CoChat

Machine-Aided Self-diagnostic Prediction Models for Polycystic Ovary Syndrome: Observational Study

Open access
Zigarelli, Angela, Jia, Ziyang, Lee, Hyunsun · 2022 · JMIR Formative Research

Published version of a preprint

The DOI matches its Crossref record; 42 of the 42 references it cites were found in a registry or an index.

Abstract

Background: Artificial intelligence and digital health care have substantially advanced to improve and enhance medical diagnosis and treatment during the prolonged period of the COVID-19 global pandemic. In this study, we discuss the development of prediction models for the self-diagnosis of polycystic ovary syndrome (PCOS) using machine learning techniques. Objective: We aim to develop self-diagnostic prediction models for PCOS in potential patients and clinical providers. For potential patients, the prediction is based only on noninvasive measures such as anthropomorphic measures, symptoms, age, and other lifestyle factors so that the proposed prediction tool can be conveniently used without any laboratory or ultrasound test results. For clinical providers who can access patients’ medical test results, prediction models using all predictor variables can be adopted to help health providers diagnose patients with PCOS. We compare both prediction models using various error metrics. We call the former model the patient model and the latter, the provider model throughout this paper. Methods: In this retrospective study, a publicly available data set of 541 women’s health information collected from 10 different hospitals in Kerala, India, including PCOS status, was acquired and used for analysis. We adopted the CatBoost method for classification, K-fold cross-validation for estimating the performance of models, and SHAP (Shapley Additive Explanations) values to explain the importance of each variable. In our subgroup study, we used k-means clustering and Principal Component Analysis to split the data set into 2 distinct BMI subgroups and compared the prediction results as well as the feature importance between the 2 subgroups. Results: We achieved 81% to 82.5% prediction accuracy of PCOS status without any invasive measures in the patient models and achieved 87.5% to 90.1% prediction accuracy using both noninvasive and invasive predictor variables in the provider models. Among noninvasive measures, variables including acanthosis nigricans, acne, hirsutism, irregular menstrual cycle, length of menstrual cycle, weight gain, fast food consumption, and age were more important in the models. In medical test results, the numbers of follicles in the right and left ovaries and anti-Müllerian hormone were ranked highly in feature importance. We also reported more detailed results in a subgroup study. Conclusions: The proposed prediction models are ultimately expected to serve as a convenient digital platform with which users can acquire pre- or self-diagnosis and counsel for the risk of PCOS, with or without obtaining medical test results. It will enable women to conveniently access the platform at home without delay before they seek further medical care. Clinical providers can also use the proposed prediction tool to help diagnose PCOS in women.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.2196/29967

Cites

42 of 42 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Illustration of the subgroup study and the flowchart of data analysis. PCA: Principal Component Analysis; PCOS: polycystic ovary syndrome.
Figure 1 Illustration of the subgroup study and the flowchart of data analysis. PCA: Principal Component Analysis; PCOS: polycystic ovary syndrome.CC BY · PubMed Central
Figure 2 Anthropomorphic variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001.
Figure 2 Anthropomorphic variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001.CC BY · PubMed Central
Figure 3 Symptom variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001.
Figure 3 Symptom variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001.CC BY · PubMed Central
Figure 4 Test result variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001. AMH: anti-Müllerian hormone.
Figure 4 Test result variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001. AMH: anti-Müllerian hormone.CC BY · PubMed Central
Figure 5 Given variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001.
Figure 5 Given variable comparison between the polycystic ovary syndrome (PCOS) positive group and the PCOS-negative group with P<.001.CC BY · PubMed Central
Figure 6 Correlation between the anthropomorphic variables.
Figure 6 Correlation between the anthropomorphic variables.CC BY · PubMed Central
Figure 7 Principal Component Analysis. Biplot of PC1 and PC2 based on the anthropomorphic attributes.
Figure 7 Principal Component Analysis. Biplot of PC1 and PC2 based on the anthropomorphic attributes.CC BY · PubMed Central
Figure 8 The optimal number of principal components to be used based on the proportion of variance explained.
Figure 8 The optimal number of principal components to be used based on the proportion of variance explained.CC BY · PubMed Central
Figure 9 Silhouette method to determine the optimal number of clusters.
Figure 9 Silhouette method to determine the optimal number of clusters.CC BY · PubMed Central
Figure 10 Subgroups on the PC1-PC2 plane after applying k-means clustering using the first 5 principal components based on the anthropomorphic attributes.
Figure 10 Subgroups on the PC1-PC2 plane after applying k-means clustering using the first 5 principal components based on the anthropomorphic attributes.CC BY · PubMed Central
Figure 11 Anthropomorphic variable comparison between subgroup 1 and subgroup 2 with P<.001.
Figure 11 Anthropomorphic variable comparison between subgroup 1 and subgroup 2 with P<.001.CC BY · PubMed Central
Figure 12 Symptom variable comparison between subgroup 1 and subgroup 2 with P<.001. PCOS: polycystic ovary syndrome.
Figure 12 Symptom variable comparison between subgroup 1 and subgroup 2 with P<.001. PCOS: polycystic ovary syndrome.CC BY · PubMed Central
Figure 13 Feature importance for all variables and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 1 in the patient subgroup model including only noninvasive predictor variables.
Figure 13 Feature importance for all variables and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 1 in the patient subgroup model including only noninvasive predictor variables.CC BY · PubMed Central
Figure 14 Feature importance for all variables and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 2 in the patient subgroup model including only noninvasive predictor variables.
Figure 14 Feature importance for all variables and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 2 in the patient subgroup model including only noninvasive predictor variables.CC BY · PubMed Central
Figure 15 Feature importance for all variables (with the middle importance variables omitted) and the SHAP (Shapley Additive Explanations) values for the 12 most important features of the one to all model (without subgroups) in the patient model.
Figure 15 Feature importance for all variables (with the middle importance variables omitted) and the SHAP (Shapley Additive Explanations) values for the 12 most important features of the one to all model (without subgroups) in the patient model.CC BY · PubMed Central
Figure 16 Feature importance for all variables (with the middle importance variables omitted) and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 1 in the provider subgroup model including all predictor variables. AMH: anti-Müllerian hormone; PRG: progrestrone; PRL: prolactin.
Figure 16 Feature importance for all variables (with the middle importance variables omitted) and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 1 in the provider subgroup model including all predictor variables. AMH: anti-Müllerian hormone; PRG: progrestrone; PRL: prolactin.CC BY · PubMed Central
Figure 17 Feature importance for all variables (with the middle importance variables omitted) and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 2 in the provider subgroup model including all predictor variables. AMH: anti-Müllerian hormone; BP: blood pressure; FSH: follicle-stimulating hormone; LH: luteinizing hormone; PRL: prolactin.
Figure 17 Feature importance for all variables (with the middle importance variables omitted) and SHAP (Shapley Additive Explanations) values for the 12 most important features of subgroup 2 in the provider subgroup model including all predictor variables. AMH: anti-Müllerian hormone; BP: blood pressure; FSH: follicle-stimulating hormone; LH: luteinizing hormone; PRL: prolactin.CC BY · PubMed Central
Figure 18 Feature importance for all variables (with the middle importance variables omitted) and the SHAP (Shapley Additive Explanations) values for the 12 most important features of the one to all model (without subgroups) in the provider model. AMH: anti-Müllerian hormone; BP: blood pressure; FSH: follicle-stimulating hormone; LH: luteinizing hormone; PRL: prolactin.
Figure 18 Feature importance for all variables (with the middle importance variables omitted) and the SHAP (Shapley Additive Explanations) values for the 12 most important features of the one to all model (without subgroups) in the provider model. AMH: anti-Müllerian hormone; BP: blood pressure; FSH: follicle-stimulating hormone; LH: luteinizing hormone; PRL: prolactin.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 42 references

Research works printed but not linked: 1

  1. 10.5555/3295222.3295230 · doi:10.5555/3295222.3295230

Cited by

See all 75

Metadata

Authors
Zigarelli, A., Jia, Z., & Lee, H.
Affiliations
University of Massachusetts Amherst (US)
Corresponding author
Lee, Hyunsun
Venue
JMIR Formative Research
Published
2022-03-15
Volume
6
Issue
3
Pages
e29967
ISSN
2561-326X (ISSN-L)
OpenAlex
W4221092010
Cited by
74
References
42
Citations by year
2026: 22 · 2025: 22 · 2024: 16 · 2023: 11 · 2022: 3
FWCI
9.75
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • JMIR Formative Research (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Topics
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
  • Health Sciences › Medicine › Reproductive Medicine › Ovarian function and disorders
  • Physical Sciences › Computer Science › Information Systems › Data Mining and Machine Learning Applications
Keywords
polycystic ovary syndrome, prediction model, CatBoost, SHAP feature importance, anti-Müllerian hormone, ovarian follicle count, acanthosis nigricans, hirsutism, menstrual irregularity, digital health platforms, women's health, risk prediction, K-means clustering, principal component analysis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.