An artificial intelligence approach for investigating multifactorial pain-related features of endometriosis

Kiser AC, Schliep KC, Hernandez EJ, Peterson CM, Yandell M, Eilbeck K

doi:10.1371/journal.pone.0297998

PloS one, 19(2), e0297998, 2024

An artificial intelligence approach for investigating multifactorial pain-related features of endometriosis

Karen C Schliep , Edgar Javier Hernandez , C Matthew Peterson

Author affiliations

University of Utah ROR

DOI10.1371/journal.pone.0297998 PMID38381710

Read at publisher

Abstract

Endometriosis is a debilitating, chronic disease that is estimated to affect 11% of reproductive-age women. Diagnosis of endometriosis is difficult with diagnostic delays of up to 12 years reported. These delays can negatively impact health and quality of life. Vague, nonspecific symptoms, like pain, with multiple differential diagnoses contribute to the difficulty of diagnosis. By investigating previously imprecise symptoms of pain, we sought to clarify distinct pain symptoms indicative of endometriosis, using an artificial intelligence-based approach. We used data from 473 women undergoing laparoscopy or laparotomy for a variety of surgical indications. Multiple anatomical pain locations were clustered based on the associations across samples to increase the power in the probability calculations. A Bayesian network was developed using pain-related features, subfertility, and diagnoses. Univariable and multivariable analyses were performed by querying the network for the relative risk of a postoperative diagnosis, given the presence of different symptoms. Performance and sensitivity analyses demonstrated the advantages of Bayesian network analysis over traditional statistical techniques. Clustering grouped the 155 anatomical sites of pain into 15 pain locations. After pruning, the final Bayesian network included 18 nodes. The presence of any pain-related feature increased the relative risk of endometriosis (p-value < 0.001). The constellation of chronic pelvic pain, subfertility, and dyspareunia resulted in the greatest increase in the relative risk of endometriosis. The performance and sensitivity analyses demonstrated that the Bayesian network could identify and analyze more significant associations with endometriosis than traditional statistical techniques. Pelvic pain, frequently associated with endometriosis, is a common and vague symptom. Our Bayesian network for the study of pain-related features of endometriosis revealed specific pain locations and pain types that potentially forecast the diagnosis of endometriosis.

Topics

Endometriosis > Symptoms > Pain Characterization Diagnostics > Artificial Intelligence > Pain Phenotyping Research Methodology > Machine Learning > Clinical Applications

artificial intelligence endometriosis pain multifactorial features, machine learning endometriosis pain prediction model, Kiser Schliep Eilbeck AI endometriosis pain analysis, endometriosis pain phenotyping computational approach, AI clinical features endometriosis pain characterization, PLoS One AI endometriosis pain 2024, supervised learning endometriosis symptom classification, multifactorial pain analysis endometriosis machine learning, endometriosis pain biomarkers computational investigation

PMID 38381710 38381710 DOI 10.1371/journal.pone.0297998 10.1371/journal.pone.0297998

Cite this article

Kiser, A. C., Schliep, K. C., Hernandez, E. J., Peterson, C. M., Yandell, M., & Eilbeck, K. (2024). An artificial intelligence approach for investigating multifactorial pain-related features of endometriosis. PloS one, 19(2), e0297998. https://doi.org/10.1371/journal.pone.0297998

Menstrual CycleSelf-Worth and MoodBasal Body TemperatureProspective Ovulation Cohort Open Access

Feeling of self-worth in healthy premenopausal women-relationships with menstrual cycles and ovulation over 1-year in the prospective ovulation cohort

Shirazian N et al., 2025PLoS One

Self-Worth is an over-arching evaluation of a person's sense of individual value. Self-worth, however, is an underappreciated concept. It has rarely been reported related to physiological data; we found no studies related to the menstrual cycle or ovulation. It is also unclear if Self-Worth is a stable trait or a variable state. We sought to discover if "Feeling of Self-Worth" (as recorded daily) was related to cycle phases and ovulation of spontaneous menstrual cycles (without hormonal contraception) in healthy premenopausal women over one-year in data from the Prospective Ovulation Cohort. Participating women were invited to complete the Menstrual Cycle Diary© (Diary©) daily; it describes cycle-related and other comprehensive everyday life experiences including negative moods as well as changes from each person's usual Feelings of Self-Worth. Women recorded their Feeling of Self-Worth on a 5-level ordinal scale as a letter translated to a number originally centered on their usual feeling (U = 3) with two higher and two lower (letters) levels. The validated Quantitative Basal Temperature© (QBT©) method was used to assess ovulation and luteal lengths. Fifty-three healthy, community dwelling, normal-weight, non-smoking premenopausal women mean age 33.9 (95% CI 32.4, 35.5) years old were studied. All participants were first required to document two consecutive normal-length (21-36 days) and normally ovulatory (luteal length ≥10 days) cycles before enrolment. Each participant completed the Diary© and QBT© records daily over a mean of 13 cycles (minimum = 8). During the study, cycle lengths were mean 28.04 (95% CI 27.80, 28.28) days with 71% of all cycles being normally ovulatory, 26% having short luteal phases (SLP, LL < 10 days) and 2.6% being anovulatory. Results in all women and all cycles showed that the mean annual Feeling of Self-Worth was 3.01 (95% CI 2.94, 3.09), thus very tightly related to their usual Feeling of Self-Worth. There were only small, inconsistent differences between Feeling of Self-Worth in the follicular versus luteal phases comparing normally ovulatory versus all ovulatory cycles (including those with short luteal phases). Analysis of Self-Worth within the 46 women having both normally ovulatory and ovulatory disturbed cycles (short luteal and anovulatory) showed that it was slightly lower in these women's normally ovulatory cycles (P = .03). Principal Components Analysis of all Diary© data showed that Feeling of Self-Worth was positively related to Interest in Sex and Feeling of Energy (together explaining 9% of all variance). In addition, Feeling of Self-Worth had a significant, negative loading on the Negative Mood Factor (that explained 14.2% of total variance). These data suggest that Feelings of Self-Worth in this comprehensive menstrual cycle and ovulation dataset in healthy women were not related menstrual phases and ovarian hormone levels.

Fertility AwarenessCreighton Model SystemProspective CohortIntentions and Compliance Open Access

Pregnancies, intentions, and fertility behaviors during use of the Creighton Model FertilityCare System after initial intention to avoid pregnancy: Results from the Creighton Model effectiveness, intentions, behaviors assessment study

Stanford JB et al., 2025PLoS One

Knowledge of the fertile and infertile phases of the menstrual cycle can be applied to conceive or to avoid pregnancy. Fertility intentions and sexual behaviors during the fertile time may influence whether and when pregnancy occurs. The Creighton Model FertilityCare System (CrMS) is a specific system of fertility appreciation used to conceive or to avoid pregnancy. The objective of this paper is to report intentions, behaviors, and pregnancy rates during use of the CrMS among couples who initially intended to avoid pregnancy.Data and methodsWe analyzed a prospective cohort study conducted in 17 CrMS centers across the USA and Canada, following 296 couples for up to one year after onset of initial use of the CrMS to avoid pregnancy. Baseline data included demographics, motivations, and pregnancy intentions for each partner. Couples contributed 2894 menstrual cycles, most of which had data collected (by questionnaires and daily diary) on cycle-specific pregnancy intentions, days of potential fertility, and fertility behaviors. Pregnancies were prospectively actively ascertained. We found a high concordance (91%) in cycle pregnancy intentions between partners. However, 44% of cycles with strong intentions to avoid pregnancy included intercourse on potentially fertile days or days of undetermined fertility status. Across all sensitivity scenarios, cumulative 13-cycle pregnancy rates with cycle intention to conceive ranged from 88.0% to 89.8%, and cumulative 13-cycle pregnancy rates with cycle intention to avoid ranged from 29.1% to 35.3%. In multivariate analysis, baseline motivations and intentions for pregnancy within 2 years were strongly correlated with the likelihood of pregnancy, more so than cycle intentions. The findings suggest that in some populations using natural family planning, baseline motivations and intentions may be more strongly related to pregnancy rates than cycle intentions. Our findings also highlight essential elements for evaluating correct use, including complete recording of intercourse and its timing.

Menstrual CycleCardiac Electrophysiology EffectsCardiovascularMenstrual Cycle Variations Open Access

Ovulatory and anovulatory cycle phase influences on QT interval dynamics during the menstrual cycle

Naderi B et al., 2025PLoS One

Ovarian hormones affect cardiovascular health yet few sufficient-sized studies with reliable ovulatory documentation have assessed the QTc-hormonal relationship. This study investigated QTc changes across ovulatory and anovulatory menstrual cycle phases. This prospective cohort investigation, a cardiac sub-study of the Menstruation and Ovulation Study 2 (MOS2), involved 62 healthy, regularly menstruating community-dwelling women during spontaneous menstrual cycles. Electrocardiographic recordings were obtained within-woman mid-follicular for all, and luteal (ovulatory) or premenstrual (anovulatory), documented by the validated Quantitative Basal Temperature© method. Fridericia's formula rate-corrected the QT interval (QTc). A subsequent meta-analysis was conducted, pooling data from three additional studies to evaluate ovulatory follicular-luteal phase QTc changes. In the 26 ovulatory cycles, QTc minimally decreased from the mid-follicular to the luteal phases (383.0 ± 12.8 vs 382.6 ± 12.8 msec, P = .859). QTc in the 36 anovulatory cycles tended to increase from mid-follicular to premenstrual phases (381.7 ± 13.1 vs 385.0 ± 16.1 msec, P = .166). The meta-analysis in ovulatory cycles yielded a random-effects weighted mean QTc shortening of 1.67 msec (P = .53) in the luteal vs the follicular phase, aligning with our cohort data. In confirmed ovulatory cycles, QTc changes were minimal, showing no meaningful luteal phase QTc shortening. QTc changes in anovulatory cycles were also insignificant, with a small QTc prolongation likely due to longer estradiol exposure not counterbalanced by progesterone. Under normal physiological conditions, QTc changes during the menstrual cycle are trivial, and menstrual status does not need to be considered when interpreting the QT interval.

Menstrual CycleBreast ChangesPhysical SymptomsCyclical Mastalgia Open Access

Breast tenderness and swelling experiences related to menstrual cycles and ovulation in healthy premenopausal women: Secondary analysis of the 1-year "Prospective Ovulation Cohort"

Wood M et al., 2025PLoS One

Breast tenderness and swelling are associated with premenstrual symptoms but are not well described in healthy women. In this 1-year prospective observational study, we examined daily breast tenderness and swelling to determine whether differences existed between normally ovulatory and ovulatory disturbed (short luteal phase and anovulatory) cycles in a cohort of community dwelling, non-smoking, healthy premenopausal women. Enrolment required two consecutive normal-length and normally ovulatory cycles by Quantitative Basal Temperature© analysis. Women (n = 53) ages 20-41 recorded their daily breast experiences in the Menstrual Cycle Diary© across an average of 13.6 cycles. In all 720 cycles, the median breast tenderness was 1.4 (on a 0-4 scale, range 0.0-3.0), in cycles with a mean length of 28.1 days (95% CI 27.5-28.8). Comparison of breast tenderness and breast size (changes from usual) parameters between all normally ovulatory cycles and all ovulatory disturbed cycles in the whole cohort showed significantly higher levels in normally ovulatory (luteal length ≥10 days) in both Breast Tenderness Score [intensity X duration in days; 6.0 (range 1.0-14.0) vs. 3.0 (0.0-11.0) (P=.005)] and breast size [4.0 (2.0-4.0) vs. 4.0 (0.0-4.0) (P=.034]). However, within-woman in the forty-seven women with both normally ovulatory and ovulatory disturbed cycles, breast tenderness (intensity, duration, and Breast Tenderness Score), did not differ between normally ovulatory cycles and cycles with ovulatory disturbances. This study also demonstrated that in all ovulatory cycles, the timing of breast tenderness increased in parallel with breast swelling; the maximum for both was in the late luteal phase.

An artificial intelligence approach for investigating multifactorial pain-related features of endometriosis

Abstract

Topics

Cite this article

Related articles

Feeling of self-worth in healthy premenopausal women-relationships with menstrual cycles and ovulation over 1-year in the prospective ovulation cohort

Pregnancies, intentions, and fertility behaviors during use of the Creighton Model FertilityCare System after initial intention to avoid pregnancy: Results from the Creighton Model effectiveness, intentions, behaviors assessment study

Ovulatory and anovulatory cycle phase influences on QT interval dynamics during the menstrual cycle

Breast tenderness and swelling experiences related to menstrual cycles and ovulation in healthy premenopausal women: Secondary analysis of the 1-year "Prospective Ovulation Cohort"