The embryonic phase of development is complete by the end of the 10th menstrual or gestational week (this corresponds to 12 weeks following the onset of the last normal menstrual period). During this critical period of development, a single fertilized cell undergoes dramatic transformation as the cell mass evolves into major organs and a recognizable human form.
Because of the complex sequence of events that occurs during this short time period, it is not unusual for complications to develop. Currently, transvaginal ultrasonography is the imaging examination of choice to evaluate the rapidly evolving intrauterine events that occur following implantation of the gestational sac and the development of a visible embryo. Although a variety of terms are used to describe early pregnancy failure, in the presence of clear-cut sonographic evidence that a nonliving embryo is present, the term embryonic demise should apply.
Ultrasonographic findings in embryonic demise are demonstrated in the images below.
During the first trimester of pregnancy, approximately 25% of women experience mild vaginal bleeding and/or cramping. Pelvic examination usually reveals a closed and normal-appearing cervix. This clinical presentation characterizes a threatened abortion. Analysis of women with these findings reveals that 50% of the pregnancies will fail and that the rest will have a normal outcome. If the cervix is dilated, the pregnancy will certainly fail, although based on clinical examination, it is not usually possible to determine whether there are retained products of conception.
Some women with embryonic demise will be asymptomatic, and in these patients the diagnosis may be suggested based on subnormal uterine growth, inability to auscultate fetal cardiac activity, or failure of the human choriogonadotropin (hCG) level to increase at the expected rate.
In most instances, the clinical pelvic examination cannot determine the cause for the patient's symptoms. Thus, the patient should be referred for a real-time pelvic ultrasonographic examination. If clinical dating suggests a gestational age (GA) of 8 weeks or older, some sonographers begin the ultrasonographic study using a transabdominal approach. This is because in a normal pregnancy, when using a transabdominal approach, cardiac activity should be visible by 8 weeks' GA. However, an increasing number of sonographers begin the ultrasonographic examination with a transvaginal approach. This is because a higher transducer frequency is used, which in a normal pregnancy can detect cardiac activity approximately 2 weeks earlier, or by 6 weeks' GA. Furthermore, in comparison to a transabdominal approach, vaginal transducers provide superior resolution with respect to examining the appearance and contents of the gestational sac as well as the ovaries and adnexa. [1, 21, 28, 29]
Limitations of techniques
Transabdominal probes are limited, because they typically use 3.5-5 MHz transducers, compared with the 5-10MHz transducers used in transvaginal probes. Even if a 5-MHz transducer were used for a transabdominal and transvaginal scan, the transabdominal images of an early intrauterine pregnancy (IUP) would be inferior to those obtained by the transvaginal probe. This is because the transvaginal probe is physically closer to the object being scanned, and the transvaginal ultrasonographic beam does not traverse the abdominal wall. This results in fewer near-field artifactual echoes. These comparative effects are most pronounced when scanning obese patients.
The transvaginal approach, however, can be limited by the presence of a large pelvic mass, which can interfere with visualization of the intrauterine contents. Most often, large or strategically placed calcified uterine fibroids cause this problem. Under these circumstances, an abdominal approach should be used in an effort to image the uterus and its contents.
Another limitation is if the ultrasonographic study is performed prior to the time a yolk sac can be detected. Using a vaginal approach, this structure should be observed by 5.5 weeks' GA. If a small, saclike structure is imaged but does not contain a yolk sac, it is often not possible to determine if the intrauterine finding is the result of an early IUP or a pseudosac associated with an ectopic pregnancy. In these instances, careful evaluation of the adnexa may aid in the detection of an ectopic pregnancy. Occasionally, serial ultrasonography and/or hCG determinations may be required to determine the etiology of the intrauterine sac.
A final, but important, admonition (that relates to all ultrasonographic examinations) is to recognize the technical adequacy of the study, to know the limitation(s) of the equipment, and, importantly, to determine the experience of the person who performs and interprets the examination.
If an embryo is identified on a transabdominal scan and cardiac activity is not visible, the prognosis is usually poor. It is important to realize that occasionally a very small embryo may be present in which cardiac activity cannot be confirmed.
According to the experience of one group of investigators who used a transabdominal approach,  21% of the time a normal IUP lacked visible embryonic heart motion when the embryonic crown rump length (CRL) was 9 mm or less. Based on their experience, these investigators recommended that when using a transabdominal approach, 9 mm should be considered the discriminatory embryonic length for detecting cardiac motion. Used in this manner, the discriminatory level denotes the numeric value when a certain finding should always be present.
Given its superior resolution, it is not surprising that vaginal ultrasonographic scans can detect cardiac activity with a smaller embryonic CRL.
One report recommended that when a transvaginal approach is used, 4 mm be considered the discriminatory embryonic length for detecting cardiac motion. According to the investigators, who reviewed the use of the transvaginal approach in embryonic ultrasonographic imaging, 18% of the time a normal IUP lacked visible embryonic heart motion when the embryonic CRL was 4 mm or less.  Other investigators suggested 5 mm as the discriminatory embryonic size for detecting cardiac motion. (See the image below.) [4, 5]
If an embryo exceeds the discriminatory length and cardiac activity is absent, a nonviable gestation should be diagnosed. Because this observation has such important clinical ramifications, this observation should be made by 2 independent observers, and interpretive caution must be exercised in any questionable case. Documentation should be available by M mode imaging and/or by obtaining a videotape or video clip.
If the length of the embryo is less than the discriminatory value, the patient should be managed expectantly, and a repeat ultrasonographic examination should be performed when the expected embryonic CRL exceeds the discriminatory value. Alternatively, or additionally, the level of serum hCG may be useful for determining whether a normal IUP is present.
Visualizing a living embryo
Although seemingly a paradox, it is well known that detecting cardiac activity when using a vaginal transducer does not guarantee as favorable an outcome as detecting cardiac activity when using an abdominal transducer. With a transvaginal approach, mortality rates of 20-30% have been reported in women with threatened abortion in whom embryonic cardiac activity is documented at 6 weeks' GA. [3, 6]
Several factors account for these less favorable statistics. First, the vaginal approach detects cardiac activity earlier when the incidence of pregnancy loss is relatively higher. In addition, a number of other important observations have been made, which, when observed with a living embryo, are predictive of a poor outcome. 
At 5-6 weeks' GA, the mean embryonic heart rate is 101 beats per minute (bpm). This rate increases to 143 bpm by 8-9 weeks' GA and subsequently plateaus at approximately 140 bpm.  Therefore, it is not unusual for an initially detected embryonic heart rate to be somewhat slower than the fetal heart rate recorded later in pregnancy. An unusually slow heart rate is cause for concern. In one study, all embryos from 5+ to 8+ weeks GA in which the heart rate was less than 85 bpm resulted in spontaneous miscarriage. (See the image below.) 
As many as 18% of women with vaginal bleeding during the first half of pregnancy have ultrasonographic evidence for a subchorionic hemorrhage (displayed on the image below) as the etiology for their bleeding.  The clinical significance of this type of hemorrhage is controversial, with some investigators reporting an increased incidence of spontaneous abortion, [11, 12] and others concluding that this condition does not adversely affect pregnancy outcome.  Several authorities have suggested that the size of the blood clot can be used to predict the outcome  ; this has not been universally accepted. 
Abnormal yolk sac/amnion
The yolk sac normally forms by 28 menstrual days and is the first structure visible in the gestational sac. Normally, it should be seen on a transabdominal scan when the mean sac diameter (MSD) is 20 mm or larger.  This corresponds to a GA of 7 weeks. Transvaginal transducers can uniformly detect the yolk sac when the MSD is 8 mm or larger.  This corresponds to a GA of 5.5 weeks. Failure to visualize a yolk sac when the GA has reached these discriminatory values signals that the pregnancy is not progressing normally. A normal-appearing yolk sac at 5.5 weeks' GA is seen below.
An abnormal-appearing yolk sac also can predict subsequent demise. Abnormal features include large size (diameter greater than 6 mm, as seen in the image below), calcification or echogenic material within the yolk sac, and a double appearance to the yolk sac. [16, 17]
The amnion develops somewhat earlier than the yolk sac, but because this membrane is so thin, it is more difficult to visualize than the yolk sac. Normally, the amnion is visible on transabdominal scans late in the embryonic period. If the amnion is easily seen, it is probably too thick and most likely is abnormal.
Other features consistent with pregnancy failure include a visible amnion without a simultaneously visible yolk sac, embryo, or cardiac activity. An enlarged amniotic sac is another sonographic sign that predicts a failed pregnancy or embryonic death. 
To date, conflicting reports exist with regard to the usefulness of first trimester Doppler for predicting pregnancy outcome. Some reports suggest that if the resistive index is measured at the subchorionic level and exceeds 0.55, a high likelihood of spontaneous abortion exists  . Others claim, however, that Doppler analysis of these vessels is not predictive of outcome. [20, 21, 22, 23]
Visualizing an "empty" gestational sac
An "empty " gestational sac is the product of a normal early IUP or an abnormal IUP; another alternative is that the structure is actually a pseudogestational sac in a patient with an ectopic pregnancy. Based on careful ultrasonographic sac analysis, it may be possible to distinguish which of these alternatives is correct. Not infrequently, however, it is difficult or impossible to make this determination, in which case a follow-up ultrasonographic examination should be considered if clinically feasible.
A normal sac first appears as a small fluid collection surrounded by high-amplitude echoes embedded in the decidualized endometrium. This appearance has been termed the "intradecidual sac sign" (IDSS) (seen in the image below). 
Abnormal sac size
From 5.5-9 weeks' GA, the mean gestational sac size (MSS) is normally at least 5 mm greater than the CRL. When this difference is less than 5 mm, the subsequent spontaneous abortion rate exceeds 90%.  The etiology for first trimester oligohydramnios (seen in the image below) is unclear, but this observation suggests that with suboptimal first trimester gestational sac growth, a high likelihood of pregnancy loss exists.
An early normal intrauterine gestational sac often can be identified transabdominally by 31 days' GA and can consistently be identified by 35 days' GA. To confidently diagnose an IUP, most sonographers rely on the double decidual sac (DDS) finding, which is not universally present until the MSD is 10 mm (40 d GA). 
Specific size criteria can be used to distinguish normal from abnormal intrauterine gestational sacs. Using a transabdominal approach, size criteria that unequivocally suggest an abnormal sac include failure to detect a DDS when the MSD is equal to or greater than 10 mm, failure to detect a yolk sac when the MSD is equal to or greater than 20 mm, or failure to detect an embryo when the MSD is equal to or greater than 25 mm. 
Using vaginal ultrasonography, a normal intrauterine gestational sac can be detected reliably at 4-5 weeks' GA, at which time the MSD approaches 5 mm. Using vaginal transducers, criteria that suggest an abnormal sac include failure to detect a yolk sac when the MSD is 8 mm or greater, and failure to detect cardiac activity when the MSD exceeds 16 mm. 
Abnormal sac growth rate
The term "blighted ovum" (or anembryonic pregnancy) is used to describe an abnormal IUP with developmental arrest occurring prior to formation of the embryo or at a stage when it is not detectable using currently available equipment. In normal gestation, mean sac growth is 1.13 mm/d; in comparison, mean sac growth in an abnormal intrauterine gestation is 0.70 mm/d.  Based on these observations, abnormal sac growth can be diagnosed confidently if the gestational sac fails to grow by at least 0.6 mm/d. (See the image below.)
Choriodecidual appearance of sac
This refers to the ultrasonographic appearance of the echoes that surround an early intrauterine gestational sac. An abnormal appearance includes a distorted sac shape; a thin (< 2 mm), weakly echogenic, or irregular choriodecidual reaction; and absence of the double decidual sac sign when the MSD exceeds 10 mm. Features of choriodecidual sac appearance are seen in the image below.
Visualizing a central cavity complex
If the uterus appears normal on ultrasonography, or if the central echoes are prominent, most often the outcome will be unfavorable. This is because most patients with recognized pregnancy loss are approximately 11 weeks' pregnant (GA) when the ultrasonographic examination normally reveals intrauterine products of conception.
When the central cavity complex is abnormally thickened (and often irregularly echogenic), the differential diagnosis includes intrauterine blood; retained products following an incomplete spontaneous abortion; decidual changes secondary to an early, but not yet visible, intrauterine pregnancy; or a decidual reaction from an ectopic pregnancy. If the uterus has this appearance and the patient does not desire her pregnancy, uterine evacuation should be performed to detect the presence or absence of chorionic villi. If chorionic villi are absent, the patient remains at risk for an ectopic pregnancy.
If the patient desires to continue her pregnancy, the clinical status should determine whether serial tests (pregnancy and/or ultrasonographic) should be performed or if laparoscopy or laparotomy is required.
Degree of confidence
When performed by an experienced examiner, and with state of the art equipment, vaginal ultrasonography can detect an early intrauterine pregnancy with a high degree of confidence. This is particularly true once a yolk sac is identified. Using the criteria outlined above for predicting a poor outcome, it is usually possible to determine which pregnancies will fail.
However, it is important to note that these discriminatory criteria are guidelines. If certain findings are not observed at the appropriate time, if the ultrasonographic findings are equivocal, if the examination is technically difficult, or if the sonographer is inexperienced, caution is warranted. The embryo always should be given the benefit of the doubt, and a follow-up ultrasonographic examination should be performed to obviate any risk of terminating a normal intrauterine pregnancy.
Prior to visualizing the yolk sac, it is often not possible to be certain if the presence of a small, saclike intrauterine structure is the result of an early intrauterine pregnancy (normal or abnormal) or if the structure is a pseudosac associated with an ectopic pregnancy. This is because it may not be possible to clearly identify the IDSS. Under these circumstances, a follow-up examination should be performed if clinically feasible.
Occasionally, a subchorionic hemorrhage may resemble a second intrauterine sac. However, since most of these women are bleeding, the correct diagnosis usually can, with careful scanning, be made. Whenever uncertainty exists, perform a short interval follow-up examination at 5-7 days.
Later in the first trimester, several anatomic structures undergo developmental changes that can be misinterpreted as abnormal. One potential pitfall is misinterpreting the developing rhombencephalon as an abnormal intracranial cystic structure, such as hydrocephalus or a Dandy Walker cyst. Note that because these anomalies require second trimester imaging, these conditions should not be diagnosed during the first trimester.
Another potential source of confusion is misinterpreting physiologic herniation of the bowel within the umbilical cord as an abdominal wall defect, such as an omphalocele. In normal patients, the diameter of the base of the cord (containing herniated bowel) should be less than 7 mm; in addition, no appreciable herniation should be seen once the CRL is greater than 45 mm. In questionable cases, perform careful follow-up ultrasonographic imaging.