Detection of ovulation, a review of currently available methods

Abstract The ability to identify the precise time of ovulation is important for women who want to plan conception or practice contraception. Here, we review the current literature on various methods for detecting ovulation including a review of point‐of‐care device technology. We incorporate an examination of methods to detect ovulation that have been developed and practiced for decades and analyze the indications and limitations of each—transvaginal ultrasonography, urinary luteinizing hormone detection, serum progesterone and urinary pregnanediol 3‐glucuronide detection, urinary follicular stimulating hormone detection, basal body temperature monitoring, and cervical mucus and salivary ferning analysis. Some point‐of‐care ovulation detection devices have been developed and commercialized based on these methods, however previous research was limited by small sample size and an inconsistent standard reference to true ovulation.

detecting ovulation, is vital for encouraging or discouraging contraception. For physicians or women who wish to know if a menstrual cycle is normal or to evaluate ovarian function, a test that retrospectively confirms ovulation should suffice, but for artificial reproductive techniques, the time of ovulation and the fertility window must be defined clearly.
By performing ultrasonography, the maximum growth of the dominant follicle and the subsequent decrease in size can be observed, so the time of ovulation, which lies in between, can be determined.
Because this time can be clearly defined in this manner, it is recognized as the standard reference examination for ovulation detection and is used mainly in artificial reproductive techniques. Detection of the luteinizing hormone (LH) surge, whether in serum or in urine, is very sensitive and specific for ovulation and provides great accuracy for determining conception capacity. However, because sperm ejaculated before a woman's LH surge may survive long enough to fertilize the ovum, methods that simply determine this surge are not ideal for contraception. Before the LH surge, serum estrogen level rises and several changes occur in body fluid components, including cervical mucus and saliva. Observation of these differences may provide a better view of the fertility window. Progesterone is secreted by the corpus luteum only after ovulation. Detection of progesterone or its metabolites can retrospectively confirm the occurrence of ovulation. Because progesterone causes a rise in basal body temperature (BBT), a measure of this temperature may also be useful for determining ovulation. Because the oocyte dies shortly after ovulation, methods that correlate to progesterone and its effect identify fertility window closure.

| Ultrasonography
Transvaginal ultrasonography can clearly define ovulation time and is recognized as the standard reference examination for detecting ovulation. It is performed by experienced technicians, radiologists, or gynecologists. Being invasive, expensive, and inconvenient, this technique is not broadly used; it is used mainly in gynecological clinics and is often performed as a step in artificial reproductive techniques. Using serial ultrasonography examinations, the time of ovulation can be determined as the point between maximum follicular diameter and follicular collapse ( Figure 1). Indications of ovulation include the following: 1. Disappearance or sudden decrease in follicle size.
3. Free fluid in pelvis (or pouch of Douglas). 4. Replacement of "triple-line appearance" of endometrium by homogenous, hyperechoic "luteinized" endometrium. 9,10 2.2 | Urinary luteinizing hormone LH belongs to a group of glycoprotein hormones that share a common a subunit, and biological specificity for each hormone is conferred by their divergent b subunit. While serum estradiol (the most potent estrogen secreted by the dominant follicle) concentration reaches a threshold level (greater than 200 pg/ml for approximately 50 hr), a positive feedback mechanism works on the hypothalamus and anterior pituitary gland, which results in an abrupt secretion of LH into bloodstream. The onset of the LH surge precedes ovulation by 35-44 hr, and the peak serum level of LH precedes ovulation by 10-12 hr. 11,12 A study of 155 cycles from 35 women demonstrated that the onset of the LH surge primarily occurs between midnight and early morning (37% between 00:00 and 04:00, 48% between 04:00 and 08:00). 13 Detection of LH in urine using an over-the-counter device is much more convenient and less invasive than measuring serum LH level by multiple venipuncture. Women wishing to accurately determine their fertility window are instructed to record several menstrual cycles to estimate the possible time of ovulation. Starting from the 10th to 11th day of a new cycle (day 1 is defined as the first day of menstrual bleeding), or 4 days before the estimated ovulation day, women can test their urinary LH once or twice daily. Highly sensitive urinary LH kits detect concentrations as low as 22 mIU/ml, while natural LH surge concentration in urine ranges from 20 to 100 mIU/ml. 14 The mean time interval after a positive urinary LH test to follicular rupture detected by sonography was reported to be 20 6 3 hr (95% CI [14][15][16][17][18][19][20][21][22][23][24][25][26], 15 and in a study focused on infertile women, sensitivity, specificity, and accuracy of the urinary LH test to detect ovulation reached 1.00, 0.25, and 0.97, respectively. 16  and, (c) plateau (13.9%). 23 Furthermore, two (4.3%) women demonstrated LH surge without ovulation. In infertile women, premature LH surge that did not trigger ovulation was detected in 46.8% of cycles. 24 Also, a situation reported as "luteinized unruptured follicle syndrome" was reported to occur in 10.7% of menstrual cycles in normally fertile women. Women with this syndrome have a normal LH surge, functioning corpus luteum, and menstruation, but no oocyte is released. 25,26 Because a positive urinary LH test precedes ovulation, it is theoretically helpful for timed intercourse or intrauterine insemination because the clinical pregnancy rate after a single incident of intercourse is highest from a point 2 days before ovulation to the day of ovulation. 27 In a SU ET AL. study mainly focused on the psychological stress in women using a urinary LH kit, a trend toward increased self-reported pregnancy rate can be seen, but this result did not reach statistical significance (the odds of pregnancy rate in the study group were 1.77 (95% CI: 0.9992, 3.1585) compared with the control group). 28 A 2015 Cochrane review concluded that timed intercourse using urinary hormone monitoring was associated with an increased pregnancy rate (RR 1.36, 95% CI 1.06-1.73, 3 RCTs, n 5 1,370). 29 A promising unpublished study that may provide a better answer to this question is the Oxford Conception Study, which targets conception rate as its primary outcome. 30

| Serum progesterone and urinary pregnanediol 3-glucuronide
After ovulation, the dominant follicle turns into a corpus luteum and begins to secrete progesterone. To confirm ovulation, serum progesterone or its metabolite in urine, can be measured. A single serum proges-terone level >3 ng/ml in mid-luteal phase has been used to retrospectively detect ovulation. A recent European study proposed a random serum progesterone 5 ng/ml for confirming ovulation with sensitivity and specificity at 89.6% and 98.4%, respectively. 31 The same study group also demonstrated that levels of urinary pregnanediol 3-glucuronide (PDG), a metabolite of progesterone, measured at levels over 5 lg/ml for three consecutive days could be used as positive confirmation of ovulation with a sensitivity of 92.2% and a specificity of 100%. 32 Nevertheless, a convenient POC device to detect urinary PDG has not yet been developed.

| Urinary follicular stimulating hormone
Li et al. reported that peak FSH level occurred within 1 day of ultrasonography-detected follicular collapse in 97% of menstrual cycles.
The threshold value of urinary FSH was not mentioned. 33 No further study has been published regarding this method to detect ovulation.
FIG URE 1 Transvaginal ultrasonography of ovary and endometrium in early follicular phase (a-c), late follicular (shortly before ovulation) phase (d-f), and post-ovulatory luteal phase (g-i). In early follicular phase, the endometrium (between the blue arrows) just after menstruation appeared thin and homogenous (a). Multiple small follicles (red arrows) could be seen inside the ovary (c). In late follicular phase, the endometrium was thickened, and a typical "triple line" appearance could be seen (d). A dominant follicle, approximately 2 cm in diameter, is about to ovulate in late follicular phase (f). After ovulation, the endometrium becomes "luteinized," with increased echogenicity (whitening) (g). The dominant follicle transforms into a corpus luteum (i). Note the increased echogenicity and inhomogenous content (with filament-like structure) inside the corpus luteum compared to the dominant follicle in picture f. Free fluid can be seen in the pelvis (white arrows in pictures g and f)

| Basal body temperature
In 1906, Theodoor Hendrik van de Velde noticed a biphasic change of BBT in women during menstruation. Monitoring of BBT has become one of the simplest and least invasive methods to detect ovulation.
The rise of BBT results from the thermogenic effect of progesterone.
During the follicular phase of the menstrual cycle, BBT keeps in the lower range, generally between 97.0 and 98.08F, until approximately 1 day before ovulation, when BBT reaches its lowest point (nadir, or dip).
After ovulation, the corpus luteum begins to secret progesterone. The BBT rises 0.5-1.08F and plateaus throughout the luteal phase. In late luteal phase, when the corpus luteum regresses and serum progesterone level decreases, the BBT returns to the lower range within 1-2 days before, or just at, the onset of menstrual bleeding. This biphasic pattern of BBT retrospectively suggests ovulation ( Figure 2).
Women interested in determining their fertility window are instructed to measure their oral, vaginal, or rectal temperature every day when they wake up and before any activity is initiated. Modern digital thermometers have recording abilities that make monitoring BBT more convenient. However, a traditional glass thermometer that is accurate to 1/10th of a degree is good enough for determining BBT.
Women can record their BBT on specialized charts, online services, or applications on smart-phone devices.
There are many ways to interpret BBT. In one approach, the "coverline" method, a horizontal coverline (which means a threshold temperature) is drawn on a BBT chart. When temperature above this coverline is recorded, ovulation is suggested. The coverline can be determined by adding 0.158F to the highest temperature recorded during the first 10 days of a cycle, or by using previously recorded temperatures. Using the "three over six" rule, 34 three temperatures are required to be 0.28F above the highest point of the previous six temperatures with at least one of the higher temperatures being 0.48F above the lower ones.
Factors influencing BBT include fever, alcohol, emotional or physical stress, sleep disturbance, change of room temperature, change of waking time, change of climate, and recent start or discontinuation of birth control pills or anti-pyretics. A "GAP" technique was described in 2005 as a superior approach to the coverline method for women seeking contraception. By subtracting the BBT of a cohabiting male partner from the female partner's BBT, the temperature "gap" may be recorded. This gap is theoretically not influenced by environmental factors other than hormonal influences of the female. However, the study group was small (33 cycles), and the approach is not suitable for women who have not, or do not, live with their male partners. 35 Interpretation of BBT is not always easy. Even for natural family planning experts, there has been poor agreement on the first day of BBT rise between different observers. 36  concluded that monitoring BBT was no longer a good predictor of ovulation and, therefore, should not be recommended for couples seeking pregnancy. 38 Nevertheless, BBT is still broadly used for contraception and for evaluation of ovulatory function, especially for couples who are reluctant or unable to pursue more formal and costly evaluation. An illustration of BBT chart in degrees Celsius. This is a biphasic pattern in a normal ovulatory cycle. Ovulation can only be suggested after observing the rise and plateau of temperature above the purple "Coverline." A drop in temperature occurs at the end of the luteal phase when progesterone decreases, which is followed by menstruation SU ET AL. | 241 menstrual cycle. Outside of the periovulatory period, this mucus is mainly composed of high-molecular-weight glycoprotein (mucin), which forms a mesh-like structure that provides a barrier to sperm and microorganisms. The mucus appears thick, scant, and viscous by inspection.

| Cervical mucus
During the periovulatory period, under the effect of estrogen, the production of acellular water increases and the production of mucin decreases. The mesh-like structure consequently loosens and becomes highly receptive for sperm penetration. 18,39,40 Women experience an increased amount of watery discharge resembling raw egg white during this period.
Observing cervical mucus is the least expensive method to detect ovulation. Women may simply observe mucus found externally at the vulva or collect vaginal mucus using their fingers. While uncomplicated, this approach has still been shown to be effective. In a small study involving 12 cycles recorded by 6 women, finding the most abundant fertile type of mucus correlated to 6 1 day of ovulation detected by ultrasonography. 41 In a larger study involving 148 cycles recorded by 40 women, mucus sensation and characteristics yielded a 48.3% correlation to ultrasonography-detected ovulation. 42 In another large study, cervical mucus-detected methodology for determining expected date of ovulation correlated to 6 1 day of ultrasonography-detected ovulation in 160/215 cycles (74.4%). 43 In a study involving 29 cycles recorded by 15 women, the sensitivity of detecting fertile mucus from the vulva and vagina compared to 6 1 day of ultrasonography-detected ovulation were 75.9% and 75.9%, respectively. When the period was extended to 21 to 12 of ovulation, the sensitivities were 96.9% (vulva) and 89.6% (vagina), respectively. 44 The combination of measuring BBT and observing cervical mucus for contraception is called the symptothermal method. Among women using this method for contraception in one study, the unintended pregnancy rate was 1.8%. 45

| Salivary ferning and analysis
The history of salivary ferning test can be traced back to 1969. Dr. Biel Cassal found that the arborization (or ferning) of saliva could be seen under a microscope during the periovulatory period. Increasing levels of estrogen and adrenocorticotropic hormone before ovulation stimulates the secretion of aldosterone, which regulates the electrolytes and fluid status in human body. 46 Crystallization of NaCl produces the ferning appearance of saliva observed under a microscope.
A pocket microscope specifically designed for salivary ferning observation has been developed. A study conducted by Barbato et al. declared that salivary ferning correlated well to BBT and cervical mucus methods. 46 It was later reported that salivary ferning can be detected in postmenopausal, pregnant, prepubertal women, and even in men. In women with regular menstruation, salivary ferning predicted ovulation with a sensitivity of only 53% correlated to ultrasonography and serum LH. 48,49 Guida et al. compared several methods to ultrasonographydetected ovulation and found that positive salivary ferning test correlated to 21 to 11 days of actual ovulation in 42% of cases. However, a high percentage of uninterpretable patterns (58.7%) was reported. 41 A recent study compared salivary ferning using a Geratherm ovu control microscope to a urinary LH test using EXACTO monitors and found a high correlation between these methods. 50 To perform this test, women could use a small microscope with built-in or removable slides. A drop of saliva may be applied to a slide, allowed to dry and inspected for ferning. The U.S. FDA announced that a positive test indicates that women may be near ovulation, but a negative test is unreliable for contraception.
Sex hormones were first measured in saliva by radioimmunoassay in 1978. Early publications focused on the correlation of salivary steroids to ovarian function, 19,20,[51][52][53][54]  Recently, attention was given to examining salivary proteins for various diagnostic purposes. Protein concentration is at its highest during the ovulation phase. By single dimension SDS-PAGE analysis, a 48 kDa protein was identified to exhibit predominantly during ovulatory phase. 57 The author suggested that the protein level and the presence of 48 kDa band might be recognized as ovulation indicators. Further research is needed on this topic.

| Urinary luteinizing hormone
Nowadays, many POC and over-the-counter ovulation detection devices are easily accessible for women planning conception or contraception. Most of these commercialized products determine ovulation by detecting urinary LH level. Several computerized fertility monitors have been used to detect urinary LH and estrone-3-glucoronide, which, if presented in urine, have also indicated ovulation. 58 The manufacturer of the Clearblue easy fertility Monitor (CEFM) reported a significantly higher pregnancy rate during the first 2 cycles in women using its product compared to control groups (22.7% compared to 14.4%). 59 (Figure 3a) Moreover, the same manufacturer developed an additional product, Clearblue DIGITAL Ovulation Test with Dual Hormone Indicator, which detects estrogen as well as LH. The manufacturer claims that this product can identify 4 or more fertile days with an accuracy rate of over 99%. Both monitors are digital immunoassays incorporating a disposable microspectrophotometer. 66 Primarily used for contraceptive purposes, the Persona monitor, achieved a 94% correct-use effectiveness. 67,68 (Figure 3a) The sensitivity of CEFM and Persona to accurately determine ovulation are 97% and 95.8%, respectively. 61

| Basal body temperature
In addition to urinary LH level detection, some commercialized POC devices adopt BBT monitoring methods to determine ovulation.
OvuSense is a vaginal temperature sensor with a 99% claimed accuracy rate for ovulation date detection and an 89% accuracy rate for ovulation date prediction. 69 In addition to measuring temperature intravaginally, some devices base their results on temperature measured elsewhere on the body. NaturalCycles, for example, which consists of thermometers and mobile application, is stated to misidentify only 0.05% of nonfertile days as fertile days. 70 By measuring and recording temperature on a daily basis, the NaturalCycles application can detect ovulation and fertile days. Another product, DuoFertility® employs a small sensor worn under the arm that takes thousands of measurements throughout the day and night and reports temperature data to a server (Figure 3b). DuoFertility automatically recognizes ovulation through proprietary algorithms with 100% sensitivity. 62  Another device, Geratherm® ovu control, is reported to have a specificity of 78% and a sensitivity of 80%. 71 Both of the abovementioned devices require patients to self-evaluate salivary ferning. The agree-ment rate between saliva test carried out by patients and by laboratory staff is 89.4% with Geratherm® ovu control. 71 To avoid potential selfevaluation error WU, Hui-Ching et al. introduced an automatic salivary ferning pattern recognition system for ovulation detection. They report an accuracy rate of 84% from 100 saliva samples. 64 60 The right bar chart is the Serum LH surge relative to CPFM peak fertility. The X-axis 22 and 2 represents >1 day before or after CPFM peak fertility, respectively. Cycles with no CPFM peak fertility (n 5 13), no serum LH surge day (n 5 10), or neither (n 5 1) are excluded from the table. 61 63 The accuracy of three proposed algorithms. Algorithm 1 is binarizing 1 dark pixel density. Algorithm 2 is binarizing 1 dark pixel density 1 thinning. Algorithm 3 is binarizing 1 Hough transform 1 thinning 1 decision tree. 64

| Paper-based approaches
Aside from smart-phone based devices, paper-based ovulation detection devices are full of potential. Paper has been successfully used for biological assays, such as ELISA 73 and cell assays. 74 Paper is economical, prevalent, disposable, and suitable for large-scale manufacture. 75 Paperbased ovulation detection devices based on LH detection, such as One Step® Ovulation & Pregnancy Test Kit Strips, are currently available.

| SUM M A R Y A N D PE R SP ECT IV E S
Most of the currently available methods to detect ovulation were developed decades ago (Table 1). Research tests for the methods described here were limited to small sample sizes and subjects of limited cultural background (most studies were performed in European countries in less than 100 women with less than 1,000 cycles). Even fewer studies examined or included women having irregular menstruation, which is characterized by less predictable ovulation timing. Such women would particularly benefit from improved ovulation predictors.