Technical note: Validation of an automated in-line milk progesterone analysis system to diagnose pregnancy in dairy cattle
Bibliographic record
Abstract
The in-line milk analysis system (IMAS) is an automated biosensor technology that samples and quantifies milk progesterone concentrations (P4c) at frequent intervals starting early postpartum until pregnancy. The objective was to validate the use of pregnancy notifications (PregN) generated by an IMAS based on P4c profiles after artificial insemination (AI) to determine pregnancy and nonpregnancy status in dairy cows. Records of 1,821 AI events from 715 Holstein cows that had milk P4c (ng/mL) measured every 2.2 ± 1.9 d (mean ± standard deviation) between 24.5 ± 8.2 and 173.4 ± 49.3 d in milk through a real-time IMAS (Herd Navigator, DeLaval International, Tumba, Sweden) were evaluated. Based on variations in adjusted milk P4c (< vs. ≥ the 5.0 ng/mL threshold), the system determined the sampling frequency, onset and cessation of luteal phases, and pregnancy. If a luteal phase initiated (P4c increased to ≥5.0 ng/mL) after AI and remained uninterrupted, a PregN was generated starting at (mean ± standard deviation) 31.0 ± 4.3 d until 53.4 ± 7.9 d after AI, when sampling stopped, unless a decline in P4c (to <5.0 ng/mL) occurred indicating nonpregnancy and imminent estrus. The assessment of IMAS PregN at 4 weekly intervals was tested, and a confirmed calving occurrence between 262 and 296 d after AI, with no other subsequent AI recorded, was the gold standard for pregnancy. In total, 14.1 (256/1,821), 41.0 (746/1,821), and 50.7% (924/1,821) of AI events were followed by a decline in P4c before 19, 23, and 30 d after AI, respectively. Frequency of the last 3 sampling events preceding P4c decline was greater if P4c decline occurred between 18 and 25 d after AI (1.4 ± 0.5 samples per day) compared with before 17 or beyond 26 d after AI (1.0 ± 0.5 samples per day). At 30 ± 3 (27 to 33) d after AI, PregN occurred in 46.8% (853/1,821) of AI events, of which 15.2% (130/853) had a decline in P4c between 30 and 55 d after AI and 17.1% (146/853) was later confirmed nonpregnant based on the gold standard. A total of 40.7% (742/1,821) of AI events was confirmed pregnant by the gold standard, which was no different than the proportion of PregN at 51 ± 3 (48 to 54) d (40.9%; 744/1,821). At any time point between 27 and 54 d after AI, sensitivity and negative predictive values for PregN were greater than 95.0 and 96.0%, respectively, whereas specificity values were less than 90.0% for PregN before 40 d but greater than 94.0% for PregN beyond 41 d after AI. In conclusion, IMAS is able to diagnose pregnancy based on P4c profiles with high precision and determine early nonpregnancy based on the spontaneous cessation of the luteal phase. However, for accuracy greater than 95.0%, pregnancy declaration based on IMAS notifications alone should occur no earlier than 41 d after AI.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.010 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".