Skip to content
IVF BIG DATA

IVF BIG DATA

Personalized Lab Report

Verify your email, tell us your lab name, then upload your clinic's embryo file. Review Outcomes-Grades scoring and validation, and receive a PDF report personalized for your lab. Analysis uses only your upload — historic master files are not merged.

1. Lab name and email verification

Enter your lab name and email, then confirm the 6-digit code we send. Upload unlocks after verification.

2. Upload embryo file

Upload a CSV or Excel file (.csv, .xlsx, .xls). Required: embryo age (numeric ~0–100) and lab grade (e.g. 4AB). Optional when present: lab location, patient ID, Age at TVA, cycle ID, date created. Column names can vary — we match headers and column content.

Drag and drop a file here, or click to browse

Complete email verification above to unlock upload.

How Outcomes-Grades scores were designed

The site uses an additions-only Outcomes-Grades score derived from Gardner morphology. Gardner expansion (the leading number in grades such as 4AB) is parsed for reporting but is not scored. The score sums day-of-biopsy points, ICM (first letter), TE (second letter), and a no-minus bonus. Maximum score is 24.

score = day_points + ICM_points + TE_points + no_minus_points

Weights prioritize biopsy day and ICM over TE: day 5 receives the largest credit, ICM A is worth three times ICM C, and TE contributes a smaller range. Any minus on ICM or TE removes the six-point “no minus” bonus.

Day of biopsy
CategoryPoints
Day 510
Day 64
Day 70
ICM (first letter)
CategoryPoints
A6
B3
C0
TE (second letter)
CategoryPoints
A2
B1
C0
Minus modifier
CategoryPoints
No minus6
Any minus (−)0

Model validation and testing

Separately from the additions-only KPI score above, morphology features were evaluated against clinical pregnancy (IUP) on a historical FET morphology cohort. A logistic model was trained on single-embryo transfers (SET) so the outcome belongs to that embryo. The model score is 100 × P(IUP). This section summarizes discrimination and selection performance; it is supporting evidence for morphology–IUP association, not a bedside calculator.

Cohort

~7,113 FET rows

IUP ≈ 73%; SET train n ≈ 6,997

Validation

5-fold stratified CV

StratifiedKFold, out-of-fold predictions

Logistic AUC / AP

0.623 / 0.791

Heuristic AUC / AP ≈ 0.582 / 0.768

Top 10% selection

81.3% IUP

vs heuristic 78.1% (SET ranking)

Feature odds ratios (logistic)
FeatureOR
Day 6 (vs day 5)0.59
Day 7 (vs day 5)0.33
ICM A (vs B)1.45
TE A (vs B)1.10
Minus present0.55
IUP rate if keeping top-ranked SET embryos (%)
Keep topLogisticHeuristic
10%81.378.1
25%80.478.8
50%79.977.7
100%73.073.0

Informational only. Not a clinical decision device. Clinic protocols and physician judgment take precedence. The additions-only Outcomes-Grades KPI score used in charts on this page is not identical to the logistic 100 × P(IUP) model score.