Open Access
Add to BibSonomy
Abstract
Female infertilities are highly associated with poor endometrial receptivity. A receptive endometrium is generally characterized by the normal uterine cavity, intact endometrial surface, appropriate endometrial thickness, and echo pattern. Acquiring comprehensive structural information is the prerequisite of endometrium assessment, which is beyond the ability of any single-modality imaging method. In this paper, we introduce a custom-made intrauterine dual-modality (OCT/ultrasound) endoscopic imaging system and achieve in vivo imaging of rabbit uteri, for the first time to our knowledge. The endometrial features of the injured uteri in both ultrasonic and OCT images are consistent with their corresponding pathology. The quantified parameters, including uterine thickness and endometrial surface roughness, show the correlation with the endometrial injury degree but with poor performance for injury classification. The combination of these parameters was proved to assess the degrees of endometrial injury more accurately. Our work shows the potential of the dual-modality system to be translated into a clinical tool, providing multiple quantitative imaging information and helping evaluate the endometrial receptivity and diagnose infertility.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Nowadays, a prevalence of 9% to 18% of the general population suffers from infertility worldwide [1]. Successful fertilization depends on a functional blastocyst, a receptive endometrium, and the interaction between the blastocyst and endometrial epithelium [2,3]. Besides embryo quality improvement, developing new technologies for the accurate evaluation of endometrium receptivity is of significance and has become what the researchers and physicians pursue [4,5]. Endometrial receptivity is defined as the ability that makes the uterus receptive to blastocyst attachment and implantation [6,7]. Impaired endometrial receptivity and altered endometrium-embryo dialogue are responsible for 67% of implantation failures and have been commonly recognized as the bottleneck of the reproductive process [8]. Endometrial receptivity is decided by complex factors, both morphological and pathological [9,10]. Current researchers regard that the acquirement of comprehensive morphological information, including uterine cavity abnormalities [11], endometrial thickness [12–14], endometrial echo pattern [12–14], and endometrial surface [15] is the prerequisite of endometrial receptivity evaluation, thus great efforts have been made to acquire the morphological information using different imaging modalities.
In clinics, although endometrial biopsy is used as a “golden standard” to assess receptivity in cell and molecule levels by invasive sampling procedures [16–18], endoscopic imaging technologies are generally performed first due to the wide imaging field and noninvasiveness to the patients. Hysteroscopy allows direct visualization of the whole uterine cavity and endometrial surface [19]. It can exclude uterine cavity abnormalities that might impede embryo implantation, such as anatomical malformations, intrauterine adhesions, and polyps [20,21]. However, it lacks the penetration ability to access the depth information of the endometrium [22]. Transvaginal ultrasonography can provide cross-sectional images of uteri and structural parameters to assess endometrial receptivity, such as endometrial thickness and endometrial echo pattern [23–25]. But the resolution and contrast of ultrasonography are insufficient to resolve the surface detail information for a more specific evaluation [26,27]. Due to the limitation of the insufficient resolution, sensitivity, or penetration depth, current clinical imaging methods cannot fully satisfy the need of evaluating endometrial receptivity.
Besides the clinically used intrauterine imaging modalities mentioned above, several endoscopic methods have been reported to directly visualize the intrauterine targets and provide surface and depth information with high resolution, high sensitivity, and sufficient penetration, such as endoscopic ultrasound, optical coherence tomography (OCT), photoacoustic imaging and polarimetry. A portable colposcope Mueller matrix polarimeter was demonstrated to provide useful information about the collagen structure in uterine cervical tissues [28]. Due to the ability of micro-scale imaging, OCT has been demonstrated to reconstruct the 3D ultrastructure of human uterine tissue [29,30]. Previous research indicated that endometrial OCT intensity is a sensitive marker to differentiate endometrium with repeated implantation failure from that with other conditions in an in vivo trial [31]. In addition, a polarization-sensitive OCT [32] can combine OCT and polarimeter as a dual-modality system and provide both functional and structural information. However, the penetration depth of intrauterine OCT and polarimeter cannot cover the whole thickness of the endometrium. An intrauterine ultrasonic endoscopic probe was demonstrated to achieve deep-tissue imaging of a pig uterine cavity ex vivo, which has the potential of measuring the total thickness and visualizing anatomical abnormalities of the uterus [33]. Previously, our group achieved dual-modality photoacoustic/ultrasonic imaging of rat rectum in vivo [34]. However, the catheter is oversize for imaging the uterus of small animals. And the intrauterine endoscopic ultrasound does not have sufficient resolution to resolve the detailed surface morphology. The need of acquiring both surface and depth information to make a comprehensive assessment of the endometrium receptivity was not fully satisfied by current imaging methods. The combining of OCT and ultrasound modality has the potential to bring new opportunities for the evaluation of the endometrium.
In this paper, we present an intrauterine dual-modality OCT/ultrasound endoscopic system and achieve endoscopic imaging of endometrial injury in rabbit models in vivo. The acquired OCT and ultrasound images provide surface and depth information simultaneously and show the uterine features associated with the evaluation of the endometrial injury, including intrauterine abnormalities, surface injury, uterine thickness, and echo pattern. In addition, a 2D scatter plot of quantified parameters acquired from these images is proved to classify the degree of endometrial injury. Our works on small animals show the potential of translation into a clinical tool for evaluating endometrial receptivity.
2. Materials and methods
2.1 Construction of endometrial injury models
All procedures involving experimental animals were carried out in accordance with the protocols approved by the animal study committee of Shenzhen institute of advanced technology, Chinese academy of sciences. Four healthy female New Zealand rabbits (4-4.5 kg) were purchased from Kangda biological Co., Ltd (Qingdao, China). All animals were maintained under a 12/12 light/dark cycle at 21-24°C with a relative humidity of 40-60% and were fed with rabbit chow and water ad libitum.
Based on studies of ethanol-induced cell apoptosis [35,36], 95% ethanol was injected into the uterine horns to cause injury to the endometrium during general anesthesia. The degree of endometrial injury is dependent on the retention time of ethanol in the uterine horns. In order to establish the endometrial injury rabbit model, 95% ethanol was injected into the left uterine horn of rabbits for 5 minutes or 10 minutes and then extracted out and slowly rinsed with saline to remove residual ethanol, while the contralateral uterine horn was injected with the equal volume of saline for the same time. The 5-minute models and 10-minute models were allocated to the EtOH-5-min group and the EtOH-10-min group of endometrial injury respectively, which represent different degrees of endometrial injury [36]. Their contralateral uterine horns were allocated to the control group. All the left uterine horns were randomized divided into two groups, and each group had two horns. Therefore, the EtOH-5-min group and the EtOH-10-min group have two ethanol-injured uterine horns and the control group has four healthy horns. The OCT/ultrasound imaging of the rabbits was performed 20-30 days post-surgery. Nine B-scans in each group for each modality were randomly selected for further imaging processing.
2.2 Histopathology preparation
Rabbits were sacrificed after imaging, then the uteri were harvested, flushed, and fixed in 4% paraformaldehyde solution for further processing. Tissue samples were dehydrated, made transparent, dipped in wax, embedded in paraffin blocks, cut into sections with a thickness of 4 µm, and then stained by hematoxylin and eosin (H&E). 10-20 uterine sections have been cut in each uterus. The endometrial morphology was observed, and the images were captured by using a research slide scanner (Olympus VS200, Japan).
2.3 Imaging system and procedure
The OCT/ultrasound endoscopic system used in this study was developed from our previous custom-built prototype [37]. The design of the dual-modality catheter is similar to the previous work, but a single-mode fiber (SMF28) was used to transmit optical beams instead of a double cladding fiber. As shown in Fig. 1(B), the ultrasonic transducer, optic prism, GRIN lens, and single-mode fiber were placed in sequence and aligned coaxially under a stereomicroscope, and then fixed with epoxy in a stainless-steel housing. A protective sheath with optical and acoustic transparency was used for encapsulation of the catheter to protect both the catheter and uterus. The outer diameter of the catheter and its proximal sheath is 1.2 mm. In addition, a custom-designed metal coil (Tu’s Cheng Fa, China) was used in the catheter design to deliver the torque during rotational scanning.
Fig. 1. (A) Schematic of the intrauterine dual-modality endoscopic system; (B) the imaging catheter and its proximal sheath. (C) Imaging procedure of a rabbit uterus in vivo. RPU: rotational and pull-backing unit; FC: fiber coupler; LT: fiber-optic light trap; FOC: fiber optical circulator; CM: collimator; RM: reflective mirror; BD: balanced detector.
Before the imaging procedure, the experimental rabbits were anesthetized using a mixture of isoflurane and oxygen (2% isoflurane concentration) at a flow rate of 300 ml/min. Due to the pelvic girdle of the rabbit over the vagina, it is difficult to observe the cervix orifice and let the catheter pass through. Therefore, two incisions with a length of about 3-5 cm were conducted on both the abdominal wall and vagina near the uterine cervix to observe the cervix orifice. Afterward, saline solution was injected into the uterus by using a peristaltic pump. Finally, the imaging catheter was inserted into the uterine horns through the cervix orifice to perform a rotational scanning of 10 frames per second and a pull-backing with the step of 50 µm. In both OCT and ultrasound modalities, 300 B-scans were acquired individually in a single pull-backing. Each OCT B-scan occupies 996 A-lines and each ultrasound B-scan occupies 1000 A-lines. Two data acquisition (DAQ) cards (ATS9325, Alazar Tech, Canada) were used for ultrasound and OCT signal data acquisition separately. The data acquisition was performed in LabVIEW. Both ultrasound and OCT images were post-processed by Matlab from the raw recorded data.
2.4 Image processing
In this study, quantified parameters, including uterine thickness (UT), uterine thickness distribution (UTD), and endometrial surface roughness (ESR), were acquired from the B-scan images (nine B-scans for each group) in ultrasonic and OCT modalities separately. Then, they were used to evaluate the degree of endometrial injury. In the ultrasonic B-scan images, the inner and outer surfaces of the uterus were manually marked by two radiologists, based on the boundary between the high echo region and low echo region. The UT of each B-scan image was regarded as the mean value of the minimum distance from each pixel at the inner surface to the outer surface. The uterine thickness distribution (UTD) of each B-scan was defined as the percentage of uterine thickness that is less than a threshold value. The threshold value ${T_{th}}$ is defined to cover 95% of the control group, which is calculated by the mean value ${M_c}$ of UT in the control group minus two folds of its standard deviation ${\sigma _c}$.
The overall process of calculating ESR consists of three steps [38]: (1) image segmentation; (2) surface detection and surface flattening; (3) surface roughness quantification. In the first step, a Gaussian filter and a median filter were applied to OCT images in the polar coordinates to diminish the effect of speckle noise. Then, binarization, Sobel edge detection, and image dilation were applied to the OCT images to segment the endometrium. In the second step, the endometrial surface curvature was extracted by another Sobel edge detection process and subtracted by itself after a Gaussian filter to produce the flattened endometrial surface curvature. Roughness characterizes irregularities on surfaces. It is quantified by the deviations in the direction of the normal vector of a real surface from its ideal form [38]. Thus, in the last step, ESR can be defined as the standard deviation of the height of the flattened endometrial surface.
In order to classify different groups of endometrial injury, a 2D scatter plot was developed by using ESR as X-axis and UT(UTD) as Y-axis. The grouping boundary is defined as a confidence ellipse [39] with the equation as follow:
In the pathological results, the parameters UT and ESR of each uterine section were also calculated. The outer surface is manually marked by the two radiologists, and the inner surface was detected by the same method in OCT image processing.
3. Results
Figure 2(A) and (C) show the cross-sectional OCT and ultrasound images of the rabbit uterus from the healthy group in Cartesian coordinates, respectively. The OCT image provides detailed surface information of the endometrium, while the ultrasound image provides the depth information of the uterus. With rotation and pull-backing of the scanning catheter, longitudinal OCT and ultrasound images in Fig. 2(B) and (D) can be reconstructed. The red lines indicated the position of B-scans in Fig. 2(A) and (C). The combination of cross-sectional and longitudinal views ensures clinicians a better 3D visualization of uterine volume for accurate diagnosis. In addition, Fig. 2(E) presented the unwrapped 3D endometrial surface with polyp-like structure (blue arrows) in polar coordinates, which was reconstructed from cross-sectional OCT images. The unwrapped image of the 3D endometrial surface provides clinicians a direct visualization of morphological information, which allows for diagnosing any damages on the endometrial surface.
Fig. 2. Image results of rabbit uterus from the healthy group in OCT and ultrasound modalities. (A) cross-sectional and (B) longitudinal OCT images of rabbit uterus. (C) cross-sectional and (D) longitudinal ultrasound images of rabbit uterus. The red lines indicated the position of B-scans in A and C. (E) unwrapped 3D OCT images of the uterus. Blue arrows: polyp-like structure; Scale bar: 1 mm.
Figure 3 showed the intrauterine OCT images of healthy endometrium and ethanol-injured endometrium and their corresponding pathological images. In Fig. 3(A) and (C), the healthy, continuous endometrium was presented with polyp-like protrusions and the uterine cavity was stellate. The red arrow in the enlarged OCT image Fig. 3(B) showed a smooth surface of the endometrium, indicating the integrated columnar epithelial cells lining on the endometrial surface in Fig. 3(D) (red arrow). In Fig. 3(E) and (I), the endometrial protrusions become discontinuous, atrophic, and the uterine cavity becomes hollower. In Fig. 3(F), the enlarged OCT images revealed that the endometrial surface was rough. In Fig. 3(J), the polyp-like protrusions were disrupted and reduced to about 100-200 µm. These features can be verified by the corresponding pathological images in Fig. 3(H) and (L). In addition, Fig. 3(F) showed dilated cavities (blue arrows) within the myometrium, indicating the depth-resolved ability of the OCT modality up to 1 mm.
Fig. 3. Uteri in OCT modality and their pathology. (A) an OCT image of a healthy uterus and its enlarged image (B) with a smooth surface (red arrows). (C) the corresponding pathological section image and its enlarged image (D), with the integrated and continuous surface (red arrows); (E) ethanol injured uterus and its enlarged image (F), with a rough endometrial surface (red arrows) and hollow cavities beneath endometrium (blue arrows). (G) the corresponding pathological image and its enlarged image (H), with the broken epithelial cells (red arrows) and myometrial diverticulum; (I) ethanol injured endometrium and its enlarged image (J), with an atrophic endometrial surface (red arrows); (K) the corresponding pathological section image and its enlarged image (L), with the atrophic endometrium (red arrows). Scale bar: 1 mm.
The ultrasound results and their corresponding pathological images were presented in Fig. 4. Comparing Fig. 4(A) and (C) (red arrows), the uterine thickness of the injured group in ultrasound modality is significantly thinner than that of the healthy group, which is consistent with the pathological results in Fig. 4(B) and (D). Meanwhile, as shown in Fig. 4(A) and (C), the boundary of the endometrium and myometrium (green dash) is clearer and the ultrasound signal intensity (blue arrow) of the endometrial surface is higher in the image of the injured endometrium. This feature is similar to the endometrial echo patterns in transvaginal ultrasonography, which indicated the acoustic impedance change of endometrial tissues.
Fig. 4. Uteri in ultrasound modality and their pathology. (A) ultrasound image of a healthy uterus and its corresponding pathological section (B); (C) ultrasound image of the injured uterus and its corresponding pathological section (D); Several features were shown in the ultrasound images including the thickness degradation (red arrows), the reduction in ultrasound intensity (blue arrows), and the clearer boundary (green dash) between the endometrium and myometrium. Scale bar: 1 mm.
Figure 5 showed the OCT and ultrasound image results of the uteri with different degrees of injury (the control group, the EtOH-5-min group, and the EtOH-10-min group), and their corresponding pathological results. The endometrial surface curvature in the OCT images was marked in the form of white lines, as shown in Fig. 5(A), (D), and G. In Fig. 5(B), (E), and H, the inner and outer surfaces of the uterus in the ultrasound modality were marked in the form of red dashes. Then, the value of ESR and UT can be calculated from the white lines and red dashes respectively. According to Fig. 5, the ultrasound and OCT images suggested that the uterus becomes thinner, and the surface becomes rougher with the increasing degree of endometrial injury. The corresponding pathological results in Fig. 5(C), (F), and I present the same changes in UT and ESR of different groups.
Fig. 5. Imaging results of the control group, EtOH-5-min group, and EtOH-10-min group (from top to bottom). (A), (D), (G) OCT images with the borderline of endometrial surface (white line). (B), (E), (H) ultrasound images with the borderline of endometrial surface and uterine outline (red dashes). (C), (F), (I) pathological sections with the borderline of endometrial surface and uterine outline (yellow line). Scale bar: 1 mm.
Statistical results of the control group, the EtOH-5-min group, and the EtOH-10-min group for OCT, ultrasound, and pathological images were shown in Fig. 6 (detailed data in supplemental document Table S2). As shown in Fig. 6(A), data from OCT and pathological results indicated that ESR in ethanol-induced groups was significantly increased compared with that in the control group and the degree of ESR is positively correlated with ethanol treatment. As shown in Fig. 6(B), UT in ethanol-induced groups was significantly reduced compared with that in the control group and the value of UT is negatively correlated with ethanol treatment. Due to the large standard deviation of ESR and UT, these parameters cannot identify the degree of endometrial injury individually. Figure 6(C) demonstrated that 2D scatter plot of both ESR and UT parameters from OCT/ultrasound can be introduced to classify the normal uterus and different degrees of the ethanol-injured uterus. However, as shown in Fig. 6(C), there are overlapping areas between different groups in the 2D scatter plot, which bring difficulties to the endometrial injury classification. Therefore, the UTD was introduced, indicating the thickness distribution for each uterine image, where the threshold is set as 1.91 mm according to the method in the previous section. Figure 6(D) showed 2D scatter plot of ESR and UTD parameters, which leads to a separation of the 95% confidence ellipse for uterine groups with different endometrial injury degrees. Thus, the combination of OCT and ultrasound modalities presents a more accurate injury classification than either OCT or ultrasound individually.
Fig. 6. Statistical results of different uterus groups (n = 9 for each group). (A) endometrial surface roughness (ESR) for the control group, EtOH-5-min group, and EtOH-10-min group in OCT images and pathological sections. (B) Uterine thickness (UT) for ultrasound images and pathological sections. (C) UT vs ESR scatter plot of different groups in OCT and ultrasound modality. (D) Uterine thickness distribution (UTD) vs ESR scatter plot of different groups in OCT and ultrasound modalities.
4. Discussion
The endometrium receptivity is critical for successful embryo implantations and is decided by a variety of characteristics, such as the uterine cavity, endometrial surface, the thickness of the endometrium, endometrial echo pattern, etc. The accurate evaluation of endometrium receptivity involves the acquirement of the above-mentioned multiple characteristics, which encourages the combination of complementary endoscopic imaging methods, such as intrauterine ultrasonic imaging and OCT.
In this study, we presented an intrauterine OCT/ultrasound endoscopic system to assess the rabbit uterus. To the best of our knowledge, it is the first intrauterine endoscopic imaging for small animals in vivo and achieves two imaging modalities simultaneously. Due to the large size of the catheter, previous intrauterine endoscopic imaging studies focused on humans and pigs, rather than small animals, which is more suitable for modeling. A home-made catheter with an outer diameter of 1.2 mm integrated the small-sized optical and ultrasonic elements and offered OCT and ultrasound images, simultaneously. Although the replacement of DCF with SMF lead to the possible incompatibility with other modality, it improves the imaging performance of OCT modality (see supplemental document Fig. S1 and Table S1). A metal coil was applied to assure the catheter entering into the uterine cavity flexibly and transmit rotational torque smoothly from the rotation and pull-back unit (RPU) to the catheter tip. We finally achieved 3D scanning of the uterus in vivo by using the metal coil and RPU. In addition, due to the small size of the dual-modality imaging catheter, the in vivo imaging experiment of the uterus was performed non-invasively to the endometrium, which allowed to repeat examinations of female genital tracts and guide the endometrial biopsy.
By imaging the rabbit models, the endometrial changes between uteri with and without endometrial injury can be recognized in both OCT and ultrasound modalities. In OCT images, the surface details of the healthy uteri appeared as continuous and smooth, while the surface of ethanol-induced uteri became discontinuous. In ultrasound images, it can be observed that the thickness of ethanol-induced uteri is significantly degraded compared to the healthy uteri. Several quantified parameters, including UT, UTD, and ESR, were concluded from the dual-modality images. In the ethanol-induced endometrium models, the statistical results of UT showed an overall negative correlation with the degree of endometrial injury and the ESR showed an overall positive correlation. However, large standard deviation always existed in the statistical results of the parameter acquired from each single imaging modality, leading to the overlapping between confidence intervals of different groups and bringing great difficulties to distinguish the degrees individually. Thus, it is not accurate to classify the endometrial injury degree by either OCT or ultrasound imaging results. In this work, we developed a 2D classification method based on parameters (UTD and ESR) in two imaging modalities. The 2D scatter plot showed superior performance than that using a single parameter in classifying the endometrium with different injury degrees, which proved the significance of combining OCT and ultrasound modalities in this work.
Additionally, the dual-modality imaging results also presented other prominent features that may contribute to the endometrial receptivity evaluation. The ultrasound results of the ethanol-induced group showed a hyperechoic endometrium as well as a well-defined hypoechoic line between the endometrium and the myometrium, while the endometrium in the healthy group had a relatively lower echo and the echogenic line was poorly defined. These endometrial echo patterns presented by the endoscopic ultrasound were essentially the same as those in transvaginal ultrasonography with a better resolution, which reflected the physical properties of endometrial tissues and was a good indicator to evaluate endometrial receptivity [40]. In the OCT modality, both 2D cross-sectional and 3D reconstructed images of healthy uteri showed the features of the stellate uterine cavity and the polyp-like endometrial infoldings. The smallest polyp-like endometrial infolding of around 200 µm has been identified in OCT images, demonstrating the potential of our technology to identify endometrial lesions as tiny as 200 µm. In the images of ethanol-induced endometrium, the infoldings become atrophic and the surface became rough, which was consistent with the corresponding pathological results. These features of the endometrium were convincing evidence of endometrial injury and poor endometrial receptivity.
There are still a few limitations in our intrauterine imaging work that need to be improved in the future. Firstly, the whole imaging procedure does not harm the endometrium, but still needs incisions of the abdominal wall and vagina, due to the pelvic girdle. A special speculum for small animals can be designed in the future, helping the catheter pass through the vagina and uterine cervix non-invasively to experimental animals. Secondly, the echogenic line between the endometrium and the myometrium in the control group can be hardly defined. It may be due to the mismatching frequency or insufficient sensitivity of the transducer used in this work. In order for clinical translation, future studies can focus on the selection of transducers with the suitable frequency and sensitivity and optical design with matching imaging depth, resolution and scanning method for intrauterine imaging of different animals and humans. In addition, further pregnancy experiments shall be conducted to prove the capability of our dual-modality system to evaluate endometrial receptivity.
5. Conclusion
In summary, we presented an intrauterine dual-modality endoscopy and achieved in vivo imaging of rabbits for the first time. Both surface and depth information was acquired from rabbit models with endometrial injury simultaneously without harm to the endometrium. Multiple features in both ultrasound and OCT imaging results, which are associated with the endometrial injury and receptivity, showed consistency with their corresponding pathological results. The quantified parameters acquired from ultrasound and OCT images showed a correlation with endometrial injury. Furthermore, by combining ultrasound and OCT modalities, the 2D scatter plot of quantified parameters was developed and applied to classify the degrees of endometrial injury accurately, demonstrating the essential of dual-modality imaging in endometrial assessment. This work paved the way for clinical translation to evaluate endometrial receptivity accurately with comprehensive imaging information using multi-modality imaging technologies.
Funding
National Natural Science Foundation of China (61975226, 82027803); National Key Research and Development Program of China (2018YFC0116302); the Scientific Instruments Funding of CAS (YJKYYQ20190077); CAS Key Laboratory of Health Informatics (2011DP173015); Natural Science Foundation of Guangdong Province (2114050003154); the Guangdong Provincial Key Laboratory of Biomedical Optical Imaging Technology (2020B121201010); Science, Technology and Innovation Commission of Shenzhen Municipality (JCYJ20200109113808048, JCYJ20210324101612034, ZDSY20130401165820357).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. L. Aghajanova, J. Hoffman, E. Mok-Lin, and C. N. Herndon, “Obstetrics and gynecology residency and fertility needs,” Reprod. Sci. 24(3), 428–434 (2017). [CrossRef]
2. H. Achache and A. Revel, “Endometrial receptivity markers, the journey to successful embryo implantation,” Hum. Reprod. Update. 12(6), 731–746 (2006). [CrossRef]
3. H. Wang and S. K. Dey, “Roadmap to embryo implantation: clues from mouse models,” Nat. Rev. Genet. 7(3), 185–199 (2006). [CrossRef]
4. A. Heger, M. Sator, and D. Pietrowski, “Endometrial receptivity and its predictive value for IVF/ICSI-outcome,” Geburtshilfe Frauenheilkd. 72(08), 710–715 (2012). [CrossRef]
5. S. Altmäe and L. Aghajanova, “Growth hormone and endometrial receptivity,” Front. Endocrinol. 10, 653 (2019). [CrossRef]
6. B. A. Lessey, “Assessment of endometrial receptivity,” Fertil. Steril. 96(3), 522–529 (2011). [CrossRef]
7. N. X. Jiang and X. L. Li, “The disorders of endometrial receptivity in PCOS and its mechanisms,” Reprod. Sci. (2021). [CrossRef]
8. L. Craciunas, I. Gallos, J. Chu, T. Bourne, S. Quenby, J. J. Brosens, and A. Coomarasamy, “Conventional and modern markers of endometrial receptivity: a systematic review and meta-analysis,” Hum. Reprod. Update. 25(2), 202–223 (2019). [CrossRef]
9. G. Nikas, “Endometrial receptivity: changes in cell-surface morphology,” Semin Reprod Med 18(03), 229–236 (2000). [CrossRef]
10. H. M. Beier and K. Beier-Hellwig, “Molecular and cellular aspects of endometrial receptivity,” Human Reproduction Update 4(5), 448–458 (1998). [CrossRef]
11. J. Pundir and T. El Toukhy, “Uterine cavity assessment prior to IVF,” Women's health. 6(6), 841–848 (2010). [CrossRef]
12. F. Bonilla-Musoles, F. Raga, N. G. Osborne, J. C. Castillo, and F. Bonilla, “Endometrial receptivity: evaluation with ultrasound,” Women’s Health 29(1), 3–20 (2013). [CrossRef]
13. R. L. Schild, C. Knobloch, C. Dorn, R. Fimmers, H. van der Ven, and M. Hansmann, “Endometrial receptivity in an in vitro fertilization program as assessed by spiral artery blood flow, endometrial thickness, endometrial volume, and uterine artery blood flow,” Fertil. Steril. 75(2), 361–366 (2001). [CrossRef]
14. I. Y. Järvelä, P. Sladkevicius, S. Kelly, K. Ojha, S. Campbell, and G. Nargund, “Evaluation of endometrial receptivity during in-vitro fertilization using three-dimensional power Doppler ultrasound,” Ultrasound Obstet Gynecol 26(7), 765–769 (2005). [CrossRef]
15. T. El-Toukhy, S. Sunkara, and Y. Khalaf, “Local endometrial injury and IVF outcome: a systematic review and meta-analysis,” Reprod. BioMed. Online 25(4), 345–354 (2012). [CrossRef]
16. A. Revel, “Defective endometrial receptivity,” Fertil. Steril. 97(5), 1028–1032 (2012). [CrossRef]
17. P. Díaz-Gimeno, M. Ruiz-Alonso, D. Blesa, N. Bosch, J. A. Martínez-Conejero, P. Alamá, N. Garrido, A. Pellicer, and C. Simón, “The accuracy and reproducibility of the endometrial receptivity array is superior to histology as a diagnostic method for endometrial receptivity,” Fertil. Steril. 99(2), 508–517 (2013). [CrossRef]
18. F. Z. Rarani, F. Borhani, and B. Rashidi, “Endometrial pinopode biomarkers: Molecules and microRNAs,” J. Cell. Physiol. 233(12), 9145–9158 (2018). [CrossRef]
19. G. B. La Sala, R. Montanari, L. Dessanti, C. Cigarini, and F. Sartori, “The role of diagnostic hysteroscopy and endometrial biopsy in assisted reproductive technologies,” Fertil. Steril. 70(2), 378–380 (1998). [CrossRef]
20. B. W. Rackow, E. Jorgensen, and H. S. Taylor, “Endometrial polyps affect uterine receptivity,” Fertil. Steril. 95(8), 2690–2692 (2011). [CrossRef]
21. M. G. Munro, “Uterine polyps, adenomyosis, leiomyomas, and endometrial receptivity,” Fertil. Steril. 111(4), 629–640 (2019). [CrossRef]
22. N. Makris, K. Kalmantis, N. Skartados, A. Papadimitriou, G. Mantzaris, and A. Antsaklis, “Three-dimensional hysterosonography versus hysteroscopy for the detection of intracavitary uterine abnormalities,” International Journal of Gynecology & Obstetrics 97(1), 6–9 (2007). [CrossRef]
23. R. Fanchin, C. Righini, J. M. Ayoubi, F. Olivennes, D. de Ziegler, and R. Frydman, “New look at endometrial echogenicity: objective computer-assisted measurements predict endometrial receptivity in in vitro fertilization-embryo transfer,” Fertil. Steril. 74(2), 274–281 (2000). [CrossRef]
24. J. L. Alcázar, “Three-dimensional ultrasound assessment of endometrial receptivity: a review,” Reprod. Biol. Endocrinol. 4(1), 56 (2006). [CrossRef]
25. C. B. Coulam, M. Bustillo, D. M. Soenksen, and S. Britten, “Ultrasonographic predictors of implantation after assisted reproduction,” Fertil. Steril. 62(5), 1004–1010 (1994). [CrossRef]
26. L. Jokubkiene, P. Sladkevicius, and L. Valentin, “Appearance of the endometrium at saline contrast sonohysterography in the luteal phase of the menstrual cycle: a prospective observational study,” Ultrasound Obstet Gynecol 45(3), 339–345 (2015). [CrossRef]
27. A. Al Chami and E. Saridogan, “Endometrial Polyps and Subfertility,” J Obstet Gynecol India 67(1), 9–14 (2017). [CrossRef]
28. M. Gonzalez, K. A. Montejo, K. Krupp, V. Srinivas, E. DeHoog, P. Madhivanan, and J. C. Ramella-Roman, “Design and implementation of a portable colposcope Mueller matrix polarimeter,” J. Biomed. Opt. 25(11), 116006 (2020). [CrossRef]
29. T. Law, F. Wu, H. Xu, C. C. Wang, and T. C. Li, “Endometrium imaging using real-time rotational optical coherence tomography imaging system: A pilot, prospective and ex-vivo study,” Medicine (Philadelphia, PA, U. S.) 98(44), e17738 (2019). [CrossRef]
30. Y. Gan, W. Yao, K. M. Myers, J. Y. Vink, R. J. Wapner, and C. P. Hendon, “Analyzing three-dimensional ultrastructure of human cervical tissue using optical coherence tomography,” Biomed. Opt. Express 6(4), 1090–1108 (2015). [CrossRef]
31. R. Zhang, T. Law, B. Liang, S. Hung, S. Lin, T. Li, and C. Wang, “P–361 Endometrium optical coherence tomography (OCT) and histomorphometry in implantation window and their relationships with reproductive failure and implantation outcome,” Hum. Reprod. 36(8), 2134–2147 (2021). [CrossRef]
32. J. Wang, Y. Xu, K. J. Mesa, F. A. South, E. J. Chaney, D. R. Spillman, R. Barkalifa, M. Marjanovic, P. S. Carney, A. M. Higham, Z. G. Liu, and S. A. Boppart, “Complementary use of polarization-sensitive and standard OCT metrics for enhanced intraoperative differentiation of breast cancer,” Biomed. Opt. Express 9(12), 6519–6528 (2018). [CrossRef]
33. C. Miranda, J. Barkley, and B. Smith, “Intrauterine photoacoustic and ultrasound imaging probe,” J. Biomed. Opt. 23(04), 1–9 (2018). [CrossRef]
34. Y. Li, R. Lin, C. Liu, J. Chen, H. Liu, R. Zheng, X. Gong, and L. Song, “In vivo photoacoustic/ultrasonic dual-modality endoscopy with a miniaturized full field-of-view catheter,” J. Biophotonics 11(10), e201800034 (2018). [CrossRef]
35. S. Naghdi, W. S. Slovinsky, M. Madesh, E. Rubin, and G. Hajnóczky, “Mitochondrial fusion and Bid-mediated mitochondrial apoptosis are perturbed by alcohol with distinct dependence on its metabolism,” Cell Death Dis. 9(10), 1028 (2018). [CrossRef]
36. S. Zhang, Y. Sun, D. Jiang, T. Chen, R. Liu, X. Li, Y. Lu, L. Qiao, Y. Pan, Y. Liu, and J. Lin, “Construction and Optimization of an Endometrial Injury Model in Mice by Transcervical Ethanol Perfusion,” Reprod. Sci. 28(3), 693–702 (2021). [CrossRef]
37. J. Leng, J. Zhang, C. Li, C. Shu, B. Wang, R. Lin, Y. Liang, K. Wang, L. Shen, K. H. Lam, Z. Xie, X. Gong, J. Ge, and L. Song, “Multi-spectral intravascular photoacoustic/ultrasound/optical coherence tomography tri-modality system with a fully-integrated 0.9-mm full field-of-view catheter for plaque vulnerability imaging,” Biomed. Opt. Express. 12(4), 1934–1946 (2021). [CrossRef]
38. C. Trojahn, G. Dobos, C. Richter, U. Blume-Peytavi, and J. Kottner, “Measuring skin aging using optical coherence tomography in vivo: a validation study,” J. Biomed. Opt. 20(4), 045003 (2015). [CrossRef]
39. B. Wang, W. Shi, and Z. Miao, “Confidence analysis of standard deviational ellipse and its extension into higher dimensional Euclidean space,” PloS one 10(3), e0118537 (2015). [CrossRef]
40. I. Brosens, I. Derwig, J. Brosens, L. Fusi, G. Benagiano, and R. Pijnenborg, “The enigmatic uterine junctional zone: the missing link between reproductive disorders and major obstetrical disorders?” Hum. Reprod. 25(3), 569–574 (2010). [CrossRef]
