Datasets:
id int64 1 10k | sex stringclasses 2
values | gestational_age_weeks float64 24.1 42 | birth_weight_g int64 590 4.97k | head_circumference_cm float64 18 42 | maternal_age_years int64 14 47 | parity int64 0 14 | delivery_mode stringclasses 2
values | apgar_1min int64 0 10 | apgar_5min int64 0 10 | temperature_c float64 34.2 39.5 | heart_rate_bpm int64 61 192 | respiratory_rate_bpm int64 27 100 | spo2_percent int64 73 100 | bw_category stringclasses 5
values | preterm_category stringclasses 4
values | primary_outcome stringclasses 6
values |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | F | 35.4 | 2,690 | 26.6 | 24 | 2 | vaginal | 6 | 3 | 37.2 | 154 | 53 | 97 | Normal_BW | Late_preterm | healthy |
2 | M | 37.1 | 2,830 | 38.5 | 15 | 0 | vaginal | 4 | 4 | 36.9 | 93 | 44 | 83 | Normal_BW | Term | birth_asphyxia |
3 | F | 37.1 | 2,600 | 36.3 | 27 | 1 | vaginal | 10 | 8 | 36.8 | 120 | 51 | 94 | Normal_BW | Term | healthy |
4 | F | 39.3 | 3,280 | 34.9 | 31 | 4 | vaginal | 9 | 7 | 37.1 | 141 | 54 | 95 | Normal_BW | Term | healthy |
5 | M | 38.5 | 3,540 | 42 | 27 | 0 | vaginal | 8 | 9 | 37 | 157 | 49 | 99 | Normal_BW | Term | healthy |
6 | F | 38.5 | 2,990 | 27.2 | 14 | 0 | vaginal | 7 | 6 | 36.9 | 123 | 49 | 96 | Normal_BW | Term | healthy |
7 | F | 35.9 | 2,680 | 36.7 | 32 | 4 | vaginal | 7 | 7 | 37.2 | 123 | 40 | 96 | Normal_BW | Late_preterm | healthy |
8 | F | 39.2 | 2,970 | 27.7 | 22 | 1 | vaginal | 7 | 6 | 36.6 | 147 | 47 | 97 | Normal_BW | Term | healthy |
9 | M | 34.4 | 2,210 | 42 | 32 | 4 | caesarean | 8 | 6 | 35.2 | 143 | 51 | 98 | LBW | Late_preterm | healthy |
10 | M | 37.8 | 2,470 | 26.6 | 15 | 1 | vaginal | 8 | 7 | 36.5 | 154 | 45 | 95 | LBW | Term | healthy |
11 | M | 41.4 | 3,650 | 42 | 26 | 5 | vaginal | 10 | 10 | 36.8 | 128 | 46 | 98 | Normal_BW | Term | healthy |
12 | F | 39.7 | 2,980 | 32 | 26 | 6 | vaginal | 7 | 6 | 37 | 137 | 51 | 98 | Normal_BW | Term | healthy |
13 | F | 39.2 | 2,600 | 27.8 | 37 | 9 | vaginal | 7 | 8 | 36.1 | 145 | 50 | 96 | Normal_BW | Term | healthy |
14 | F | 40.8 | 2,420 | 34.4 | 26 | 0 | vaginal | 7 | 9 | 36.9 | 138 | 51 | 94 | LBW | Term | low_birth_weight_term |
15 | M | 34.5 | 2,440 | 20.8 | 18 | 1 | caesarean | 9 | 7 | 36.4 | 145 | 39 | 98 | LBW | Late_preterm | healthy |
16 | M | 38.8 | 3,420 | 25.5 | 30 | 2 | vaginal | 9 | 9 | 37.2 | 151 | 51 | 97 | Normal_BW | Term | healthy |
17 | F | 37.4 | 1,990 | 42 | 21 | 1 | vaginal | 9 | 8 | 36.3 | 146 | 38 | 97 | LBW | Term | low_birth_weight_term |
18 | M | 38.9 | 3,580 | 39.3 | 24 | 4 | vaginal | 8 | 9 | 36.7 | 151 | 46 | 95 | Normal_BW | Term | healthy |
19 | F | 38.1 | 3,460 | 42 | 31 | 5 | vaginal | 9 | 10 | 37 | 140 | 60 | 100 | Normal_BW | Term | healthy |
20 | F | 37.6 | 2,790 | 28.9 | 17 | 0 | vaginal | 7 | 6 | 35.2 | 120 | 51 | 95 | Normal_BW | Term | neonatal_sepsis |
21 | F | 36.3 | 2,890 | 39.7 | 27 | 3 | vaginal | 4 | 3 | 36.6 | 137 | 50 | 96 | Normal_BW | Late_preterm | healthy |
22 | M | 38.8 | 2,910 | 36.1 | 36 | 6 | vaginal | 9 | 9 | 37 | 132 | 43 | 98 | Normal_BW | Term | healthy |
23 | F | 37.8 | 2,400 | 26.5 | 33 | 6 | vaginal | 9 | 9 | 37.3 | 111 | 57 | 97 | LBW | Term | healthy |
24 | F | 36.9 | 2,370 | 42 | 21 | 2 | vaginal | 9 | 9 | 36.7 | 143 | 50 | 97 | LBW | Late_preterm | healthy |
25 | F | 37.4 | 3,380 | 19.5 | 28 | 4 | vaginal | 1 | 3 | 36.9 | 135 | 34 | 83 | Normal_BW | Term | birth_asphyxia |
26 | M | 39.9 | 3,160 | 35.9 | 29 | 2 | vaginal | 8 | 7 | 37.2 | 132 | 44 | 95 | Normal_BW | Term | healthy |
27 | M | 31.9 | 2,320 | 35.6 | 19 | 4 | vaginal | 3 | 4 | 36.5 | 157 | 68 | 94 | LBW | Very_preterm | preterm_complications |
28 | M | 37.5 | 2,750 | 37.1 | 19 | 0 | vaginal | 9 | 10 | 37.3 | 130 | 47 | 96 | Normal_BW | Term | healthy |
29 | M | 35.2 | 1,770 | 28.2 | 27 | 4 | vaginal | 7 | 6 | 38.8 | 145 | 54 | 100 | LBW | Late_preterm | neonatal_sepsis |
30 | F | 39.4 | 3,190 | 27.9 | 31 | 3 | vaginal | 8 | 10 | 36.9 | 154 | 41 | 95 | Normal_BW | Term | healthy |
31 | F | 41.9 | 2,810 | 38.4 | 27 | 4 | vaginal | 10 | 7 | 37 | 152 | 49 | 96 | Normal_BW | Term | healthy |
32 | F | 38.3 | 3,010 | 26.5 | 18 | 0 | vaginal | 8 | 10 | 37.2 | 140 | 44 | 97 | Normal_BW | Term | healthy |
33 | M | 40.1 | 3,620 | 42 | 23 | 1 | vaginal | 9 | 10 | 37.1 | 155 | 49 | 96 | Normal_BW | Term | healthy |
34 | M | 36.5 | 3,230 | 35.8 | 25 | 1 | vaginal | 6 | 7 | 37.2 | 132 | 45 | 96 | Normal_BW | Late_preterm | healthy |
35 | M | 39.2 | 3,730 | 33.1 | 26 | 1 | vaginal | 9 | 10 | 36.7 | 129 | 48 | 97 | Normal_BW | Term | healthy |
36 | M | 36.8 | 2,920 | 34.9 | 24 | 1 | caesarean | 9 | 9 | 37.3 | 151 | 57 | 100 | Normal_BW | Late_preterm | healthy |
37 | M | 40.6 | 3,320 | 40.6 | 26 | 2 | vaginal | 8 | 10 | 36.8 | 164 | 39 | 96 | Normal_BW | Term | healthy |
38 | M | 35.8 | 2,560 | 42 | 27 | 3 | vaginal | 7 | 8 | 37.2 | 141 | 60 | 95 | Normal_BW | Late_preterm | healthy |
39 | M | 37.3 | 3,040 | 38 | 20 | 2 | vaginal | 9 | 10 | 36.9 | 151 | 69 | 98 | Normal_BW | Term | healthy |
40 | F | 36.1 | 2,520 | 34.7 | 32 | 1 | vaginal | 7 | 9 | 36.7 | 137 | 45 | 96 | Normal_BW | Late_preterm | healthy |
41 | M | 36.9 | 2,780 | 42 | 23 | 2 | vaginal | 6 | 8 | 36.3 | 147 | 56 | 95 | Normal_BW | Late_preterm | healthy |
42 | F | 39.1 | 2,720 | 20 | 21 | 2 | vaginal | 7 | 7 | 36.8 | 163 | 53 | 98 | Normal_BW | Term | healthy |
43 | F | 39.6 | 3,420 | 29 | 21 | 0 | vaginal | 8 | 10 | 36.5 | 130 | 53 | 98 | Normal_BW | Term | healthy |
44 | M | 38.1 | 3,160 | 32.9 | 29 | 4 | vaginal | 9 | 9 | 36.6 | 137 | 43 | 96 | Normal_BW | Term | healthy |
45 | F | 37.6 | 2,560 | 24.8 | 31 | 3 | vaginal | 9 | 8 | 37.1 | 149 | 42 | 94 | Normal_BW | Term | healthy |
46 | F | 36.4 | 3,080 | 40.1 | 29 | 3 | vaginal | 6 | 8 | 36.4 | 148 | 45 | 98 | Normal_BW | Late_preterm | healthy |
47 | M | 41.2 | 3,540 | 35.7 | 34 | 5 | caesarean | 9 | 9 | 37.3 | 140 | 45 | 96 | Normal_BW | Term | healthy |
48 | M | 39.5 | 2,400 | 23.7 | 22 | 2 | vaginal | 8 | 9 | 36.8 | 140 | 41 | 97 | LBW | Term | low_birth_weight_term |
49 | F | 40.5 | 3,660 | 22.3 | 25 | 3 | vaginal | 9 | 10 | 37.3 | 137 | 43 | 97 | Normal_BW | Term | healthy |
50 | M | 35.5 | 2,340 | 32.8 | 21 | 0 | vaginal | 5 | 8 | 37 | 145 | 54 | 98 | LBW | Late_preterm | healthy |
51 | M | 32.5 | 1,900 | 37.2 | 24 | 1 | caesarean | 5 | 6 | 37.8 | 112 | 57 | 89 | LBW | Late_preterm | birth_asphyxia |
52 | M | 40.8 | 3,350 | 36.2 | 36 | 6 | vaginal | 8 | 9 | 36.8 | 119 | 43 | 99 | Normal_BW | Term | healthy |
53 | F | 40.2 | 3,590 | 42 | 18 | 0 | caesarean | 9 | 10 | 37 | 152 | 63 | 98 | Normal_BW | Term | healthy |
54 | F | 39.4 | 3,100 | 32.4 | 24 | 0 | vaginal | 8 | 10 | 36.8 | 125 | 39 | 94 | Normal_BW | Term | healthy |
55 | F | 36.5 | 2,360 | 27 | 39 | 5 | vaginal | 6 | 6 | 38.2 | 146 | 41 | 96 | LBW | Late_preterm | neonatal_sepsis |
56 | F | 37.4 | 2,240 | 42 | 28 | 2 | vaginal | 9 | 7 | 35.4 | 144 | 37 | 97 | LBW | Term | healthy |
57 | M | 37.7 | 2,590 | 37.4 | 19 | 1 | vaginal | 8 | 9 | 37 | 148 | 37 | 96 | Normal_BW | Term | healthy |
58 | F | 27.6 | 1,020 | 38.7 | 25 | 5 | vaginal | 2 | 4 | 35.3 | 143 | 76 | 97 | VLBW | Extremely_preterm | preterm_complications |
59 | M | 39.2 | 2,950 | 42 | 30 | 4 | vaginal | 9 | 10 | 36.8 | 150 | 45 | 98 | Normal_BW | Term | healthy |
60 | M | 24.7 | 790 | 19.5 | 27 | 2 | caesarean | 8 | 9 | 36.7 | 141 | 70 | 80 | ELBW | Extremely_preterm | respiratory_distress |
61 | F | 38.7 | 2,180 | 28.2 | 28 | 1 | vaginal | 7 | 7 | 37.2 | 143 | 46 | 96 | LBW | Term | low_birth_weight_term |
62 | M | 37.3 | 2,620 | 41.2 | 23 | 3 | vaginal | 9 | 10 | 36.9 | 153 | 40 | 97 | Normal_BW | Term | healthy |
63 | F | 37.5 | 2,550 | 25.3 | 25 | 0 | vaginal | 8 | 10 | 37.1 | 116 | 37 | 96 | Normal_BW | Term | healthy |
64 | F | 36.6 | 2,660 | 28.2 | 24 | 0 | vaginal | 6 | 6 | 36.8 | 139 | 34 | 98 | Normal_BW | Late_preterm | healthy |
65 | F | 40.8 | 2,430 | 18 | 32 | 8 | vaginal | 7 | 9 | 36.9 | 124 | 46 | 98 | LBW | Term | healthy |
66 | F | 39 | 3,450 | 32.7 | 23 | 0 | vaginal | 9 | 10 | 36.6 | 146 | 50 | 98 | Normal_BW | Term | healthy |
67 | F | 37 | 2,750 | 42 | 32 | 4 | vaginal | 8 | 10 | 36.8 | 132 | 42 | 98 | Normal_BW | Term | healthy |
68 | M | 36.4 | 2,750 | 33.7 | 16 | 0 | vaginal | 7 | 8 | 37.1 | 139 | 34 | 99 | Normal_BW | Late_preterm | healthy |
69 | M | 36.4 | 2,540 | 18 | 16 | 0 | vaginal | 2 | 8 | 37.4 | 150 | 53 | 98 | Normal_BW | Late_preterm | healthy |
70 | M | 40.4 | 3,670 | 42 | 24 | 2 | vaginal | 9 | 9 | 36.5 | 138 | 54 | 98 | Normal_BW | Term | healthy |
71 | M | 34.4 | 1,780 | 25.6 | 30 | 2 | vaginal | 6 | 6 | 36.9 | 95 | 39 | 88 | LBW | Late_preterm | birth_asphyxia |
72 | M | 36.9 | 3,230 | 34.2 | 28 | 2 | vaginal | 7 | 6 | 36.9 | 132 | 40 | 96 | Normal_BW | Late_preterm | healthy |
73 | F | 38.9 | 2,400 | 37.5 | 33 | 3 | vaginal | 8 | 8 | 36 | 148 | 52 | 100 | LBW | Term | healthy |
74 | M | 36.4 | 2,740 | 42 | 39 | 13 | vaginal | 8 | 8 | 36.8 | 152 | 35 | 98 | Normal_BW | Late_preterm | healthy |
75 | M | 36.6 | 2,770 | 38.2 | 18 | 0 | vaginal | 9 | 8 | 36.8 | 123 | 39 | 96 | Normal_BW | Late_preterm | healthy |
76 | M | 36.7 | 2,880 | 32.2 | 25 | 2 | vaginal | 9 | 7 | 37 | 153 | 56 | 97 | Normal_BW | Late_preterm | healthy |
77 | M | 39.1 | 3,750 | 34.7 | 14 | 0 | caesarean | 7 | 9 | 36.7 | 156 | 44 | 96 | Normal_BW | Term | healthy |
78 | F | 39.5 | 2,610 | 27.4 | 32 | 6 | vaginal | 10 | 8 | 36.7 | 128 | 47 | 98 | Normal_BW | Term | healthy |
79 | F | 39.8 | 2,630 | 24.1 | 20 | 0 | vaginal | 9 | 10 | 36.9 | 152 | 40 | 97 | Normal_BW | Term | healthy |
80 | F | 38.2 | 1,950 | 30.8 | 27 | 2 | vaginal | 5 | 9 | 36.6 | 144 | 54 | 95 | LBW | Term | low_birth_weight_term |
81 | F | 39 | 2,080 | 42 | 22 | 5 | vaginal | 4 | 9 | 36.9 | 151 | 47 | 98 | LBW | Term | low_birth_weight_term |
82 | M | 35.1 | 3,200 | 21.9 | 16 | 0 | vaginal | 5 | 7 | 37.3 | 168 | 53 | 96 | Normal_BW | Late_preterm | healthy |
83 | F | 38.6 | 3,350 | 20.8 | 14 | 0 | vaginal | 9 | 10 | 37 | 141 | 44 | 94 | Normal_BW | Term | healthy |
84 | M | 38.5 | 3,310 | 36.3 | 17 | 1 | vaginal | 8 | 9 | 36.6 | 148 | 40 | 98 | Normal_BW | Term | healthy |
85 | M | 41.3 | 3,790 | 42 | 19 | 0 | vaginal | 9 | 10 | 37.4 | 135 | 46 | 95 | Normal_BW | Term | healthy |
86 | M | 35 | 2,860 | 25.8 | 22 | 2 | caesarean | 7 | 5 | 36.7 | 136 | 42 | 96 | Normal_BW | Late_preterm | healthy |
87 | F | 39.6 | 2,920 | 24.6 | 20 | 0 | vaginal | 9 | 10 | 37.4 | 117 | 44 | 97 | Normal_BW | Term | healthy |
88 | M | 38.4 | 2,970 | 38.2 | 32 | 3 | caesarean | 9 | 9 | 36.9 | 143 | 42 | 97 | Normal_BW | Term | healthy |
89 | M | 37.8 | 2,260 | 35.9 | 32 | 7 | vaginal | 4 | 9 | 36.8 | 129 | 39 | 96 | LBW | Term | low_birth_weight_term |
90 | M | 36.3 | 2,530 | 42 | 34 | 3 | vaginal | 8 | 8 | 36.8 | 126 | 52 | 98 | Normal_BW | Late_preterm | healthy |
91 | M | 37.7 | 2,650 | 42 | 19 | 0 | vaginal | 8 | 8 | 37.3 | 140 | 50 | 98 | Normal_BW | Term | healthy |
92 | F | 35.7 | 2,570 | 19.7 | 15 | 0 | caesarean | 3 | 5 | 36.8 | 120 | 41 | 84 | Normal_BW | Late_preterm | birth_asphyxia |
93 | M | 38.8 | 3,310 | 24.4 | 34 | 9 | vaginal | 9 | 10 | 37.2 | 170 | 58 | 97 | Normal_BW | Term | healthy |
94 | M | 40.8 | 3,800 | 18 | 23 | 1 | vaginal | 9 | 10 | 36.9 | 132 | 42 | 97 | Normal_BW | Term | healthy |
95 | M | 38.3 | 2,620 | 36.4 | 26 | 1 | vaginal | 7 | 8 | 36.9 | 157 | 52 | 99 | Normal_BW | Term | healthy |
96 | F | 38.6 | 2,930 | 26.3 | 16 | 0 | vaginal | 7 | 10 | 37.1 | 139 | 31 | 97 | Normal_BW | Term | healthy |
97 | M | 37.4 | 2,900 | 42 | 32 | 5 | vaginal | 9 | 7 | 36.9 | 113 | 45 | 94 | Normal_BW | Term | healthy |
98 | M | 41.7 | 2,900 | 30.4 | 15 | 0 | vaginal | 9 | 10 | 36.7 | 150 | 36 | 100 | Normal_BW | Term | healthy |
99 | M | 35.7 | 2,740 | 29.6 | 21 | 0 | vaginal | 7 | 8 | 36.7 | 141 | 42 | 97 | Normal_BW | Late_preterm | healthy |
100 | F | 39.3 | 3,070 | 27.3 | 29 | 7 | vaginal | 9 | 9 | 37.1 | 132 | 49 | 97 | Normal_BW | Term | healthy |
⚠️ Synthetic dataset — Parameterized from published SSA literature, not real observations. Not suitable for empirical analysis or policy inference.
Synthetic Neonatal Birth Outcomes & Vital Signs Dataset (0–28 days)
Abstract
This dataset provides 30,000 synthetic neonatal records (10,000 per scenario) representing live-born neonates in low- and middle-income country (LMIC) facility settings. Each record contains 15 clinically relevant variables including gestational age, birth weight, Apgar scores, vital signs (temperature, heart rate, respiratory rate, SpO2), head circumference, maternal demographics, and delivery mode. All distributions are parameterized from peer-reviewed epidemiological literature, WHO/UNICEF normative standards, and Intergrowth-21st reference curves. Three burden scenarios (low, moderate, high) reflect the spectrum of neonatal health contexts encountered across LMICs. The dataset is designed for training and benchmarking machine learning models for neonatal risk prediction, triage, and outcome classification, where real patient data is unavailable or restricted.
1. Introduction
Neonatal mortality accounts for approximately 47% of all under-five deaths globally, with 98% occurring in low- and middle-income countries (Lawn et al., 2014). Birth asphyxia, prematurity-related complications, and neonatal sepsis are the three leading causes. Despite this burden, open-access tabular datasets of neonatal clinical data from LMIC contexts are virtually nonexistent due to ethical constraints, fragmented health information systems, and data governance barriers.
This synthetic dataset addresses this gap by generating realistic neonatal records grounded in published epidemiological evidence. It is intended for:
- Training ML models for neonatal outcome prediction
- Benchmarking classification algorithms on realistic class distributions
- Educational use in global health and clinical data science curricula
- Prototyping clinical decision support tools for neonatal care
This dataset is entirely synthetic. It must not be used for clinical decision-making.
2. Methodology
2.1 Target Population
Live-born neonates (0–28 days of life) presenting at facility-based delivery settings in LMICs. The age range corresponds to the WHO neonatal period definition.
2.2 Causal Structure (DAG)
Variables are sampled following a directed acyclic graph that respects clinical causal relationships:
Sex, Maternal Age (roots)
→ Parity (conditional on maternal age)
→ Primary Outcome (from scenario prevalence)
→ Gestational Age (conditional on outcome)
→ Birth Weight (conditional on GA, sex, outcome)
→ Head Circumference (conditional on GA, sex)
→ Apgar Scores (conditional on outcome, GA)
→ Vital Signs: Temp, HR, RR, SpO2 (conditional on outcome, GA, BW)
→ Delivery Mode (conditional on outcome, GA, scenario)
→ Classifications (derived from measurements)
2.3 Epidemiological Parameterization
All distributions are grounded in the following primary sources:
| Parameter | Value | Source |
|---|---|---|
| Global preterm rate | 10.6% (range 5–18%) | Blencowe et al., Lancet 2012 |
| LBW prevalence (LMIC) | 14.6% global; up to 28% South Asia | UNICEF/WHO Low Birthweight Estimates 2019 |
| Birth asphyxia incidence | 2–8% of facility births | WHO Basic Newborn Resuscitation 2012 |
| Neonatal sepsis incidence | 1–6% depending on setting | Lawn et al., Lancet 2014 |
| Birth weight by GA | Intergrowth-21st reference curves | Intergrowth-21st Consortium, Lancet 2014 |
| Sex ratio at birth | 1.05:1 (M:F) | WHO Global Health Observatory |
| C-section rates (LMIC) | 3–25% by setting | DHS Program, multiple countries |
Birth weight and head circumference are generated via cubic spline interpolation of Intergrowth-21st reference curves, with sex-specific offsets (~130g for BW, ~0.6cm for HC).
2.4 Scenario Design
| Scenario | Context | Healthy | Preterm Comp. | Asphyxia | Sepsis | RDS | LBW Term | C-section |
|---|---|---|---|---|---|---|---|---|
| Low burden | Urban LMIC facility | 88.6% | 3.1% | 2.5% | 1.5% | 0.7% | 3.5% | 19.1% |
| Moderate burden | District hospital | 80.0% | 4.8% | 4.2% | 3.5% | 2.6% | 4.9% | 13.7% |
| High burden | Under-resourced/conflict | 70.0% | 6.7% | 7.1% | 5.1% | 4.0% | 7.0% | 7.4% |
2.5 Generation Process
- Outcome-first sampling: Primary outcome assigned from scenario-specific prevalence to ensure epidemiological fidelity
- Conditional GA sampling: Gestational age sampled from outcome-specific truncated normal distributions (e.g., preterm complications → GA 24–36.9 weeks)
- Spline-interpolated anthropometry: Birth weight and head circumference derived from Intergrowth-21st reference curves via cubic spline interpolation, with sex offsets and outcome-specific z-score shifts
- Outcome-conditional vitals: Apgar scores (scaled Beta distribution), temperature, HR, RR, SpO2 all conditioned on primary outcome and GA
- Biological clipping: All values constrained to physiologically possible ranges
3. Dataset Description
3.1 Schema
| Column | Type | Units | Range | Description |
|---|---|---|---|---|
| id | int | — | 1–10000 | Unique identifier |
| sex | categorical | — | M, F | Biological sex at birth |
| gestational_age_weeks | float | weeks | 24.0–42.0 | Gestational age at delivery |
| birth_weight_g | int | grams | 400–5500 | Birth weight |
| head_circumference_cm | float | cm | 18.0–42.0 | Head circumference at birth |
| maternal_age_years | int | years | 14–48 | Maternal age at delivery |
| parity | int | — | 0–14 | Number of previous deliveries |
| delivery_mode | categorical | — | vaginal, caesarean | Mode of delivery |
| apgar_1min | int | score | 0–10 | Apgar score at 1 minute |
| apgar_5min | int | score | 0–10 | Apgar score at 5 minutes |
| temperature_c | float | °C | 32.0–41.0 | Axillary temperature |
| heart_rate_bpm | int | bpm | 80–200 | Heart rate |
| respiratory_rate_bpm | int | breaths/min | 15–110 | Respiratory rate |
| spo2_percent | int | % | 40–100 | Peripheral oxygen saturation |
| bw_category | categorical | — | ELBW, VLBW, LBW, Normal_BW, Macrosomia | WHO birth weight classification |
| preterm_category | categorical | — | Extremely_preterm, Very_preterm, Late_preterm, Term, Post_term | WHO preterm classification |
| primary_outcome | categorical | — | 6 classes | Primary neonatal outcome |
3.2 Classification Criteria
| Classification | Criteria | Source |
|---|---|---|
| ELBW | Birth weight < 1000g | WHO ICD-10 P07.0 |
| VLBW | Birth weight 1000–1499g | WHO ICD-10 P07.1 |
| LBW | Birth weight 1500–2499g | WHO/UNICEF |
| Extremely preterm | GA < 28 weeks | WHO Born Too Soon 2012 |
| Very preterm | GA 28–31 weeks | WHO Born Too Soon 2012 |
| Late preterm | GA 32–36 weeks | WHO Born Too Soon 2012 |
3.3 Primary Outcome Categories
| Outcome | Description |
|---|---|
| healthy | No significant complications |
| preterm_complications | Prematurity-related complications (apnoea, feeding difficulty, hypothermia) |
| birth_asphyxia | Perinatal asphyxia (failure to establish breathing at birth) |
| neonatal_sepsis | Early or late-onset neonatal sepsis |
| respiratory_distress | Respiratory distress syndrome or transient tachypnoea |
| low_birth_weight_term | Term neonate with low birth weight (small for gestational age) |
4. Validation
4.1 Prevalence Fidelity
Cross-scenario monotonicity confirmed: all adverse outcome rates increase from low → moderate → high burden, consistent with epidemiological expectations.
4.2 Key Statistics (Moderate Burden)
| Metric | Observed | Literature Target | Source |
|---|---|---|---|
| LBW rate | 25.9% | 15–28% | UNICEF/WHO 2019 |
| Preterm rate | 29.6% | 12–18% overall | Blencowe 2012 |
| Asphyxia rate | 4.2% | 3–5% | WHO 2012 |
| Sepsis rate | 3.5% | 2–4% | Lawn 2014 |
| C-section rate | 13.7% | 8–15% | DHS |
| Mean birth weight | 2845g | 2800–3100g | Intergrowth-21st |
| GA-BW correlation | ~0.85 | 0.80–0.90 | Clinical expectation |
4.3 Diagnostic Plots
5. Usage
5.1 Loading with HuggingFace datasets
from datasets import load_dataset
# Load default (moderate burden) configuration
dataset = load_dataset("electricsheepafrica/synthetic-neonatal-birth-outcomes-vitals-WHO-0-28days", "moderate_burden")
# Load high burden scenario
high = load_dataset("electricsheepafrica/synthetic-neonatal-birth-outcomes-vitals-WHO-0-28days", "high_burden")
# Access as pandas DataFrame
df = dataset["train"].to_pandas()
5.2 Loading directly from CSV
import pandas as pd
df = pd.read_csv("data/neonatal_moderate_burden.csv")
asphyxia = df[df['primary_outcome'] == 'birth_asphyxia']
print(f"Asphyxia prevalence: {len(asphyxia)/len(df)*100:.1f}%")
print(f"Mean Apgar-1 in asphyxia: {asphyxia['apgar_1min'].mean():.1f}")
5.3 Regenerating with custom parameters
pip install numpy pandas scipy matplotlib
python generate_dataset.py --all-scenarios --n 10000 --seed 42
python validate_dataset.py
6. Limitations & Ethical Considerations
- Synthetic data: All records are computationally generated. No real patient data was used. The dataset must NOT be used for clinical decision-making.
- Simplified causal structure: Real neonatal outcomes involve complex interactions (e.g., maternal HIV, malaria in pregnancy, intrapartum events) not fully modelled here.
- LMIC focus: Distributions are parameterized for LMIC facility contexts and may not represent high-income country populations.
- No longitudinal component: Each record is a snapshot at birth; neonatal trajectory (e.g., NICU course) is not modelled.
- Outcome exclusivity: Each neonate has one primary outcome; in reality, comorbidities are common (e.g., preterm + sepsis).
- Missing data: No missing values are simulated; real clinical datasets have substantial missingness.
7. References
- Blencowe H, et al. (2012). National, regional, and worldwide estimates of preterm birth rates in 2010. Lancet, 379(9832):2162-2172.
- UNICEF/WHO (2019). UNICEF-WHO Low Birthweight Estimates. New York.
- Lee AC, et al. (2013). National and regional estimates of term and preterm babies born small for gestational age. Lancet Global Health, 1(1):e26-36.
- WHO (2012). Guidelines on Basic Newborn Resuscitation. Geneva.
- Lawn JE, et al. (2014). Every Newborn: progress, priorities, and potential beyond survival. Lancet, 384(9938):189-205.
- Mwaniki MK, et al. (2012). Long-term neurodevelopmental outcomes after intrauterine and neonatal insults. Lancet, 379(9814):445-452.
- WHO (2022). Born Too Soon: Decade of Action on Preterm Birth. Geneva.
- Intergrowth-21st Consortium (2014). International standards for newborn weight, length, and head circumference. Lancet, 384(9946):857-868.
- Demographic and Health Surveys (DHS) Program. Maternal and newborn health indicators, multiple countries 2015-2023.
- WHO (2015). Pregnancy, Childbirth, Postpartum and Newborn Care: A guide for essential practice (3rd ed). Geneva.
Citation
If you use this dataset in your research, please cite:
@dataset{esa_neonatal_2025,
title={Synthetic Neonatal Birth Outcomes and Vital Signs Dataset (0-28 days)},
author={Electric Sheep Africa},
year={2025},
publisher={Hugging Face},
url={https://huggingface.co/datasets/electricsheepafrica/synthetic-neonatal-birth-outcomes-vitals-WHO-0-28days}
}
License
This dataset is released under the Creative Commons Attribution 4.0 International (CC-BY-4.0) license.
- Downloads last month
- 96