The rodent estrous cycle modulates a range of biological functions, from gene expression to behavior. The cycle is typically divided into four stages, each characterized by distinct hormone concentration profiles. Given the difficulty of repeatedly sampling plasma steroid hormones from rodents, the primary method for classifying estrous stage is by identifying vaginal epithelial cell types. However, manual classification of epithelial cell samples is time-intensive and variable, even amongst expert investigators. Here, we use a deep learning approach to achieve classification accuracy at expert level. Due to the heterogeneity and breadth of our input dataset, our deep learning approach (“EstrousNet”) is highly generalizable across rodent species, stains, and subjects. The EstrousNet algorithm exploits the temporal dimension of the hormonal cycle by fitting classifications to an archetypal cycle, highlighting possible misclassifications and flagging anestrus phases (e.g., pseudopregnancy). EstrousNet allows for rapid estrous cycle staging, improving the ability of investigators to consider endocrine state in their rodent studies.

The rodent estrous cycle modulates a range of biological functions, from gene expression to behavior. The cycle is typically divided into four stages, each characterized by distinct hormone concentration profiles. Given the difficulty of repeatedly sampling plasma steroid hormones from rodents, the primary method for classifying estrous stage is by identifying vaginal epithelial cell types. However, manual classification of epithelial cell samples is time-intensive and variable, even amongst expert investigators. Here, we use a deep learning approach to achieve classification accuracy at expert levels in a matter of seconds. Due to the heterogeneity and breadth of our input dataset, our deep learning approach (EstrousNet) is highly generalizable across rodent species, stains, and subjects. The EstrousNet algorithm exploits the temporal dimension of the hormonal cycle by fitting classifications to an archetypal estrous cycle, highlighting possible misclassifications and flagging anestrus phases (e.g., pseudopregnancy). EstrousNet allows for rapid estrous cycle staging, improving the ability of investigators to consider endocrine state in their rodent studies.

Despite known sex differences in brain function and incidence of neurological disorders, female subjects are routinely excluded from preclinical neuroscience research, particularly behavioral studies in rats. A common rationale for excluding females is that the hormone fluctuations of the estrous cycle will increase variability in experimental data. Accounting for the estrous cycle as an experimental variable requires expert knowledge of cycle tracking methods, which presents a barrier to widespread inclusion of female subjects. Conventional tracking relies on qualitative interpretation of vaginal cytology smears, and the subjective nature of this approach combined with a lack of reporting standards likely underlies the conflicting literature on estrous cycle effects on behavior. The estrous cycle is traditionally divided into stages based on cytology, but most stages do not directly reflect hormonal events and are therefore of limited relevance to neuroscience experiments. Here we present a simple, streamlined approach to estrous cycle monitoring in rats that eliminates subjective staging. Our method instead indexes the days of the estrous cycle to the one event that is unambiguously reflected in vaginal cytology: the pre-ovulatory surge in 17β-estradiol and subsequent epithelial cornification. With this tracking method, we demonstrate that cycle length is robustly regular across conditions. We quantified long-term memory in a Pavlovian fear conditioning experiment and uterine histology in a large cohort of rats, and found that grouping subjects by day was more sensitive in detecting cycle effects than grouping by traditional cytology staging. We present several datasets demonstrating the logic and applicability of our method, and show that, in the Track-by-Day framework, the cycle is highly regular and predictable in the vast majority of rats across a range of experimental conditions.

The exclusion of female subjects from preclinical neuroscience research has traditionally been justified in part by concerns about potential effects of cycling ovarian hormones on brain function. There is evidence that some behavioral and neurobiological measures do change over the estrous cycle and, as the use of female subjects becomes increasingly routine, there is a greater demand for accessible cycle‐tracking methods. Conventional estrous cycle staging requires expert training in the qualitative interpretation of vaginal cytology smears, which serves as a barrier for novice researchers. In addition, definitions and reporting practices are not standardized across laboratories, which makes it difficult to compare results across studies and likely contributes to a false perception of the cycle as ephemeral and inconsistent. Here, we describe a streamlined method for monitoring the estrous cycle in rats, which we term Track‐by‐Day. It is simple to implement and inherently produces consistent reporting. Our protocol should serve to demystify and facilitate adoption of cycle tracking for those new to the practice. © 2023 Wiley Periodicals LLC. Basic Protocol 1: Collection and staining of vaginal smears Basic Protocol 2: Track‐by‐Day classification of vaginal smears Support Protocol: Preparation of gelatin‐subbed slides