Have a personal or library account? Click to login

Application of machine learning algorithm for predicting gestational diabetes mellitus in early pregnancy†

Open Access
|Sep 2021

Figures & Tables

Figure 1

Technical route.
Technical route.

Figure 2

Accuracy curve of the training set and the test set of the F1 model.
Accuracy curve of the training set and the test set of the F1 model.

Figure 3

Accuracy curve of the training set and the test set of the F2 model.
Accuracy curve of the training set and the test set of the F2 model.

Figure 4

Accuracy curve of the training set and the test set of the F3 model.
Accuracy curve of the training set and the test set of the F3 model.

Figure 5

Weighted ranking in the F1 model (top 20).
Weighted ranking in the F1 model (top 20).

Figure 6

Weighted ranking in the F2 model (top 20).
Weighted ranking in the F2 model (top 20).

Figure 7

Weighted ranking in the F3 model (top 20).
Weighted ranking in the F3 model (top 20).

Definition of input data variables for model training_

DimensionVariablesInput variables
1. Physical examination1.1 AgeAge
1.2 BMIBMI
1.3 Blood type-AA
1.4 Blood type-BB
1.5 Blood type-ABAB
1.6 Blood type-OO
1.7 Systolic blood pressureSBP
1.8 Diastolic blood pressureDBP
2. Past history2.1 HypertensionHypertension
2.2 Heart diseaseHeart disease
2.3 Thyroid diseaseThyroid disease
2.4 Gynecological diseasesGynecological diseases
2.5 Kidney diseaseKidney disease
2.6 Congenital spina bifidaCongenital spina bifida
2.7 Benign tumorBenign tumor
2.8 Past surgical historyPast surgical history
2.9 HPVHPV
2.10 ColpomycosisColpomycosis
2.11 PCOSPCOS
2.12 Negative reproductive historyNegative reproductive history
2.13 History of multiple pregnanciesHistory of multiple pregnancies
2.14 Regular menstruationRegular menstruation
3. Personal history3.1 Body weight at birthBody weight at birth
3.2 Mother's weightMother's weight
3.3 Number of births by motherNumber of births by mother
3.4 SmokingSmoking
4. Family history4.1 Family history of diabetesFamily history of diabetes
5. Specialist examination5.1 Uterine heightUterine height
5.2 Abdominal circumferenceAbdominal circumference
5.3 Weight gainWeight gain
5.4 Gravidity historyGravidity history
5.5 Parity historyParity history
5.6 Multiple pregnanciesMultiple pregnancies
5.7 Fetus numberFetus number
6. Laboratory indicators6.1 Fasting glucoseFasting glucose
6.2 HepatitisHepatitis
6.3 HIVHIV
6.4 SyphilisSyphilis
7. Food frequency questionnaire (FFQ)7.1 Porridge intakePorridge intake (kg/month)
7.2 Flour foods intakeFlour foods intake (kg/month)
7.3 Sweet food IntakeSweet food Intake (kg/month)
7.4 Fried food IntakeFried food Intake (kg/month)
7.5 Filling food IntakeFilling food Intake (kg/month)
7.6 Coarse grain intakeCoarse grain intake (kg/month)
7.7 Potato intakePotato intake (kg/month)
7.8 Milk intakeMilk intake (kg/month)
7.9 Egg intakeEgg intake (kg/month)
7.10 Red meat intakeRed meat intake (kg/month)
7.11 Poultry intakePoultry intake (kg/month)
7.12 Processed meat intakeProcessed meat intake (kg/month)
7.13 Freshwater fishes intakeFreshwater fishes Intake (kg/month)
7.14 Seafood intakeSeafood intake (kg/month)
7.15 Bean products intakeBean products intake (kg/month)
7.16 Nuts intakeNuts intake (kg/month)
7.17 Dark vegetables intakeDark vegetables intake (kg/month)
7.18 Light vegetables intakeLight vegetables intake (kg/month)
7.19 Mushrooms intakeMushrooms intake (kg/month)
7.20 Fruit intakeFruit intake (kg/month)
7.21 Sweet drinks intakeSweet drinks intake (kg/month)
7.22 Alcohol intakeAlcohol intake (l/month)
8. International Physical Activity Questionnaire (IPAQ)8.1 Heavy physical activityHeavy physical activity (min/week)
8.2 Moderate physical activityModerate physical activity (min/week)
8.3 Light physical activityLight physical activity (min/week)
8.4 Walking timeWalking time (min/week)
8.5 Sedentary timeSedentary time (min/day)
8.6 Sleep duration on weekdaysSleep duration on weekdays (hour/day)
8.7 Sleep duration in rest daysSleep duration in rest days (hour/day)

Confusion matrix from prediction results of GDM incidence in the F1 model_

Actual valueTotalPredicted valuePrediction accuracy (%)

01
07357350100
182562631.70
In all81793.15

Confusion matrix from prediction results of GDM incidence in the F2 model_

Actual valueTotalPredicted valuePrediction accuracy (%)

01
07356954094.56
14679637179.44
In all1202 88.69

Results of prediction models using random forest_

Data setsAccuracy for training setAccuracy for test setAUC
F10.9570.9310.66
F20.9110.8870.87
F30.9300.9160.58

Confusion matrix from prediction results of GDM incidence in the F3 model_

Actual valueTotalPredicted valuePrediction accuracy (%)

01
07357350100
182691315.85
In all81791.55
DOI: https://doi.org/10.2478/fon-2021-0022 | Journal eISSN: 2544-8994 | Journal ISSN: 2097-5368
Language: English
Page range: 209 - 221
Submitted on: Nov 24, 2020
Accepted on: Jan 11, 2021
Published on: Sep 21, 2021
Published by: Shanxi Medical Periodical Press
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year

© 2021 Li-Li Wei, Yue-Shuai Pan, Yan Zhang, Kai Chen, Hao-Yu Wang, Jing-Yuan Wang, published by Shanxi Medical Periodical Press
This work is licensed under the Creative Commons Attribution 4.0 License.