against which innovations like STAR are measured. He has also been cited in peer-reviewed systematic reviews examining AI predictive models for non-obstructive azoospermia (NOA) and microdissection TESE outcomes (Schlegel, 2023; Human Reproduction Open, 2025).
Editorial note for INCIID: Weill Cornell’s Center for Male Reproductive Medicine has established the clinical benchmarks — including guidelines for managing non-obstructive azoospermia — that make STAR’s results measurable and significant. Reaching out to Dr. Goldstein or Dr. Schlegel for direct comment on STAR would be a high-value next step for INCIID reporting.
Broader Scientific Context: AI in Male Infertility
A 2025 scoping review published in Current Urology Reports (PubMed/NIH) found that AI may offer proactive, cost-effective, and efficient management of male infertility across areas including semen analysis, hypogonadism, and assisted reproductive technology, but cautioned that quality of clinical care must remain paramount as AI integration expands (Naik et al., 2025).
A separate 2025 peer-reviewed review in Current Urology Reports (PubMed) concluded that AI techniques — including machine learning and artificial neural networks — outperform traditional methods by reducing subjectivity in sperm evaluation, identifying subtle abnormalities often missed during manual assessments, and enhancing sperm selection for ART (Nashed et al., 2025).
Disclosure of Competing Interests
Suryawanshi and Williams are inventors on patent applications filed by Columbia University related to this technology. All other authors declared no competing interests. De-identified technical and imaging data are available from the corresponding authors upon reasonable request, subject to institutional review and patient privacy regulations (Suryawanshi et al., 2025).
Sources restricted to peer-reviewed publications and accredited scientific/institutional authorities All citations in APA 7th Edition format
RESOURCE LIST
APA 7th Edition Format
I. Primary Peer-Reviewed Publication
Suryawanshi, H., Gemmell, L., Morgan, S., Koustas, G., Prosser, R. W., Fu, R., Forman, E., & Williams, Z. (2025). First clinical pregnancy following AI-based microfluidic sperm detection and recovery in non-obstructive azoospermia. The Lancet. https://doi.org/10.1016/S0140-6736(25)01623-X
II. Weill Cornell Medicine — Peer-Reviewed Publications
Marinaro, J., & Goldstein, M. (2025). Current and future applications of artificial intelligence to diagnose and treat male infertility. In C. Y. Cheng & F. Sun (Eds.), Molecular male reproductive medicine (Advances in Experimental Medicine and Biology, Vol. 1469, pp. 1–23). Springer. https://doi.org/10.1007/978-3-031-82990-1_1
Schlegel, P. N. (2023). New CUA guideline: A valuable reference for counselling men with azoospermia [Commentary]. Canadian Urological Association Journal. https://pmc.ncbi.nlm.nih.gov/articles/PMC10426417/
Lee, R., Li, P. S., Schlegel, P. N., & Goldstein, M. (2008). Reassessing reconstruction in the management of obstructive azoospermia: Reconstruction or sperm acquisition? Urologic Clinics of North America, 35(2). https://doi.org/10.1016/j.ucl.2008.01.007
III. Weill Cornell Medicine — Institutional Source
Weill Cornell Medicine, Center for Male Reproductive Medicine and Microsurgery. (n.d.). Center for male reproductive medicine and microsurgery. NewYork-Presbyterian Hospital / Weill Medical College of Cornell University. https://urology.weill.cornell.edu/research/center-male-reproductive-medicine-and-microsurgery
IV. Columbia University — Institutional & Academic Sources
Columbia University Irving Medical Center. (2025, October 31). First pregnancy with AI-guided sperm recovery method developed at Columbia [Press release]. EurekAlert! / American Association for the Advancement of Science. https://www.eurekalert.org/news-releases/1104028
ColumbiaDoctors. (2026, February). Columbia fertility looks to the stars to help men with infertility. Columbia University Department of Obstetrics and Gynecology. https://www.columbiadoctors.org/news/columbia-fertility-looks-stars-help-men-infertility
ColumbiaDoctors. (2025). Sperm recovery and analysis. Columbia University Fertility Center. https://www.columbiadoctors.org/specialties/obstetrics-gynecology/our-services/columbia-university-fertility-center/our-services/sperm-recovery-and-analysis
Columbia University Vagelos College of Physicians and Surgeons. (2025, Fall). Columbia develops STAR technology for men with infertility. Columbia Medicine Magazine. https://www.vagelos.columbia.edu/about-us/columbia-medicine-magazine/fall-2025/clinical-advances/columbia-develops-star-technology-men-infertility
NewYork-Presbyterian Hospital. (2025). Building an AI-powered system to improve fertility success [Audio podcast episode]. In Advances in Care, Episode 38. https://www.nyp.org/advances/podcast/building-an-ai-powered-system-to-improve-fertility-success
V. Peer-Reviewed Reviews — AI & Male Infertility (Broader Scientific Context)
Naik, N., Roth, B., & Lundy, S. D. (2025). Artificial intelligence for clinical management of male infertility: A scoping review. Current Urology Reports, 26. https://pmc.ncbi.nlm.nih.gov/articles/PMC11550229/
Nashed, J. Y., Liblik, K., Dergham, A., Witherspoon, L., & Flannigan, R. (2025). Artificial intelligence in andrology: A new frontier in male infertility diagnosis and treatment. Current Urology Reports, 26(1), 29. https://doi.org/10.1007/s11934-025-01257-5
Human Reproduction Open. (2025). AI predictive models and advancements in microdissection testicular sperm extraction for non-obstructive azoospermia: A systematic scoping review. Human Reproduction Open, 2025(1). https://doi.org/10.1093/hropen/hoae070
Soubry, A. (2025). Introducing artificial intelligence and sperm epigenetics in the fertility clinic: A novel foundation for diagnostics and prediction modelling. Frontiers in Reproductive Health. https://doi.org/10.3389/frph.2025.1506312
VI. Referenced AI Architecture (Cited in The Lancet Publication)
Varghese, R., & Sambath, M. (2024). YOLOv8: A novel object detection algorithm with enhanced performance and robustness. Proceedings of the 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS), April 18–19, 2024, 1–6.
Compiled for INCIID — The InterNational Council on Infertility Information Dissemination Last updated: May 2026 All citations formatted in APA 7th Edition