Overview
To classify subtypes of Polycystic Ovary Syndrome (PCOS) using machine-learning algorithms, and compare the reproductive and metabolic characteristics and IVF outcomes across these identified subtypes.
Description
In this study, we've developed a machine-learning model to classify PCOS patients into four subtypes based on nine clinical characteristics.
The goal of this observational study is to:
- Learn about different PCOS subtypes using our classification model.
- Compare the reproductive and metabolic features of these subtypes.
- Assess the outcomes of IVF among different PCOS subtypes.
- Prospective 6.5-year follow-up data will be collected.
Participants will:
- Undergo a telephone interview to gather details on:
- Current physical stats like height and weight.
- Reproductive history, including pregnancies and childbirths in recent years.
- Details about any IVF treatments.
- Current status of conditions such as PCOS, Type 2 Diabetes, hypertension, and dyslipidemia.
- Be invited for a physical examination that includes:
- Measurements such as height, weight, blood pressure, and body circumferences.
- Laboratory tests for endocrine and metabolic conditions.
- Ultrasound scans of the ovaries and liver.
Eligibility
Inclusion Criteria:
- PCOS patients diagnosed using the Rotterdam criteria, which requires the presence of
at least two of the following:
- Menstrual Irregularities: A menstrual cycle length of fewer than 21 days or more than 35 days, and/or fewer than eight cycles per year.
- Hyperandrogenism: Defined either by an elevated total testosterone level (as per local laboratory criteria) or by a modified Ferriman-Gallwey (mFG) score of 5 or higher.
- Polycystic Ovaries on Ultrasound: Presence of 12 or more follicles measuring 2-9 mm in diameter in each ovary and/or an ovarian volume exceeding 10 mL.
Exclusion Criteria:
Patients with congenital adrenal hyperplasias, androgen-secreting tumours, or Cushing's
syndrome) will be excluded.