News

On 23 July 2026, the Oocyte Spindle Research Network held its third session, on age-related changes in spindle morphology. Pavlo Mazur presented the results of the second and third blind validations of ReSpindle, covering 361 oocytes scored without knowing patient age.

The data shows score distributions shifting cleanly with maternal age, the most common score moving from A to B to C as age rises. The session covered what this means for spindle patterns across age groups, and how it translates into identifying, within a single patient, the egg-cell least affected by age-related change.

Full session recording available for OSRN members in the OSRN Library.

Constantine Hnennyi walks through the ReSpindle web platform: how to upload a polarised-light video, run the analysis, read the Score A, B or C result, and generate PDF reports for individual oocytes and the full cycle.

Available for OSRN members in the OSRN Library.

The second OSRN session. Pavlo Mazur draws on a dataset of over 2000 oocytes to walk through the edge cases embryologists meet during ICSI - no spindle visible, non-MII spindle, and atypical morphology.

Session recording available for OSRN members in the OSRN Library.

In partnership with the Alpha IVF Group, we completed the second and third blind validations of ReSpindle. In both, the clinic withheld patient age until after every oocyte was scored. Combined, that is 361 oocytes assessed fully blind.

ReSpindle reads the meiotic spindle from a short video and gives each oocyte one of three scores: A (young-looking), B (intermediate) or C (aged-looking).

Across the combined data, the distribution shifts cleanly with age. The most common score moves A → B → C as maternal age rises (p < 0.001). The algorithm consistently captures age-related change in spindle morphology, in proportions that match what is biologically expected from oocyte ageing.

We continue to see "spindle not found" appear on strong morphological deviations, such as twisted spindles - itself a meaningful signal about the oocyte.

Read on LinkedIn

On 26 May 2026, the Oocyte Spindle Research Network (OSRN) held its inaugural online gathering. More than 120 IVF professionals from over 39 countries registered for the live session.

Pavlo Mazur and Daniela Nogueira walked through two decades of spindle morphology research now arriving in clinical practice - AI spindle assessment, its implementation in ICSI routine, and the research directions that open from here. The live Q&A ran throughout the session, with embryologists from clinics around the world sharing their experience.

An 80-second teaser is available below. Full recordings of OSRN events are available to members.

OSRN brings together IVF professionals moving spindle morphology from research into clinical practice. To join, visit the OSRN page.

ReSpindle has completed its second independent blind validation, conducted with Alpha IVF & Women's Specialists (Malaysia) on a sample of 153 oocytes - more than five times larger than the first test. The system reached 72.5% accuracy on Score A and Score C classifications, up from 71.2% in the first study.

This validation used patient age as ground truth. ReSpindle identifies "young-looking" spindle morphology in older patients and "old-looking" morphology in younger ones, regardless of the patient's actual age. Score B is a meaningful third category - generally favourable in advanced-age patients and unfavourable in young ones.

The most striking finding is the natural emergence of a biological signal across age groups. In young patients (≤35 years, n=113), Score A dominates: 52 Score A, 35 Score B, 17 Score C, 11 MII spindle not found. In intermediate age (36-38, n=24), Score B leads clearly: 3 Score A, 11 Score B, 5 Score C, 5 MII spindle not found. In advanced maternal age (≥39, n=16), Score C is dominant: 5 Score A, 5 Score B, 6 Score C. This transition was not optimised for - it emerged directly from the data and reflects what is biologically expected from oocyte aging.

In practice, this means ReSpindle identifies the youngest-looking spindle in any cohort. For a 28-year-old patient, it flags the oocytes with unexpectedly poor spindle morphology. For a 42-year-old, it highlights the rare oocytes that still carry a young-looking signature - the ones most worth prioritising for transfer.

ReSpindle now performs reliably across all video quality levels. We are scaling the technology to more partner clinics worldwide.

Read on LinkedIn

In March 2026, ReSpindle completed the first independent blind validation of AI-powered oocyte spindle morphology analysis, conducted at Alpha IVF & Women's Specialists (Malaysia), an R&D partner since December 2024.

The AI system analyzed 186 unique polarized light images from 23 oocytes. All predictions were made before PGT-A results were disclosed.

Results: 71.2% accuracy between ReSpindle Score A/C classifications and PGT-A ground truth, with 25% of oocytes assigned Score B - a deliberate grey zone for cases with inconclusive spindle morphology.

These are early results on a limited sample. However, they represent the first independent blind validation confirming that AI analysis of oocyte spindle morphology via polarized light microscopy can correlate with PGT-A outcomes in a real clinical setting, without any contact with the embryo.

The dataset continues to grow as additional clinical partners contribute validation data.

Read on LinkedIn