Morphokinetic analysis and

embryonic prediction for blastocyst

formation through an integrated
time-lapse system
Yamileth Motato, Ph.D., María Jose
de los Santos, Ph.D., María Jose

Escriba, Ph.D.,

n Aparicio Ruiz, Ph.D., Jose
Bele
Remohí, M.D., and Marcos Meseguer, Ph.D.
Instituto Valenciano de Infertilidad, Universidad de Valencia, Valencia, Spain

Objective: To describe the events associated with the blastocyst formation and implantation that occur in embryos during preimplan-
tation development based on the largest sample size ever described with time-lapse monitoring.
Design: Observational, retrospective, single-center clinical study.
Setting: University-affiliated private IVF center.
Patient(s): A total of 7,483 zygotes from 990 first treatments of intracytoplasmic sperm injection (ICSI; 627 of oocyte donor vs. 363
autologous oocyte cycles), of which 832 blastocysts were transferred.
Intervention(s): No patient intervention. Embryos were cultured in a time-lapse monitoring system, and the embryos were transferred
on day 5 after ICSI. Embryo selection was based on the multivariable model previously developed and on blastocyst morphology.
Main Outcome Measure(s): Using a time-lapse system, embryo images were acquired every 15 minutes for 120 hours. Embryos
cleavage time points up to the 9-cell stage (t2–t9) as well as to the morula stage (tM) and blastocyst formation (tB) were registered
in hours after ICSI. Additionally, duration of the cell cycle and synchrony of the second and third cell cycles were defined. As a
result, we have monitored the embryonic development of a total of 3,215 blastocysts, of which 832 were transferred. Finally, we
analyzed the characteristics of embryonic development of blastocyst (phase 1) and of implanted and not implanted (phase 2)
embryos as finally validated in an independent data set (phase 3).
Result(s): A detailed retrospective analysis of cleavage times was made for 7,483 zygotes. We analyzed 17 parameters and found
several significantly correlated with subsequent blastocyst formation and implantation. The most predictive parameters for blas-
tocyst formation were time of morula formation, tM (81.28–96.0 hours after ICSI), and t8–t5 (%8.78 hours) or time of transition of
5-blastomere embryos to 8-blastomere embryos with a receiver operating characteristic curve (ROC) value ¼ 0.849 (95% confidence
interval [CI], 0.835–0.854; phase 1). These parameters were less predictive of implantation, with a ROC value ¼ 0.546 (95% CI,
0.507–0.585). We also observed that time for expansion blastocyst, tEB (107.9–112.9 hours after ICSI), and t8–t5 (%5.67 hours
after ICSI) predict blastocyst implantation, with a ROC value ¼ 0.591 (95% CI, 0.552–0.630; phase 2). The model was validated
on an independent data set and gave a ROC of 0.596 (0.526–0.666; phase 3).
Conclusion(s): The inclusion of kinetic parameters into score evaluation may improve blastocyst selection criteria and can predict
blastocyst formation with high accuracy. We propose two multivariable models based on our findings to classify embryos according
to their probabilities of blastocyst stage and implantation in the largest data set ever reported
of human blastocysts. (Fertil SterilÒ 2016;105:376–84. Ó2016 by American Society for Repro-
ductive Medicine.) Use your smartphone
Key Words: Embryo quality, implantation, time lapse, assisted reproductive technologies, to scan this QR code
morphokinetic parameters and connect to the
discussion forum for
this article now.*
Discuss: You can discuss this article with its authors and with other ASRM members at http://
fertstertforum.com/motatoy-blastocyst-prediction-time-lapse/ * Download a free QR code scanner by searching for “QR
scanner” in your smartphone’s app store or app marketplace.

Received February 5, 2015; revised October 9, 2015; accepted November 2, 2015; published online November 17, 2015.
Y.M. has nothing to disclose. M.J.d.l.S. has nothing to disclose. M.J.E. has nothing to disclose. B.A.R. has nothing to disclose. J.R. has nothing to disclose. M.M.
was supported by the Spanish Ministry of Economy and Competitiveness (PI14/00523) through the Instituto de Salud Carlos III program.
This work has not received any financial support from any commercial entity, and the instrumentation, disposables, and utensils have been fully paid for by
Instituto Valenciano de Infertilidad (IVI). None of the authors have any economic affiliation with Unisense FertiliTech A/S. IVI is a minor shareholder in
Unisense FertiliTech A/S.
Reprint requests: Marcos Meseguer, Ph.D., Instituto Valenciano de Infertilidad, Plaza de la Policía Local, 3, Valencia 46015, Spain (E-mail: marcos.
meseguer@ivi.es).

Fertility and Sterility® Vol. 105, No. 2, February 2016 0015-0282/$36.00

Copyright ©2016 American Society for Reproductive Medicine, Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.fertnstert.2015.11.001

376 VOL. 105 NO. 2 / FEBRUARY 2016


T
he morphological classification of gametes and em-
bryos has been used since the beginning of IVF until
today as a tool to assess embryo development (1)
and to select the best embryo for transfer (2). Embryos are
routinely evaluated at a single or few time points on day 2
or 3 (D2 or D3). This handling is likely to cause fluctuations
in temperature and pH during routine microscopic evalua-
tion, which can harm embryos. Although used in most lab-
oratories as a general indicator of embryo quality, this
methodology may not select the most competent embryo,
as a consequence of the nonspecific and subjective charac-
teristics of this method. In addition, the conventional em-
bryo selection method is associated with a relatively low
IVF success rate (3). Therefore, the final decision on which
embryo to transfer is traditionally based predominantly on
the number and symmetry of blastomeres at the day of trans-
fer, taking into account morphology evaluation at earlier
observations. As a compensative approach, extended em-
bryo culture and transfer at the blastocyst stage (D5 or D6)
becomes an alternative that permits the selection of embryos
at more advanced stages of development and minimizes the
risk of multiple pregnancies, a figure that in 2010 reached
an alarming threshold of 20.6% (4). Although multiple ob-
servations give a better understanding of embryo develop-
ment, it is well known that each observation also involves
exposure to suboptimal conditions outside the controlled
environment of an incubator, potentially affecting the suc-
cess of the treatment.
In contrast, automatic time-lapse systems offer the pos-
sibility to monitor embryo development continuously
throughout the culture period (5). This is obtained through
the digital image capture defined by programmed time
intervals and increases the quality and quantity of the
information. In addition, time-lapse technology allows
maintaining more stable culture conditions and is considered
a safe tool in IVF laboratories, mainly because no adverse ef-
fects on the embryo can be detected (6–9). Since the first
analysis of human embryonic development with time-lapse
imaging that included the process of fertilization and assess-
ment of early events (10), numerous studies have used this
technology (11, 12), including the search for noninvasive
prognostic markers that predict embryo development and
the outcome of IVF treatments. Meseguer et al. (13) introduced
specific temporal development markers that were related
to subsequent implantation. They evaluated the chronological
pattern of cell divisions as well as other morphological
features to identify connections between the time taken
to reach each embryonic event and the implantation
potential of the specific embryo and proposed a hierarchical
classification based on the parameters t5 (time division
5 cells), t4–t3, and t3–t2 to select the most viable embryos
for transfer in each treatment cycle. However, this model
only assesses embryos until D3 of development and is not
associated with the formation and quality of blastocysts or
the outcomes of success in IVF treatments that focus on
blastocyst culture.
Conversely, other studies using this tool have focused on
identifying morphokinetic parameters that predict blastocyst
formation and blastocyst morphology (14–18), which could
VOL. 105 NO. 2 / FEBRUARY 2016
Fertility and Sterility®
potentially improve the reproductive outcome compared
with the methodology of conventional incubation (8).
Wong et al. (19) found that development of human em-
bryos to the blastocyst stage was linked to key timings in early
embryos development like the duration of the first cytoplasm
cleavage from 1 to 2 cells (cytokinesis) and the length of the
interval between divisions in the first stages of embryonic
development. Later Campbell et al. described the risk to em-
bryos of having single or multiple aneuploid chromosome
constitution, and their results showed that multiple aneuploid
embryos were delayed at the initiation of compaction, initia-
tion of blastulation, and the time to reach full blastocyst stage.
So they developed a predictive algorithm based on the dy-
namic events (20). Although in the retrospective analysis to
evaluate the effectiveness and potential impact of this model
(21) their results showed that this model offers a good predic-
tion of pregnancy rates and outcomes and although other in-
vestigators (20, 22–24) defined optimal ranges of embryonic
development associated with chromosomal normality, no
prospective study or test comparing these algorithms with
the conventional embryo selection procedure and conversely
has been published.
Thus, time-lapse technology becomes an important tool
in IVF that allows identification of a number of critical events
during embryonic development. Therefore, the present study
offers a unique opportunity for a comprehensive retrospective
analysis and evaluates blastocyst development through time-
lapse monitoring (TLM) and the effect on treatment outcome.
The purpose of this study was to assess the time of embry-
onic events and determine which have the ability to predict
blastocyst formation and implantation potential using a
TLM and a multivariable morphokinetic model until D5.
MATERIALS AND METHODS
This research project was conducted at the Instituto Valen-
ciano de Infertilidad (IVI) Valencia and was performed from
May 2010 until May 2014. The procedure and protocol for
analysis of embryos were approved by an Institutional Review
Board (IRB reference 1404-VLC-014-YM), which regulates
and approves database analysis and clinical IVF procedures
for research at IVI. Additionally, the project complies with
the Spanish law governing assisted reproductive technologies
(14/2006). The present retrospective cohort study was drawn
from a total of 990 first treatments of intracytoplasmic sperm
injection (ICSI). Among the embryos included in the study,
3,215 (42.96%) developed to the blastocyst stage, 832
(11.11%) blastocysts were transferred in fresh treatment,
and the remaining 2,383 (31.84%) were vitrified. Finally,
4,268 (57.03%) embryos did not reach blastocyst stage
(nonviable and discarded; Supplemental Fig. 1).
Phase 1 of the study included the generation of the algo-
rithm for blastocyst formation. The data to develop the new
algorithm were obtained from a total of 990 cycles giving
rise to 7,483 zygotes and 3,215 blastocysts. Phase 2 included
the generation of the algorithm for implantation blastocysts.
The data to develop the new algorithm were obtained from
832 blastocyst with known implantation (KID) transferred
blastocysts, from which 383 implanted. Phase 3 included
377
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
the validation of the algorithm for implantation blastocysts
once they develop. The algorithm was validated from 328
cycles with 257 KID transferred blastocysts from which 123
implanted, between May 2013 and May 2014. Clinical preg-
nancy rate was confirmed by the existence of a fetal heartbeat
at ultrasound after 7 weeks of gestation.
All embryos were obtained after fertilization by ICSI.
Annotation of morphokinetic characteristics was performed
on all available embryos, including all transferred, vitrified,
and also finally discarded, that were cultured until D5.
The exclusion criteria for standard patients and recipients
with respect to this study were low response (less than five
metaphase II oocytes), endometriosis, polycystic ovary syn-
drome (PCOS), hydrosalpynx, body mass index (BMI)
>30 kg/m2, uterine pathology (myomas, adenomyiosis, endo-
crinopathies, thrombophilia, chronic pathologies, acquired or
congenital uterine abnormalities), recurrent pregnancy loss,
patients included in the preimplantation genetic diagnosis
and preimplantation genetic screening (PGS) program,
maternal age over 45 for oocytes donation recipients, and se-
vere male factor (presenting less than 1 million sperm cells in
total in the ejaculate).
Ovarian Stimulation in Standard Patients and
Oocytes Donors
All donors were from our egg donation program. The selection
criteria for donors can be found in Garrido et al. (25) as stated
by Spanish law. All donors had normal menstrual cycles of
26–34 days' duration, normal weight (BMI of 18–28 kg/m2),
no endocrine treatment (including gonadotropins and oral
contraception) in the 3 months before the study, normal
uterus and ovaries at transvaginal ultrasound (no signs of
PCOS), and antral follicle count >20 on the first day of
gonadotropin administration (13).
Donors and standard patients had started ovarian con-
traceptive pills (OCPs; Microgynon30, Organon) on D1–D2
of the menses of the previous cycle and continued for
12–16 days. In the donors, 5 days after stopping OCPs they
received a starting dose of recombinant FSH (Gonal-F; Se-
rono; Puregon; MSD) ranging from 150 to 225 IU. GnRH
antagonist (0.25 mg ganirelix, Orgalutran, Organon) was
administered daily starting on D5 or D6 after FSH adminis-
tration, and hCG (Ovitrelle, Serono Laboratories) was admin-
istered SC for final oocyte maturation when at least three or
more follicles reached a mean size of R18 mm (26).
The protocol for endometrial preparation of recipients can
be found in Mu~ noz et al. (26). After ET for luteal phase sup-
port, the oocytes recipients received a daily dose of 400 mg
vaginal micronized P (Progeffik) every 12 hours.
Ovum Pick-up and ICSI
Oocyte retrieval was performed 36 hours after hCG, under ul-
trasound guidance. Follicles were aspirated, and the oocytes
were washed in Gamete Medium (COOK). After washing, oo-
cytes were cultured in fertilization medium (Cleavage Me-
dium, COOK) at 5.5% CO2 in air and 37 C for 4 hours before
oocyte denudation. Oocyte denudation was carried out by
378
mechanical pipetting with 40 IU/mL of hyaluronidase in the
same medium. Subsequently, ICSI was performed on all meta-
phase II oocytes in a medium containing HEPES (Gamete
Medium) at 400 magnification using an Olympus IX7
microscope.
Incubation and Imaging System
Immediately after ICSI, the injected oocytes were placed indi-
vidually in Cleavage Medium (Cook) preequilibrated in cul-
ture dishes (EmbryoSlide Vitrolife) until D3 (72 hours after
ICSI) covered with 1.2 mL of mineral oil. On D3, the medium
was changed to CCM medium (Vitrolife) until D5 (120 hours
after ICSI), and culture was continued at 37 C and 5.5%
CO2. All incubations were done in a time-lapse incubator
(EmbryoScope, Unisense FertiliTech A/S).
Image acquisition in the TLM system was performed
every 15 minutes. For each embryo and time point, we ac-
quired stacks of five images at different focal planes to enable
accurate assessment of the embryo morphology. The images
were acquired with low-intensity red light with exposure
times of 15–30 ms per image. Thus, the total light exposure
in this methodology was lower than the exposure during
routine microscopy (13, 27).
Embryo Score
Successful fertilization was assessed at 16–19 hours post-ICSI
and confirmed by the presence of two pronuclei and two polar
bodies, based on digital images acquired with the TLM system
and evaluated on an external computer with software for
analysis (EmbryoViewer workstation, Vitrolife). Similarly,
embryo morphology was analyzed on the basis of the ac-
quired digital images on D2 (48 hours post-ICSI); D3 (72 hours
post-ICSI) taking into account the number, symmetry, and
granularity of the blastomeres, type and percentage of frag-
mentation, presence of multinucleated blastomeres, and de-
gree of compaction as previously described by Alikani et al.
(28); and D5 (120 hours post-ICSI) according to the expansion
of blastocoele cavity and the number and integrity of both the
inner cell mass and trophectoderm cell. According to this
grading system, we selected the embryos for transfer on D5
to make a comparison between morphology and the time-
lapse classification tree categories, as we used ASEBIR classi-
fication as previously published (2, 29); a summary of this
classification tree is described in Supplemental Figure 2.
Time-lapse Images and Selection
In this study, we analyzed the different cellular divisions:
time of cleavage from 2 to 9 cells (t2, t3, t4, t5, t6, t7, t8,
and t9). The times were annotated at the first frame in which
the cells (blastomeres) were seen as separated by individual
membranes; at timing of morula (tM), which is defined as
the stage at which it was no longer possible to visualize
the individual blastomere membranes and full embryo
compaction was achieved; and at timing of blastocyst (tB),
which consisted of the time from insemination to the forma-
tion of a ‘‘full blastocyst’’ when the blastocoele cavity filled
the embryo and the inner cell mass and trophectoderm
VOL. 105 NO. 2 / FEBRUARY 2016

tissues were distinguishable from each other. Thus, the blas-
tocyst is composed of three parts: the two cell types and the
fluid cavity. When development of blastocysts progresses,
cells in the two components divide and the fluid cavity
enlarges. The successive step was the time of expanded blas-
tocyst or tEB; consistent with the increase of the overall vol-
ume of the embryo and expansion of the blastocoele cavity,
this is the time frame after the zona pellucida starts to thin, so
that the blastocoel was as full as possible, which caused any
expansion/zona thinning. The final step is the timing of
hatching of the blastocyst or tHB, the time that the trophec-
toderm starts herniating through the zona pellucida. For
blastocysts, development of the inner cell mass was assessed
as follows: A—tightly packed many cells (compact); B—
loosely grouped several cells (noncompact); C—very few
cells; and D—degenerate or nonexistent. The trophectoderm
was assessed as follows: A—many cell forming a cohesive
epithelium; B—homogenous but few cells; C—very few cells;
and D—degenerate or degenerating (2). The time of ICSI was
taken as time zero (t0) for determination of all durations.
Variables related to the duration of cell cycles were also
determined and designated: duration of second cell cycle
as the time from division to a 2-blastomere embryo until di-
vision to a 3-blastomere embryo or t3–t2 duration of third
cell cycle or t5–t3 the duration of the transition from a 3-
blastomere embryo to a 4-blastomere embryo or t4–t3 tran-
sition of a 5-blastomere embryo to an 8-blastomere embryo
or t8–t5 and the interval between 2 and 5 cells as the variable
t5–t2, which combines the concepts of cell cycle and syn-
chrony (Supplemental Fig. 3).
Embryos were selected for transfer according to a com-
bination of standard morphological grading in D5 and using
the approach described previously by our team (13) based on
the hierarchical classification procedure for the selection of
viable embryos for transfer, which subdivided embryos
into 10 categories from A to F according to optimal range
for t5 (48.8–56.6 hours), t4–t3 (%0.76 hours), and
(%11.9 hours; Supplemental Fig. 4) including novel mor-
phokinetic exclusion criteria (direct cleavage from 1 to 3 cells
and presence of multinucleation at the 4-cell stage embryo),
which are not correlated to blastocyst formation tB, tEB, and
tHB and which were not used as selection criteria but were
studied retrospectively in relation to the outcome of the ET.
ET
The number of embryos transferred was normally two accord-
ing to embryo quality. Supernumerary embryos were frozen
for potential future transfer using the IVI standard vitrifica-
tion technique (30). The b-hCG value was determined
13 days after ET, and clinical pregnancy was confirmed
when a gestational sac with fetal heartbeat was visible by ul-
trasound examination after 7 weeks of pregnancy.
Statistical Analysis
The investigated embryos were divided into two groups based
on whether or not they developed to blastocyst stage. Anal-
ysis of variance was used to test for significance differences
VOL. 105 NO. 2 / FEBRUARY 2016
Fertility and Sterility®
in the mean times for embryonic events. Variables included
in this study did follow a normal distribution. Chi-square tests
were used to compare categorical data. The analysis included
blastocysts in the fifth day of culture, transferred or frozen,
and embryos that had not developed to the blastocyst stage.
The results of these observations cannot be obtained by
means of direct comparison of analyzed parameters between
both groups (blastocyst and not blastocyst) because very high
as well as very low values exist and they are not good predic-
tors of an embryo's ability to reach the blastocyst stage.
The statistical model chosen was logistic regression and
was performed by using defined variables to select the most
relevant. The binary response parameter blastocyst stage
was either ‘‘1’’ for blastocyst formation or ‘‘0’’ for absence
of blastocyst stage. In a second analysis, we defined the
binary response parameter as the presence of a gestational
sac, ‘‘1’’ (implanted blastocyst), or no gestational sac, ‘‘0’’.
To describe the distribution of the probabilities of blas-
tocyst formation or successful implantation, timings were
converted from continuous variables into categorical vari-
ables by dividing them into groups based on their quartiles.
P< .05 was considered statistically significant. By this pro-
cedure we avoided bias due to differences in the total number
of embryos in each category. We then calculated the per-
centage of embryos with blastocyst formation for each
timing quartile to assess the distribution of blastocyst stage
in the different categories. Chi-square tests were used to
compare between categorical data. For each timing variable,
an optimal range was defined as the combined range
spanned by the two quartiles with the highest percentage
of blastocysts formed. Additionally, a binary variable was
defined with the value inside (outside) if the value of the
timing variables was inside (outside) the optimal range. In
the regression analysis, we determined how the significant
binary variables were combined and related with blastocyst
formation or implantation potential. In this way, binary var-
iables were submitted to computer analysis by the forward
step method (likelihood method), and those with P>.05
were not considered in the final model. Following this statis-
tical analysis, the variables in question were classified as
those that significantly conditioned blastocyst formation
or implantation potential.
The odds ratio (OR) of the effect of all binary variables
generated on blastocyst formed or implanted was expressed
in terms of 95% confidence interval (95% CI) and signifi-
cance. Receiver operating characteristic (ROC) curves were
used to test the predictive value of all the variables included
in the model with respect to blastocyst stage and implanta-
tion. ROC curve analysis provides area under the curve
(AUC) values that are between 0.5 and 1 and can be inter-
preted as a measurement of the global classification ability
of the model. A further analysis was performed combining
embryo morphological classification according to ASEBIR
(A, B, C, D) and our resulting algorithm for implantation (A,
B, C, D); in this model we included oocyte donation cycles
source and the age of the patient (own cycles) and the donor
(oocyte donation cycles) as a potential bias factor to be
controlled. The resulting model was also tested for predictive
value by ROC analysis.
379
ORIGINAL ARTICLE: ASSISTED REPRODUCTION

Statistical analysis was performed using the Statistical mental in embryos cultured until D5 (t2, t3, t4, t5, t6, t7, t8,
Package for the Social Sciences 19 (SPSS). and t9, tM, tB, tEB, tHB, t3–t2, t4–t3, and t5–t2).
The next step in the model was to classify them according
RESULTS to binary timing variables defined by the logistic regression
model: tM (81.28–96.00 hours) and t8–t5 (%8.78 hours). The
Our study analyzes two groups of embryos from 7,483 zygotes
predictive value of the variables included in the model with
of 990 treatments of ICSI. Group 1 consisted of 3,215 embryos
respect to blastocyst formation presented a ROC value ¼
that reached the blastocyst stage, giving rise to a 42.96% blas-
0.849 (95% CI, 0.835–0.854; P%.05). If the values of tM fall
tocyst formation rate. Group 2 consisted of 4,268 embryos
inside the optimal range 81.28–96.00, the embryo is catego-
that did not develop to the blastocyst stage. Supplemental
rized as A or B depending on whether the values t8–t5 fall in-
Table 1 shows the annotation of the different morphokinetics
side the optimal range (%8.78 hours); similarly, if the tM falls
parameters for these two groups of embryos.
outside the optimal range, the embryo is categorized as C or D
In group 1, the first cleavage occurs earlier compared with
depending on t8–t5.
the second group (27.23  0.2 hours vs. 28.78  0.2 hours;
Thus the hierarchical classification procedure divides all
P< .05). Such a difference persisted in almost all timing divi-
the 7,483 evaluated embryos into four different categories,
sions, except for t5, t9, and t5–t3, in which there was no sig-
with decreasing potential of blastocyst formation (Table 1).
nificant difference. Other characteristics associated with
We also tested the previously defined binary variables
embryo quality (e.g., fragmentation or multinucleation)
and their relation to implantation potential with no signifi-
were not taken into consideration in this analysis to assess
cant results (data not shown). When the predictive value of
separately the specific impact of cleavage dynamics on subse-
variables that were significant for blastocyst formation (tM
quent development.
and t8–t5) was examined for its predictivity in regard to im-
Supplemental Table 2 describes the four quartiles for the
plantation, the ROC value was 0.546 (95% CI, 0.507–0.585;
timing of each parameter together with the percentages of
P¼ .023), which indicates that this hierarchal model has little,
blastocyst in each quartile.
if any, utility for predicting viable blastocysts and, thus, for
selecting viable embryos.
Phase 1: Embryo Scoring Based on a Classification
Tree to Select Embryos with Higher Blastocyst
Phase 2: Blastocyst Transferred and Implantation
Formation Probabilities
Rate
The observed correlations between morphokinetic parameters
Owing to the lack of a relationship between the previously
and blastocyst formation are the basis for a proposed hierar-
described variables with implantation potential, we identified
chical classification procedure to select viable embryos with a
new variables by comparing transferred blastocysts (n ¼ 383)
high blastocyst (when the blastocoele cavity filled the em-
that implanted with those that did not implant (n ¼ 449). We
bryo) formation potential. Using our database, we subdivided
analyzed 17 morphokinetic parameters to identify temporal
the embryos into four categories from A to D as shown in
developmental markers that predict blastocyst implantation.
Figure 1.
For each of the identified parameters, we selected the two
The hierarchical classification procedure starts with a reg-
quartiles with the highest frequency of implanting blastocyst.
ister of timings of 17 morphokinetic parameter of develop-
We observed a significant difference in blastocysts with an
optimal range for %112.9 hours and timing of transition
FIGURE 1 from 5-blastomere embryos to 8-blastomere embryos (t8–t5)
%5.67 hours as shown in Figure 2.
Using the present data, we made a hierarchical model rep-
resenting a classification tree, which subdivided blastocysts
into four categories from A to D with higher or lower implan-
tation rate (i.e., from 72.2% in category A to 39.7% in cate-
gory D) as show in Table 2. The logistic regression on 832
blastocysts transferred and analyzed with the morphokinetic
model for blastocyst selection—383 with full implantation

TABLE 1

Distribution of blastocyst according to morphokinetic model category.

Blastocyst
Hierarchical classification of embryos based on [1] timing of morula, Category Total formations % (95% CI)
optimal range 81.28–96.00 hours; [2] timing of transition from A 1,624 1,370 84.4 (82.64–86.16)
5-blastomere embryo to until 8-blastomere embryo (t8–t5), optimal B 1,461 1,095 75.0 (72.89–77.1)
range %8.78 hours. The classification generates four categories of C 960 276 28.8 (26.6–31.0)
embryos with increasing expected blastocyst formation (right to left). D 3,438 474 13.8 (12.12–15.48)
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016. Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

380 VOL. 105 NO. 2 / FEBRUARY 2016


FIGURE 2
Implantation rate of embryos exhibiting or not optimal ranges for tEB and t8–t5 kinetic parameters. Time ranges are described in hours after ICSI.
*P<.05.
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.
(number of gestational sacs matched the number of trans-
ferred embryos) and 449 not implanted—gives a ROC of
0.591 (95% CI, 0.552–0.630; P< .001). We performed a further
analysis in which blastocyst morphology was included; the
four morphology categories were related to implantation
rate (Supplemental Table 3). We additionally aimed to
combine blastocyst morphology and the hierarchical model
by logistic regression to confirm the relevance of the latter
when morphology is taken into consideration, including
oocyte donation and age (patient for own oocytes and donors
for oocyte donation) as potential bias factor (Supplemental
Table 4). The combined model gives a ROC of 0.602 (95%
CI, 0.559–0.645; P< .001), while the model of implantation
with blastocyst morphology only gives a ROC of 0.561 (95%
CI, 0.517–0.605; P¼ .007).
Phase 3: Validation of the Implantation Model
After the conclusion of phase 2, the created model was vali-
dated on an independent data set composed of 257 embryos,
123 blastocysts with known implantation (KID embryos;
Supplemental Table 5), which gives a ROC of 0.596 (95% CI,
0.526–0.666; P¼ .008).
DISCUSSION
Morphological evaluation has been the primary method used
for embryo assessment, and it remains the most commonly
TABLE 2
Distribution of implanted blastocyst according to morphokinetic
categories.
Blastocyst
Category Total implantations
A 36 26
B 65 43
C 147 82
D 584 232
Note: Implantation in the blastocyst categories of the hierarchical classification tree model
applied in the TLM. Results refer only to those blastocystd with known implantation (KID
embryos).
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.
VOL. 105 NO. 2 / FEBRUARY 2016
Fertility and Sterility®
used approach for selection because embryo quality is the
most predictive factor of success of treatment (31). However,
this method limits the amount of information available,
introduces an important element of subjectivity, and is associ-
ated with low IVF success rate (clinical pregnancy rate of
30% per transfer) (32), especially in programs with premature
ET on D3 or earlier. For this reason, the selection of viable em-
bryos for transfer is decisive, although the characteristics and
the methods to identify ‘‘optimal embryos’’ with high develop-
ment and implantation potential are still not comprehensive.
Initial observations describing human fertilization and
early development events through time-lapse microscopy
were reported by Payne et al. (10) and several years later by
Lemmen et al. (5). Time lapse is a noninvasive technology
that has attracted considerable interest during the last years
especially after the development of equipment suitable for
routine application in clinical embryology. One important
reason is that this technology allows the study of events that
usually happen at a speed imperceptible to the human eye.
Thus, TLM is an ideal tool to study the dynamic biolog-
ical processes of embryo development, permits the selection
of embryos at more advanced stages of development, and
provides morphological, dynamic, and quantitative timing
data, that is, morphokinetic parameters. These parameters
may be used to estimate embryo viability and implantation
potential (5). Therefore, these recent findings suggest that
morphokinetic parameters may supplement traditional em-
bryo selection methods to increase clinical pregnancy rates
for IVF treatments.
However, this method has some drawbacks and limita-
tions, as reported by Herrero et al. (33). For instance, the
time-lapse system does not permit rotation of the embryos,
making visual observation limited, in particular when blas-
% (95% CI) tomeres overlap or a high level of cytoplasmic fragmentation
72.2 (57.56–86.84) is present.
66.2 (54.70–77.70) To date, several published studies have used time-lapse
55.8 (47.77–63.83) technology, however, most have been limited mainly to mea-
39.7 (35.73–43.67)
surements of early development, such as pronuclear formation
and fusion and time to first cleavage (5, 12, 34, 35), whereas
relatively few, based on a small number of samples (17),
aimed at studying the blastocyst formation (36). For this
381
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
reason, in our study we sought to overcome these limitations
and to define events of human embryo development
associated with the ability of an embryo to reach the
blastocyst stage and continue to implant, mainly because the
transfer of embryos at later stages allows greater selection
and is associated with substantially higher implantation
rates compared with transfer of early-stage embryos (37, 38).
To our knowledge, this retrospective study uses the
largest database reported for human blastocysts, and it is
the first that provides a robust model suggesting a new clas-
sification algorithm based on late events of embryo develop-
ment. We found that development to blastocyst could be
predicted within 96 hours of culture by morula formation
(tM) and short duration of the transition from five to eight
blastomeres (t8–t5). In addition, our results in general indicate
that the timings of cleavages from two to the morula stage
occur progressively earlier (except t5 and t9) in embryos
with the ability to develop to blastocyst and are in particular
faster during progression from three to four blastomeres (t4–
t3). The duration of the second and third cell cycles is longer in
these embryos compared with in embryos arresting before
blastocyst stage, which may indicate that a reasonable time
in these cell stages is required for an ordered and efficient
DNA synthesis as part of the cell cycle. Thus, these embryos
showed significant differences in the temporal patterns to
achieve the formation of the blastocyst.
As a result, it is evident that later kinetic parameters
represent one of the most potent noninvasive tools to improve
selection of embryos that develop to blastocyst. Some inves-
tigators, for example, Herrero et al., studied 9,530 embryos
and found that late events such as t5 and t8 are significantly
more predictive of embryo viability to later stages such as
blastocyst stage than earlier timings measured on D2 of
development (39). Another investigation recently reported
by Cetinkaya et al. confirmed that cleavage kinetics assessed
by time-lapse microscopy could be adopted as a tool to predict
blastocyst development and quality (18). By monitoring 3,354
embryos, the investigators analyzed the ratios of cleavage
timings (CS2-8, CS4-8) and a time interval (t8–t5) and pro-
pose the equation CS2-8 ¼ [(t3–t2) þ (t5–t4)]/(t8–t2) that
reflect the synchronicity of cell cycles. In the same way, other
investigators suggest that times of early cleavage (from the 2-
to the 8–cell stage) can also determine the ability to develop to
blastocyst stage, undergo blastocoele expansion, and
implant; for example, the study by Cruz et al. 2012 (14) pre-
sented a time-lapse analysis of 834 embryos, in which they
related blastocyst formation capacity with the timings t5
and cell cycle 2. Their observations indicate that embryos
that cleave earlier have a significantly improved chance of
continuing development to D5 when compared with embryos
that develop more slowly.
On the other hand, the low reliability of predicting suc-
cess rates by embryo selection within the first 2–3 days of
culture has increasingly favored blastocyst transfer,
although some studies have suggested a potential influence
of in vitro culture on the offspring in terms of epigenetic
modifications (40, 41). Transfer on D5 is also considered a
strategy that more resembles the physiological course of
implantation (42), helps to select embryos after activation
382
of the embryonic genome, and has proved efficient in
preventing multiple pregnancies in particular.
In fact, a review by Blake et al. (43) concluded that there
were significant differences in live-birth rate in favor of blas-
tocyst transfers. One of the main reasons for this statement is
that more chromosomally normal embryos continue to devel-
opment to D5 or D6 (44). Similarly, Basile's group (23),
through the combination of time-lapse technology with
PGS, analyzed the morphokinetic behavior of chromosomally
normal and abnormal embryos in a cohort of 504 embryos,
and they have proposed an algorithm based on D3 data that
can increase the probability of selecting chromosomally
normal embryos in the absence of PGS. In the same way,
Campbell et al. (20) indicate that with TLM until the blastocyst
stage, it is possible to avoid the selection of embryos having a
high risk of aneuploidy. Thus, the selection of embryos
through TLM, mainly using late parameters (t5–t2, t5–t3,
and tB), represents a good tool, especially for patients who
have no obvious indication for PGS.
To conclude, in an attempt to supplement existing data
about cleavages times and duration of cell cycle length ob-
tained by time lapse, and considering our previous study,
Meseguer et al., that was based on 247 transferred and im-
planted embryos, we have contrasted our averages for blas-
tocyst formation with implantation rate according to tM
and t8–t5. Interestingly our selection model that was devel-
oped for prediction of blastocyst formation does not correlate
with implantation rates. By adding the timing of later events
such as tEB together with the duration of the transition from
5 to 8 blastomere embryos (t8–t5), the model is related to im-
plantation potential, although as a diagnostic tool for em-
bryo outcome we should consider it as a poor performing test.
As result, we proposed two models; the first model was
developed to identify embryos that develop until blastocyst
stage, while the second model is constituted by blastocysts
transferred and clinical outcome as the endpoint. An indepen-
dent data set was also used to validate the model prediction of
implantation, and the obtained results were very similar. The
AUC was similar although slightly higher (AUC ¼ 0.596). Both
the endpoint and the study population differ between the two
models. So it is expected that a model that predicts blastocyst
development would be different from a model that predicts
clinical outcome (45). In our opinion, however, we expect
that a model that predicts blastocyst development would posi-
tively help to select embryos with high potential of implanta-
tion and clinical pregnancy. Nevertheless, we do not find any
explanation, neither in the different study populations nor in
the different endpoints, as to why a proportion of embryos
that were classified as A or B because they are located inside
the optimal range according to t8–t5 in the model selection
for blastocyst formation resulted in failed implantation. One
explanation could be the narrow time intervals for optimal
cellular division.
We used conventional morphology for grading; in conse-
quence, we take into consideration blastocyst morphology in
our morphokinetic models by multivariable analysis to assess
any additional value of the presented model in predicting im-
plantation. As reflected by the logistic regression, morphology
did not add to or change the relationship of our model with the
VOL. 105 NO. 2 / FEBRUARY 2016

implantation potential. Considering that both morphology
and morphokinetics are related to blastocysts, outcome our
study reveals an autonomous behavior of both parameters
and the possibility of using it independently with similar suc-
cess, although detailed observation reveals slightly stronger
features of the morphokinetic model. The potential bias pro-
duced by the inclusion of donors, which represent a different
population of women compared with infertility patients, was
considered and controlled and did not affect the prediction
of implantation of the model.
It is important to note, however, that our study has some
limitations; first, the morphokinetics parameters in this
study are developed on a data set from a specific clinic,
and thus patterns of developing embryos are likely to differ
for other clinics culturing embryos under different condi-
tions (46). Second, although percentage and type of frag-
mentation have been traditional criteria in selecting
embryos for transfer due to their implications in embryo
development (47) and despite the suggestion that embryo
fragmentation might disappear as the embryo develops
(48), this work did not analyze this influence in relation to
kinetics of development due to the software limitations.
Third, in this study it is not possible to discriminate between
dizygotic or monozygotic pregnancy results in the transfer of
two blastocysts. Fourth and finally, clinical pregnancy is not
the most meaningful endpoint, which should be live birth,
but unfortunately we have not be able to follow the better
approach.
In summary, the data presented in this paper improve our
knowledge of the real time distribution of the main events of
embryo development, making it possible to establish, with ac-
curacy and objectivity, the basis of the embryo development
timing. In addition, this technology offers completely new
possibilities in the evaluation of embryo development and
morphokinetic parameters that allow for the creation of a
model that significantly differentiates embryos that can
develop to the blastocyst stage, mainly timing parameters
from later stages (such as the morula, blastocyst, and blasto-
cyst expansion stages) that are still under investigation and
could potentially provide additional or complementary
time-lapse marker candidates. Therefore, the facts presented
in this manuscript encourage the inclusion of kinetic param-
eters into score evaluations, thereby providing a new and
alternative classification algorithm of embryos.
Acknowledgments: The authors thank Nicolas Garrido and
Josep Lluis Romero from IVI Valencia for their clinical sup-
port in this study.
REFERENCES
1. Edwards RG, Fishel SB, Cohen J, Fehilly CB, Purdy JM, Slater JM, et al. Factors
influencing the success of in vitro fertilization for alleviating human infer-
tility. J In Vitro Fert Embryo Transf 1984;1:3–23.
2. Alpha Scientists in Reproductive Medicine and ESHRE Special Interest Group
of Embryology. The Istanbul consensus workshop on embryo assessment:
proceedings of an expert meeting. Hum Reprod 2011;26:1270–83.
3. Mansour R, Ishihara O, Adamson GD, Dyer S, de Mouzon J, Nygren KG, et al.
International Committee for Monitoring Assisted Reproductive Technolo-
gies world report: assisted reproductive technology 2006. Hum Reprod
2014;29:1536–51.
VOL. 105 NO. 2 / FEBRUARY 2016
Fertility and Sterility®
4. Kupka MS, Ferraretti AP, de Mouzon J, Erb K, D'Hooghe T, Castilla JA, et al.
Assisted reproductive technology in Europe, 2010: results generated from
European registers by ESHRE. Hum Reprod 2014;29:2099–113.
5. Lemmen JG, Agerholm I, Ziebe S. Kinetic markers of human embryo quality
using time-lapse recordings of IVF/ICSI-fertilized oocytes. Reprod Biomed
Online 2008;17:385–91.
6. Cruz M, Gadea B, Garrido N, Pedersen KS, Martinez M, Perez-Cano I, et al.
Embryo quality, blastocyst and ongoing pregnancy rates in oocyte donation
patients whose embryos were monitored by time-lapse imaging. J Assist Re-
prod Genet 2011;28:569–73.
7. Kirkegaard K, Agerholm IE, Ingerslev HJ. Time-lapse monitoring as a tool for
clinical embryo assessment. Hum Reprod 2012;27:1277–85.
8. Meseguer M, Rubio I, Cruz M, Basile N, Marcos J, Requena A. Embryo incu-
bation and selection in a time-lapse monitoring system improves pregnancy
outcome compared with a standard incubator: a retrospective cohort study.
Fertil Steril 2012;98:1481–9.e10.
9. Nakahara T, Iwase A, Goto M, Harata T, Suzuki M, Ienaga M, et al. Evalua-
tion of the safety of time-lapse observations for human embryos. J Assist Re-
prod Genet 2010;27:93–6.
10. Payne D, Flaherty SP, Barry MF, Matthews CD. Preliminary observations on
polar body extrusion and pronuclear formation in human oocytes using
time-lapse video cinematography. Hum Reprod 1997;12:532–41.
11. Iwata K, Yumoto K, Sugishima M, Mizoguchi C, Kai Y, Iba Y, et al. Analysis of
compaction initiation in human embryos by using time-lapse cinematog-
raphy. J Assist Reprod Genet 2014;31:421–6.
12. Aguilar J, Motato Y, Escriba MJ, Ojeda M, Munoz E, Meseguer M. The hu-
man first cell cycle: impact on implantation. Reprod Biomed Online 2014;
28:475–84.
13. Meseguer M, Herrero J, Tejera A, Hilligsoe KM, Ramsing NB, Remohi J. The
use of morphokinetics as a predictor of embryo implantation. Hum Reprod
2011;26:2658–71.
14. Cruz M, Garrido N, Herrero J, Perez-Cano I, Munoz M, Meseguer M. Timing
of cell division in human cleavage-stage embryos is linked with blastocyst
formation and quality. Reprod Biomed Online 2012;25:371–81.
15. Conaghan J, Chen AA, Willman SP, Ivani K, Chenette PE, Boostanfar R, et al.
Improving embryo selection using a computer-automated time-lapse image
analysis test plus day 3 morphology: results from a prospective multicenter
trial. Fertil Steril 2013;100:412–9.e5.
16. Hashimoto S, Kato N, Saeki K, Morimoto Y. Selection of high-potential em-
bryos by culture in poly(dimethylsiloxane) microwells and time-lapse imag-
ing. Fertil Steril 2012;97:332–7.
17. Kirkegaard K, Kesmodel US, Hindkjaer JJ, Ingerslev HJ. Time-lapse parame-
ters as predictors of blastocyst development and pregnancy outcome in em-
bryos from good prognosis patients: a prospective cohort study. Hum
Reprod 2013;28:2643–51.
18. Cetinkaya M, Pirkevi C, Yelke H, Colakoglu YK, Atayurt Z, Kahraman S. Rela-
tive kinetic expressions defining cleavage synchronicity are better predictors
of blastocyst formation and quality than absolute time points. J Assist Re-
prod Genet 2015;32:27–35.
19. Wong CC, Loewke KE, Bossert NL, Behr B, De Jonge CJ, Baer TM, et al.
Non-invasive imaging of human embryos before embryonic genome acti-
vation predicts development to the blastocyst stage. Nat Biotechnol 2010;
28:1115–21.
20. Campbell A, Fishel S, Bowman N, Duffy S, Sedler M, Hickman CF. Modelling
a risk classification of aneuploidy in human embryos using non-invasive mor-
phokinetics. Reprod Biomed Online 2013;26:477–85.
21. Campbell A, Fishel S, Bowman N, Duffy S, Sedler M, Thornton S. Retro-
spective analysis of outcomes after IVF using an aneuploidy risk model
derived from time-lapse imaging without PGS. Reprod Biomed Online
2013;27:140–6.
22. Chavez SL, Loewke KE, Han J, Moussavi F, Colls P, Munne S, et al. Dynamic
blastomere behaviour reflects human embryo ploidy by the four-cell stage.
Nat Commun 2012;3:1251.
23. Basile N, Nogales Mdel C, Bronet F, Florensa M, Riqueiros M, Rodrigo L, et al.
Increasing the probability of selecting chromosomally normal embryos by
time-lapse morphokinetics analysis. Fertil Steril 2014;101:699–704.
24. Montag M. Morphokinetics and embryo aneuploidy: has time come or not
yet? Reprod Biomed Online 2013;26:528–30.
383
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
25. Garrido N, Zuzuarregui JL, Meseguer M, Simon C, Remohi J, Pellicer A.
Sperm and oocyte donor selection and management: experience of a
10 year follow-up of more than 2100 candidates. Hum Reprod 2002;
17:3142–8.
26. Munoz M, Cruz M, Humaidan P, Garrido N, Perez-Cano I, Meseguer M.
Dose of recombinant FSH and oestradiol concentration on day of HCG affect
embryo development kinetics. Reprod Biomed Online 2012;25:382–9.
27. Li R, Pedersen KS, Liu Y, Pedersen HS, Laegdsmand M, Rickelt LF, et al. Effect
of red light on the development and quality of mammalian embryos. J Assist
Reprod Genet 2014;31:795–801.
28. Alikani M, Calderon G, Tomkin G, Garrisi J, Kokot M, Cohen J. Cleavage
anomalies in early human embryos and survival after prolonged culture in-
vitro. Hum Reprod 2000;15:2634–43.
29. de los Santos MJ, Arroyo G, Busquet A, Calderon G, Cuadros J, Hurtado de
Mendoza MV, et al. A multicenter prospective study to assess the effect of
early cleavage on embryo quality, implantation, and live-birth rate. Fertil
Steril 2014;101:981–7.
30. Cobo A, Meseguer M, Remohi J, Pellicer A. Use of cryo-banked oocytes in an
ovum donation programme: a prospective, randomized, controlled, clinical
trial. Hum Reprod 2010;25:2239–46.
31. Walker MC, Murphy KE, Pan S, Yang Q, Wen SW. Adverse maternal out-
comes in multifetal pregnancies. BJOG 2004;111:1294–6.
32. Ferraretti AP, Goossens V, de Mouzon J, Bhattacharya S, Castilla JA,
Korsak V, et al. Assisted reproductive technology in Europe, 2008: re-
sults generated from European registers by ESHRE. Hum Reprod 2012;
27:2571–84.
33. Herrero J, Meseguer M. Selection of high potential embryos using time-lapse
imaging: the era of morphokinetics. Fertil Steril 2013;99:1030–4.
34. Fenwick J, Platteau P, Murdoch AP, Herbert M. Time from insemination to
first cleavage predicts developmental competence of human preimplanta-
tion embryos in vitro. Hum Reprod 2002;17:407–12.
35. Nagy ZP, Liu J, Joris H, Devroey P, Van Steirteghem A. Time-course of oocyte
activation, pronucleus formation and cleavage in human oocytes fertilized
by intracytoplasmic sperm injection. Hum Reprod 1994;9:1743–8.
36. Wong C, Chen AA, Behr B, Shen S. Time-lapse microscopy and image anal-
ysis in basic and clinical embryo development research. Reprod Biomed On-
line 2013;26:120–9.
384
37. Graham J, Han T, Porter R, Levy M, Stillman R, Tucker MJ. Day 3 morphology
is a poor predictor of blastocyst quality in extended culture. Fertil Steril 2000;
74:495–7.
38. Scholtes MC, Zeilmaker GH. A prospective, randomized study of embryo
transfer results after 3 or 5 days of embryo culture in in vitro fertilization. Fer-
til Steril 1996;65:1245–8.
39. Herrero J, Tejera A, Albert C, Vidal C, de los Santos MJ, Meseguer M. A time
to look back: analysis of morphokinetic characteristics of human embryo
development. Fertil Steril 2013;100:1602–9.e1-4.
40. van Montfoort AP, Hanssen LL, de Sutter P, Viville S, Geraedts JP, de Boer P.
Assisted reproduction treatment and epigenetic inheritance. Hum Reprod
Update 2012;18:171–97.
41. Dumoulin JC, Land JA, Van Montfoort AP, Nelissen EC, Coonen E,
Derhaag JG, et al. Effect of in vitro culture of human embryos on birthweight
of newborns. Hum Reprod 2010;25:605–12.
42. Gardner DK, Vella P, Lane M, Wagley L, Schlenker T, Schoolcraft WB. Cul-
ture and transfer of human blastocysts increases implantation rates and re-
duces the need for multiple embryo transfers. Fertil Steril 1998;69:84–8.
43. Blake DA, Farquhar CM, Johnson N, Proctor M. Cleavage stage versus blas-
tocyst stage embryo transfer in assisted conception. Cochrane Database Syst
Rev 2007:CD002118.
44. Mercader A, Garcia-Velasco JA, Escudero E, Remohi J, Pellicer A, Simon C.
Clinical experience and perinatal outcome of blastocyst transfer after cocul-
ture of human embryos with human endometrial epithelial cells: a 5-year
follow-up study. Fertil Steril 2003;80:1162–8.
45. Kirkegaard K, Campbell A, Agerholm I, Bentin-Ley U, Gabrielsen A, Kirk J,
et al. Limitations of a time-lapse blastocyst prediction model: a large multi-
centre outcome analysis. Reprod Biomed Online 2014;29:156–8.
46. Best I, Campbell A, Duffy S, Montogomery S, Fishel S. Does one model fit all?
Testing a published embryo selection algorithm on independent time-lapse
data. Hum Reprod 2013;28(i87–i90).
47. Alikani M. The originals and consequences of fragmentation in mammalian
eggs and embryos. In: Preimplatation embryo evaluation and selection, Elder
and Cohen, editors. 2007:89–100.
48. Mio Y, Maeda K. Time-lapse cinematography of dynamic changes occurring
during in vitro development of human embryos. Am J Obstet Gynecol 2008;
199:660.e1–5.
VOL. 105 NO. 2 / FEBRUARY 2016
 Fertility and Sterility®

SUPPLEMENTAL FIGURE 1

Flowchart for phases accomplished in this study to develop the model.

Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

VOL. 105 NO. 2 / FEBRUARY 2016 384.e1

ORIGINAL ARTICLE: ASSISTED REPRODUCTION

SUPPLEMENTAL FIGURE 2

Blastocyst morphology classification categories based on trophectoderm, inner cell mass, and grade of expansion.
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

384.e2 VOL. 105 NO. 2 / FEBRUARY 2016

 Fertility and Sterility®

SUPPLEMENTAL FIGURE 3

Graphic description of morphokinetic variables discerned by time-

lapse analysis.
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

VOL. 105 NO. 2 / FEBRUARY 2016 384.e3

ORIGINAL ARTICLE: ASSISTED REPRODUCTION

SUPPLEMENTAL FIGURE 4

Hierarchical classification of embryos based on morphological screening; the new morphological criteria and timing of cell divisions based on
optimal range for t5 or timing of cell division to 5 cells (48.8–56.6 hours); optimal range for second synchrony t4–t3 (%0.76 hours); and
optimal range for t3–t2 or duration of second cell cycle (%11.9 hours). This classification was applied for embryo selection as described by
Meseguer et al. (13).
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

384.e4 VOL. 105 NO. 2 / FEBRUARY 2016

 Fertility and Sterility®

SUPPLEMENTAL TABLE 1

Mean timings from embryos analyzed by time lapse.

Median time-lapse observations
Morphokinetic variable Blastocyst stage (3,215) Non–blastocyst stage No. of nonblastocysts Total P value
t2 27.23  0.2 28.78  0.2 4,268 7,483 < .05
t3 37.77  0.2 38.46  0.2 4,191 7,406 < .05
t4 39.54  0.2 41.5  0.2 3,997 7,212 < .05
t5 51.37  0.2 51.13  0.3 3,941 7,156 .318
t6 54.15  0.3 55.22  0.3 3,568 6,783 < .05
t7 57.07  0.3 59.34  0.4 3,417 6,632 < .05
t8 60.96  0.4 63.07  0.5 1,948 5,193 < .05
t9 71.68  0.4 72.14  0.8 278 3,493 .286
tM 87.60  0.3 91.79  0.9 112 3,327 < .05
t3–t2 10.54  0.1 9.82  0.2 4,191 7,406 < .05
t5–t3 13.62  0.2 13.41  0.3 3,941 7,156 .219
t4–t3 1.78  0.1 3.30  0.2 3,997 7,212 < .05
t8–t5 10.27  0.3 13.73  0.4 1,948 5,193 < .05
t5–t2 24.15  0.3 22.97  0.3 3,941 7,156 < .05
Note: Values are mean  SD time in hours.
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

VOL. 105 NO. 2 / FEBRUARY 2016 384.e5

ORIGINAL ARTICLE: ASSISTED REPRODUCTION

SUPPLEMENTAL TABLE 2

Exact timing of events according to quartiles (Q1, Q2, Q3, and Q4) and the numbers and proportions of embryos reaching the blastocyst stage.
Q1 Q2 Q3 Q4
Cell cycle stage Limit % Blastocyst (n) Limit % Blastocyst (n) Limit % Blastocyst (n) Limit % Blastocyst (n)
t2 %24.56 51.9 (976) 24.57–26.66 48.3 (902) 26.67–29.48 43.4 (812) >29.49 28.1 (525)
t3 %34.26 38.7 (711) 34.27–37.61 55.20 (1,015) 37.62–41.22 48.1 (882) >41.23 33.1 (607)
t4 %36.32 48.3 (873) 36.33–39.45 53.2 (956) 39.46–43.38 47.5 (857) >43.39 29.3 (528)
t5 %45.28 31.7 (591) 45.29–50.88 51.8 (975) 50.89–56.29 49.8 (938) >56.30 38.2 (711)
t6 %49.08 40.0 (752) 49.09–53.74 49.5 (929) 53.75–59.06 47.6 (882) >59.07 34.7 (652)
t7 %51.60 45.5 (852) 51.61–56.59 46.9 (878) 56.60–62.83 44.8 (837) >62.84 34.7 (648)
t8 %53.98 46.1 (863) 53.99–59.83 44.3 (829) 59.84–67.20 41.0 (767) >67.21 40.4 (756)
t9 %63.60 38.9 (728) 63.61–70.26 43.1 (807) 70.27–79.01 49.2 (921) >79.02 40.6 (759)
tM %81.27 44.2 (827) 81.28–88.95 44.9 (771) 88.96–96.00 47.0 (951) >96.01 35.6 (666)
t3–t2 %9.75 31.3 (575) 9.76–11.51 55.7 (1,045) 11.52–12.72 54.1 (972) >12.73 33.9 (623)
t5–t3 %11.34 35.5 (656) 11.35–13.90 50.2 (924) 13.91–16.49 49.8 (932) >16.50 37.6 (703)
t4–t3 %0.25 48.8 (963) 0.26–0.89 47.7 (779) 0.90–2.33 48.9 (913) >2.34 32.2 (559)
t8–t5 %3.75 50.1 (943) 3.76–8.78 46.8 (871) 8.79–18.00 38.3 (716) >18.01 36.6 (685)
t5–t2 %17.34 28.7 (537) 17.35–24.81 50.1 (936) 24.82–28.40 52.8 (989) >28.41 40.3 (753)
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

384.e6 VOL. 105 NO. 2 / FEBRUARY 2016

 Fertility and Sterility®

SUPPLEMENTAL TABLE 3

Distribution of blastocyst according to morphological categories

(ASEBIR).
Blastocyst
Category Total formations % (95% CI)
A 168 93 55.4 (46.9–64.0)
B 386 180 46.6 (41.0–51.2)
C 245 98 39.9 (32.8–46.7)
D 33 12 35.6 (16.0–54.2)
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

VOL. 105 NO. 2 / FEBRUARY 2016 384.e7

ORIGINAL ARTICLE: ASSISTED REPRODUCTION

SUPPLEMENTAL TABLE 4

Distribution of implanted blastocyst according to morphokinetic

categories (validation phase).
Model effect OR P value
Blastocyst morphology (ASEBIR)
A vs. D 1.69 (0.69–4.15) .249
B vs. D 1.30 (0.56–1.86) .542
C vs. D 1.06 (0.45–2.50) .899
Morphokinetic model
A vs. D 2.33 (1.38–3.92) .002
B vs. D 1.88 (1.19–2.96) .007
C vs. D 1.44 (0.98–2.11) .062
Oocyte source
Donation vs. autologous 1.316 (0.952–1.821) .097
Oocyte age, y 0.94 (0.91–0.98) .002
Note: Logistic regression analysis of ongoing implantation (gestational sac) referring to those
blastocysts with known implantation (KID embryos) as affected by the combination of
morphological blastocyst categories (ASEBIR) and the blastocyst categories of the hierarchical
classification tree model applied in the TLM system; oocyte source and the age of the patient
or the oocyte donor are confounding factors. OR is the odds ratio, with 95% confidence in-
tervals in brackets.
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

384.e8 VOL. 105 NO. 2 / FEBRUARY 2016

 Fertility and Sterility®

SUPPLEMENTAL TABLE 5

Logistic regression analysis on implantation potential of transferred

embryos.
Blastocyst
Category Total implantations % (95% CI)
A 11 10 90.9 (88.39–93.41)
B 18 12 66.7 (62.59–70.81)
C 44 24 54.4 (50.05–58.75)
D 184 77 41.8 (37.49–46.11)
Note: Implantation in the blastocyst categories of the hierarchical classification tree model
applied in the TLM system; results only refer to those blastocysts with known implantation
(KID embryos).
Motato. Blastocyst prediction by time lapse. Fertil Steril 2016.

VOL. 105 NO. 2 / FEBRUARY 2016 384.e9