%! Author = alex
%! Date = 3/6/25


\section{Introduction}\label{sec:introduction}

In textbooks, a menstrual cycle is 28 to 30 days in length with its ovulation happening around day 14~\cite{Phy}



The average age of pregnant women in developed countries has been increasing over the past few decades.
Combined with the sharp decline in conception rates after age 35,
this trend underscores the growing need for accurate understanding of the menstrual cycle~\cite{sauer_reproduction_2015}.
Predicting the fertile days in a women's menstrual cycle is not only relevant for family planning but also for
natural contraception and general health monitoring, as the corresponding hormone levels have a significant impact on
the overall health and well-being of a woman.

Textbook cycles usually have a length of 28 days with an ovulation around day 14.
This however, does not represent the real world variability of menstrual cycles.

Ovulation is the process in which an egg cell is released from the ovaries, making fertilization possible.
This process is regulated by hormonal changes, including fluctuations in luteinizing hormone (LH) and
follicle-stimulating hormone (FSH), and is accompanied by other physiological changes such as an increase in electrical resistance
and viscosity of the cervical mucus or an increase in body temperature~\cite{wallach_prediction_1980}.
These processes remain incompletely understood and are influenced by lifestyle factors such as stress, diet or exercise or
health-related factors such as Polycystic Ovary Syndrome (PCOS), making ovulation difficult to predict.

In theory, these physiological changes provide a basis for predicting ovulation and the surrounding fertile window.
In practice, however, many of these signals are difficult to measure continuously, as they require invasive procedures,
manual tracking, or costly equipment.
Body temperature, by contrast, is a non-invasive marker that can be measured continuously using intravaginal, wrist-worn, or ear-worn thermometers.
While some studies have deemed body temperature unusable for predictive analysis~\cite{bauman_basal_1981},
others suggest its predictive value for ovulation~\cite{sato_novel_2024, royston_identifying_1991,
    luo_detection_2020, alexander_fertilitatsmonitoring_2014, luz_improved_2024, yu_tracking_2022, pratikno_pdf_2024}.

However, most existing studies rely on small, idealized datasets that exclude cycles with irregular lengths or late ovulation.
While such restrictions simplify the prediction task and yield high accuracy, they give a misleading impression of real-world model performance.
It is thus not yet fully clear, whether temperature can reliably be used as a predictive marker for ovulation or fertility.

Additionally, other studies have focused on peripheral temperature measurements of skin or in-ear temperature,
which are subject to many sources of noise that can significantly affect the quality of the resulting predictions.
Intravaginal temperature reflects true core body temperature and offers higher resolution and stability,
as it is largely unaffected by external circumstances.
This allows for more reliable detection of subtle thermal shifts associated with ovulation, especially in irregular cycles.

Generalization is critical for reliable ovulation prediction,
especially given the high variability in cycle patterns~\cite{munster_length_1992, bull_real-world_2019}.
This is particularly relevant in high-stakes applications such as natural contraception or in-vitro fertilization (IVF),
where inaccurate predictions can have serious consequences.


This study aims to advance ovulation prediction by leveraging an extensive database of more than 40,000 menstrual cycles recorded using
an intravaginal wearable device that continuously measures core body temperature.
The objective is to develop a machine learning model that performs reliably across diverse cycle types, including irregular ones.
To this end, we compare a set of time series-based machine learning architectures and evaluate their performance for
natural family planning and contraception.
Finally, we demonstrate that high predictive accuracy on highly regular, curated datasets, as commonly reported in prior work,
does not reflect general applicability, since such datasets tend to favor even simple, rule-based approaches.


The research objectives are:
\begin{itemize}
    \item To assess the performance of different model architectures on the use cases of natural family planning and contraception, across both regular and irregular cycles.
    \item To compare sophisticated machine learning models with simple rule-based baseline approaches.
    \item To study the overall predictive quality of temperature for ovulation and fertility prediction.
\end{itemize}