Measuring cumulus expansion of mammalian oocytes: comparing the reliability of methods and how artificial intelligence could automate the measurement

doi:10.21203/rs.3.rs-2572620/v1

Download PDF

Article

Measuring cumulus expansion of mammalian oocytes: comparing the reliability of methods and how artificial intelligence could automate the measurement

https://doi.org/10.21203/rs.3.rs-2572620/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Cumulus expansion is an important indicator of oocyte maturation, often correlated with greater oocyte developmental capacity. Although multiple methods have been described to assess cumulus expansion, none of them is considered a gold standard. Additionally, these methods are subjective and time-consuming. Here, the reliability of three cumulus expansion measurement methods was evaluated and a deep learning model was created to automatically perform the measurement. Cumulus-oocyte complexes were compared before and after in vitro maturation by three independent observers using three methods: (1) measurement of the cumulus area, (2) measurement of three distances between the zona pellucida and outer cumulus, and (3) scoring cumulus expansion on a 5-point Likert scale. Inter- and intra-observer agreements were calculated using intraclass-correlation coefficients (ICC). The area method resulted in the best overall agreement with an ICC of 0.89 versus 0.54 and 0.30 for the 3-distance and scoring method, respectively. Therefore, the area method served as the base to create the deep learning model, which outperformed two observers while equivalent to the third. Measuring the area is the most reliable method to manually evaluate cumulus expansion, whilst deep learning automatically performs the calculation with human-level accuracy and could therefore be a valuable prospective tool for embryologists.

Biological sciences/Developmental biology/Embryology

Biological sciences/Physiology/Reproductive biology

Mammalian oocytes are surrounded by multiple layers of specialized somatic cells, called cumulus cells¹. These cells are connected to the oocyte through transzonal projections, providing the oocyte with metabolites and regulatory molecules during the maturation process^2–4. Likewise, during fertilization, the cumulus cells attract⁵, trap⁵, and select⁶ spermatozoa, and prevent premature hardening of the zona pellucida⁷, which are all actions required for successful fertilization. As such, cumulus cells assist the oocyte in completing the processes of maturation, fertilization, and early embryonic development^6,7. To effectively carry out their function, cumulus cells need to respond to gonadotropins whereby a mucinous matrix of hyaluronic acid is created extracellularly, causing the cumulus to expand^8–12. Since adequate cumulus expansion has been correlated with high developmental potential^13,14, the degree of cumulus expansion is considered an important parameter for oocyte quality^15,16. However, when different compositions of maturation media are being compared, one should be careful in using the degree of cumulus expansion as a predictor for developmental competence, since variable concentrations of gonadotropins or protein sources in the media can influence the degree of cumulus expansion^17–19. Leibfried-Rutledge and colleagues¹⁹ demonstrated for example that cumulus expansion was stimulated when supplementing a high dosage of fetal calf serum to the maturation medium, while this was not the case when bovine serum albumin was supplemented. Yet, lack of cumulus expansion during maturation resulted in decreased ovulation rates and impaired fertility in murine oocytes^8,20. Hence, an adequate measurement of cumulus expansion could be a useful tool for assisted reproduction in human, domestic, and wildlife species.

Multiple methods for the assessment of cumulus expansion have been described^{10–12, 19,21–29}, including both invasive and non-invasive techniques. Quantitative measurement of hyaluronic acid is the most precise approach^11,30, but it impairs further embryo development and can hold health hazards for the operator when it requires the application of radioactive material³⁰. The most frequently used method for cumulus expansion assessment is a classification into groups by evaluating the degree of cumulus expansion as a whole or by counting the layers of cumulus cells^{14,18,23,25,26,28}. This procedure is fast and practical, since it can be performed directly via microscopy while manipulating the oocyte for in vitro processing. No additional equipment is necessary and further in vitro development of the oocyte remains possible. The downside of this approach is that it is highly dependent on the expertise of the embryologist and is thus considered to be a subjective method. More recently, with the development of built-in cameras and image analysis software, alternative methods such as measuring the area^10,22 and longitudinal axes of the cumulus cells^24,29,31 have been reported aiming to diminish subjectivity. Yet, all so far proposed visual methods cannot record the three-dimensional structure of the cumulus-oocyte complex (COC) and are time-consuming. Moreover, no direct comparisons have been made to evaluate their precision, accuracy, and repeatability.

In all cases, the results remain dependent on the subjectivity of the human observer. A possible solution to remove the human factor from the equation, is to apply deep learning (DL) modelling. Deep learning is a subfield of artificial intelligence that is based on learning hierarchical knowledge from data rather than rule-based programming³². There is a growing number of applications for artificial intelligence in several areas of assisted reproduction, in order to automate time-consuming and subjective tasks^33–35. Non-invasive methods to assess cumulus expansion, both qualitatively and (semi-) quantitatively, depend deeply on human interpretation and require a considerable amount of time. Hence, the development of a tool that can objectively and accurately automate this task could improve the reliability of the correlation between cumulus expansion and oocyte developmental capacity and facilitate the assessment for researchers and embryologists.

Cumulus expansion is an important marker of successful COC maturation and its assessment is often used in reproductive biotechnologies^8,36. However, none of the available non-invasive methods can be considered a gold standard, and all of them are time-consuming and involve a degree of subjectivity. In the present manuscript, the results are reported of a side-by-side comparison between three non-invasive visual evaluation methods to identify the method with the best intraclass correlation coefficient (ICC) between and within observers. Additionally, based on the results of this comparison, AIxpansion, a DL method to automate cumulus expansion measurement, is created.

No ethical approval was necessary for this experiment, since ovaries were collected post-mortem from cows in a commercial slaughterhouse.

2.1 Media and reagents

Tissue culture medium (TCM-199), gentamycin and phosphate buffered saline were purchased from Life Technologies Europe (Ghent, Belgium). All other chemicals were obtained from Sigma-Aldrich (Overijse, Belgium) unless otherwise listed. All media were filtered before use [0.22 µm; GE Healthcare-Whatman (Diegem, Belgium)].

2.2 Oocyte collection and in vitro maturation

In vitro oocyte maturation was performed as previously described by Azari-Dolatabad et al.²⁴. In brief, bovine ovaries were collected from a local slaughterhouse and processed within 2 h after collection. Ovaries were rinsed three times in warm (37˚C) physiological saline supplemented with kanamycin (25 mg/mL) and sterilized with 90% ethanol. Cumulus-oocyte complexes were aspirated from follicles between 4 and 8 mm in diameter using an 18-gauge needle attached to a 10 mL syringe and transferred into a 15 mL tube containing 2.5 mL of HEPES Tyrode’s albumin–pyruvate–lactate (HEPES-TALP) medium. Cumulus-oocyte complexes were subsequently cultured individually in droplets of 20 µL of maturation medium (TCM-199, supplemented with 20 ng/mL epidermal growth factor (EGF) and 50 µg/mL gentamycin) covered with 7.5 mL paraffin oil (SAGE oil for tissue culture, ART-4008-5P, Cooper Surgical Company) for 22 h at 38.5°C in 5% CO₂ in humidified air.

2.3 Image acquisition

Images from 300 COCs were acquired at 0 h of maturation (immature stage) and 22 h after maturation (matured stage) using an inverted Olympus stereomicroscope, connected to a ToupCam camera and ToupView software (version 3.7.13270.20181102). Images were obtained under the same magnification (56X), with the plane of focus on the zona pellucida of a single oocyte in the middle of the field. Images were saved as PNG files at a resolution of 2592 x 1944 pixels in RGB. Sixty-eight images were excluded since the zona pellucida was not clearly in focus, the demarcation of the outer cumulus cells was not clear and/or the COC rotated into the 3-dimensional field during maturation, causing the pre- and post-maturation images to be acquired from different 2-dimensional fields. After exclusion, a total of 232 paired (prior to and after maturation) images of COCs were presented to three observers as mentioned earlier.

2.4 Cumulus expansion measurement

Cumulus expansion of bovine oocytes before and after in vitro maturation was measured by three independent observers, with sufficient experience. The observers scored each oocyte by three different methods (area, 3-distance, and scoring method) in duplicate, at different times, and in random order.

2.4.1 Area

The area of the immature and mature COCs was measured by drawing the contour of the COC using the freehand selection in ImageJ software³⁷ (version 1.49 q; National Institutes of Health, USA). To calculate the absolute growth, the immature COC area was subtracted from the mature COC area. The difference in growth was divided by the area of the immature COC to compute the cumulus expansion (%) (Fig. 1).

2.4.2 3-distance

The shortest, medium, and longest distance between the zona pellucida and the extremes of the cumulus cells was measured in immature COCs, using the straight lines tool in ImageJ. Next, the average value of these three distances was calculated. The same was done for mature oocytes. The average value related to the immature COC was subtracted from the mature COC’s average value, which resulted in absolute growth. Cumulus expansion (%) was then determined by dividing the absolute growth by the average value of the immature COC (Fig. 1).

2.4.3 Scoring

The images of immature and corresponding mature COCs were directly compared to each other to evaluate the expansion of the cumulus cells by assigning a score on a 5-point Likert scale, as previously described by Downs²⁶. In brief, the score ranged from 0 to 4, with “0”: no expansion; “1”: separation of only the outermost layers of cumulus cells; “2”: further expansion involving the outer half of the cumulus oophorous; “3”: further expansion up to, but not including the corona radiata; and “4”: complete expansion, including the innermost corona radiata cells. (Fig. 2).

2.6 AIxpansion: a deep learning model for cumulus expansion measurement

As the area method resulted in the highest ICC for inter- and intra-observer agreement, the automatization of this method was pursued for creating AIxpansion: a DL algorithm that recognizes the area of the COC automatically and thus can be used for measuring cumulus expansion. AIxpansion, technically explained in detail in Athanasiou et al.³⁸ exploits DL modeling to automatically calculate cumulus expansion. To construct the DL model, AIxpansion combines transfer learning and image pre-processing with a U-Net network inspired by Ronneberger et al.³⁹ The improvement of AIxpansion’s performance relied on transfer learning from a publicly available dataset of melanoma images⁴⁰. This was used to generate the first pre-trained DL model, which served as a starting point for training AIxpansion with images of immature and mature COCs. From data acquired in section 2.3, a subset of 100 oocytes was randomly selected, due to image annotation costs. Experts provided segmentation masks by manually drawing the area of immature and mature COCs. These images and their corresponding segmentation masks were used as the training input for AIxpansion.

To evaluate the segmentation models, i.e., the ability of AIxpansion to detect the cumulus demarcation, a 10-fold cross-validation was used since the dataset size was limited and this approach provided stable results⁴¹. At each fold, 90 oocytes (immature and mature; 180 images) were used for training and 10 oocytes (20 images) for validation. The model was trained on minibatches of 32 for 200 epochs. For each fold, a segmentation mask was generated for the 20 validation images, resulting in a total of 200 masks after going through all the folds. Then, the generated masks were compared with masks of the same oocytes that were provided by each of the experts (Fig. 3). To evaluate the similarity between two masks, the dice coefficient was used⁴². The mean dice performance of the majority-vote model converged at around 95%.

2.7 Statistical analyses

To compare the three different methods for cumulus expansion measurement, the following variables were studied: (1) median CE of all oocytes considered, (2) inter-observer agreement, i.e., the variation in cumulus expansion as scored between different observers, (3) overall inter-observer agreement, i.e., the average of the two measurement repetitions for the inter-observer agreement (as the measurements were performed in duplicate) and (4) intra-observer agreement, i.e., the variation in cumulus expansion scored between repeated measurement for each of the observers.

The data were analyzed with Python, version 3.10.6. A two-way random effects model was used to evaluate inter-observer agreement and a one-way random effects model was performed to evaluate the intra-observer agreement for each of the experts. Consequently, the ICC and its 95% confidence interval were computed using the intraclass_corr function in the Pingouin Python statistical library⁴³, version 0.5.2. The code used for the evaluation is available at the following link https://github.com/IIIA-ML/cumulus_expansion_variance_analysis. The interpretation of the ICC values was as proposed by Landis and Koch⁴⁴: <0.20, poor agreement; 0.20–0.39, fair agreement; 0.40–0.59, moderate agreement; 0.60–0.79, good agreement, > 0.80, very good agreement. Data are reported as ICC and 95% confidence intervals.

To compare the results of the DL method with the results of the human observers, the similarity among the annotations of each of the three experts was computed and compared with the similarity between each of the human experts and the proposed method. To evaluate the similarity between AIxpansion and human observers, different metrics were used: average rank, bias, and variances. The first one focused on how many of the AIxpansion estimations were closer to the experts. To do so, the average rank metric was used, in which the estimators were ranked with respect to how close their score is to the score of reference. For example, if observer 1 is set as a reference, then the scores of observer 2, observer 3, and AIxpansion, are compared to the score of observer 1. The scores are then ranked from 1 to 3, according to which estimator has the minimum distance to the reference. The same procedure was accomplished for every observer. Finally, the lowest value indicates better estimation, according to Supplementary Equation S1. Additionally, to support the results, the biases and variances among the AIxpansion estimator and the expert’s annotations were examined, according to Supplementary Equation S2 and S3 respectively.

Three different methods to measure cumulus expansion were evaluated by three observers. The working principle, user-friendliness (based on equipment and time requirements), and performance (based on level of subjectivity) of these methods are summarized in Table 1.

Table 1. Comparison of methods to measure cumulus expansion. Methods were compared by three observers and evaluated for equipment- and time requirements and the level of subjectivity. +, easy or low; ++, moderate; +++, complicated or high. ^AEquipment required: apart from a stereomicroscope, software was needed to calculate cumulus expansion for the area and 3-distance method.

Method	Working principle	Equipment^A	Time
Area	Measuring the area by drawing the COC contour	++	++
3-distance	Measuring 3 distances between zona pellucida and outer cumulus	++	+++
Scoring	5-point Likert scale	+	+

3.1 Cumulus expansion in numbers

Both the area and the 3-distance method resulted in numerical data, while the results of cumulus expansion derived from the scoring method were ordinal. Therefore, no direct comparisons were made between the scoring method and the other two methods.

Observations of cumulus expansion evaluated by the area method varied from 0.06 to 346.3% with a median growth of 74.28% (interquartile range (IQR): 31.01–154.2%; Fig. 4A). The observations resulting from the 3-distance method ranged from 0.15 to 346.9% and had a median growth of 52.68% (IQR: 15.29–106.2%; Fig. 4A). The scoring method yielded a median of 3.0 (IQR: 1.0–3.0; Fig. 4B).

3.2 The area method resulted in the highest inter-observer and intra- observer agreement

3.2.1 Inter-observer agreement

For all three methods, the agreement between the observers was evaluated by calculating the corresponding ICC, as illustrated in Fig. 5a and Table 2. This ICC was calculated in duplicate, as the measurements were performed twice by every observer for every method. The inter-observer ICCs for the area method showed a very good level of agreement, while the 3-distance method resulted in a moderate level of agreement. The inter-observer agreement for the scoring method was fair.

3.2.2 Overall inter-observer agreement

An overall agreement level was calculated for every method, where both repetitions of inter-observer agreement were considered. This resulted in a very good overall agreement for the area method, a moderate overall agreement for the 3-distance method and a poor overall agreement for the scoring method, as shown in Fig.5b and Table 2.

3.2.3 Intra-observer agreement

The intra-observer agreement was evaluated by ICC calculations for every method and every observer (Fig. 5c; Table 2). Overall, intra-observer agreements for observers 1, 2, and 3 were very good for the area method, while moderate to good for the 3-distance method. The results for the scoring method varied per observer, as the level of agreement ranged from poor, over moderate, to good.

Table 2. Intraclass correlation coefficients for three cumulus expansion measurement methods. Data are reported as intraclass correlation coefficients with their respective 95% confidence intervals.

Method	Inter-observer agreement		Overall inter-observer agreement	Intra-observer agreement
	Repetition 1	Repetition 2	Overall inter-observer agreement	Observer 1	Observer 2	Observer 3
Area	0.89 (0.88–0.92)	0.90 (0.85–0.93)	0.89 (0.85–0.93)	0.87 (0.84–0.90)	0.90 (0.87–0.92)	0.96 (0.95–0.97)
3-distance	0.56 (0.49–0.63)	0.51 (0.44–0.59)	0.54 (0.44–0.63)	0.61 (0.53–0.69)	0.59 (0.50–0.67)	0.64 (0.56–0.71)
Scoring	0.23 (0.12–0.34)	0.38 (0.3–0.47)	0.30 (0.12–0.47)	0.69 (0.63–0.76)	0.11 (-0.01–0.24)	0.51 (0.42–0.6)

3.3 AIxpansion automatically measures cumulus expansion based on the COCs area

3.3.1 AIxpansion processor capacity

As ‘area’ resulted in the highest ICC values for manually measuring cumulus expansion, a DL model, AIxpansion, was created based on this method. The pre-processing stage, in which the region of interest was detected, was able to correctly determine 98% of the region of interest. Failure of detection was due to a very low signal of the COC and to the presence of an oil droplet in the image, which interfered with the model. These 2 cases were excluded from further analyses, leading to a total of 98 COCs used to train the model as explained in section 2.5.

3.3.2 Alxpansion performs similar to human observers

The average ranking, i.e. the comparison of the different scores among the observers, is reported in Table 3. In 2 out of 3 cases, Alxpansion performs better compared to the other observers. The performance of AIxpansion in measuring the COC’s area was similar to that of the observers (p = 0.15).

Bias and variance among observers and between human observers and AIxpansion are reported in Supplementary Table S1 and S2, respectively. Measuring cumulus expansion using AIxpansion resulted in lower bias and less variance compared to the human observers in 1 out of 3 times. For the remaining 2 cases, AIxpansion scored similarly for both bias and variance to the human observers, proving that AIxpansion reaches a human-level performance.

Table 3. Average rank calculated in the comparison between the different human observers and human observers vs Alxpansion. This table shows the similarity between the average ranking of three observers (O1-O3) and the deep learning method (AIxpansion). Scores closer to zero indicate that the performance is closer to the column observer.

	O1	O2	O3
O1		2.04	1.93
O2	2.04		2.15
O3	1.97	2.11
AIxpansion	1.99	1.85	1.93

This study demonstrates that comparing the area of the COC before and after maturation is the most reliable non-invasive method to assess cumulus expansion. As a result, the area method was used as a basis to develop a DL model, AIxpansion. This model was able to measure the area of COCs with similar reliability to the human observers, in a tenth of a second.

The most frequently used method to evaluate cumulus expansion is currently the scoring method, where a subjective score from 0 to 4 is given on a Likert scale^{12,14,18,23,26,45,46}. In the present study, the scoring method provided in general lower ICC values compared to the area and 3-distance method, indicating a higher degree of subjectivity for the scoring method. As the intra-observer agreements were very good for the area- and good for the 3-distance method, the mean evaluation result of multiple random raters is expected to be close to each other and the selection of an observer is of little consequence. Still, a large variation is reported in the intra-observer agreements for the scoring method (poor, moderate, and good for observer 1, 2 and, 3, respectively). Consequently, the observer may introduce a bias and the results should be interpreted with caution when using the scoring method. The same conclusion can be drawn from the inter-observer ICC values when both repetitions are being compared: in case of the area and 3-distance method, the inter-observer ICC from the first repetition was close to the ICC of the second repetition while more variation was noticed in the scoring method.

Moreover, the general performance of the area method was more repeatable than the 3-distance method and drawing the area of the COCs was less time-consuming than calculating three specific distances. The fact that the three distances are chosen arbitrarily, and that in some images the zona pellucida is not clearly distinguishable because of an overlying cloud of cumulus cells, could explain the moderate reliability of the 3-distance method. Although all ICC values were the highest for the area method, some variation (although limited) was still reported as no ICC score was equal to 1. This variation was probably caused by the demarcation of the outer cumulus cells, which is open for interpretation, especially when the outer cumulus cells are completely expanded. The inability to evaluate the 3-dimensional shape of the COC is an important limitation associated with the area- and 3-distance methods, as positional changes of the COC in the 3-dimensional field could result in erroneous measurements and calculations. Since the scoring method is based on a simple visual inspection it is not greatly affected by changes in the position of the COCs during in vitro maturation. Other methods like spectrophotometry and high-performance liquid chromatography have been proposed to measure the amount of hyaluronic acid-degrading products in conditioned media⁴⁷. These methods are not routinely performed in the IVF lab since specific equipment and expertise are required.

In recent years, assisted reproductive technologies have benefited from the increased use of artificial intelligence techniques such as DL and image segmentation^{34, 48–53}, although segmenting images of bovine COCs had not been done yet. AIxpansion was therefore developed to automatically measure the area of immature and mature bovine COCs by image segmentation, and to consequently calculate the relative expansion of the cumulus cells. The performance of AIxpansion reaches the same level of reliability as the human observers with higher efficiency. It generates results with a lower bias and variance compared to human observers. However, sometimes the signal from largely expanded COCs becomes translucent, leading to difficulties in surpassing human-level performance. AIxpansion calculates cumulus expansion in milliseconds and generates a dataset of quantitative cumulus expansion information while the embryologist acquires COC images in the lab, with no need for manually measuring cumulus expansion afterward. Still, AIxpansion measures cumulus expansion from a 2-dimensional image. Future research may focus on 3-dimensional imaging for which the technique of lens-free imaging could be proposed to evaluate the expansion of cumulus cells in 3 dimensions, without affecting the oocyte’s capacity to be fertilized and develop into an embryo.

This study assessed the reliability of three methods that are routinely used to evaluate cumulus expansion using ICC. Although scoring cumulus expansion on a Likert scale is the most frequently used method, it showed the highest degree of subjectivity in the present study. The evaluation of the area before and after in vitro maturation resulted in the best inter- and intra-observer agreements. Therefore, the area method was used as a basis to develop a deep learning algorithm for further measurements: AIxpansion. AIxpansion successfully measured the area automatically and provided a fast and reliable way to calculate cumulus expansion. Since cumulus expansion is an important parameter to evaluate oocyte quality and the level of oocyte maturation, AIxpansion could be a useful tool for embryologists and researchers in the in vitro embryo production lab. Although AIxpansion was created using bovine oocytes, its algorithm could be extrapolated to other species where in vitro oocyte maturation is performed, such as horses⁵⁴, wildlife⁵⁵ and humans⁵⁶.

Acknowledgements

A.R. is supported by the Special Research Fund (BOF) of Ghent University, project number 01D12519. N.A.-D. is supported by “GOA: Ex ovo omnia: from stem cell to oocyte BOF GOA2018000501”. G. A. is supported by the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 860960 and by project CI-SUSTAIN (PID2019-104156GB-I00) funded by the Spanish Ministry of Science and Innovation. O.B.P. is supported by Fonds Wetenschappelijk Onderzoek: 12Y5220N. D.A.-V. is supported by Fonds of COLFUTURO and the Ministerio de Ciencia Tecnología Innovación of Colombia.

Author Contributions

Conceptualization: D. A.-V.; Collection of data: A. R., G. A., H. S., S. G.-A.; Performed the analysis: G. A. J. C.; Wrote the paper: N. A.-D., A.R., G. A., D. A.-V.; Review & editing: O. B. P., K. C. P.; Supervision: A. V. S., J. L. A., K. S., G. O. All authors read the final draft and agreed to its submission for publication.

Competing interests

The authors have no conflicts of interest.

Data availability statement

Data that supports the findings of this manuscript are available from the corresponding author, A.R., upon reasonable request.

Diaz, F. J., Wigglesworth, K. & Eppig, J. J. Oocytes determine cumulus cell lineage in mouse ovarian follicles. J Cell Sci 120, (2007).
Robert, C. Nurturing the egg: The essential connection between cumulus cells and the oocyte. Reproduction, Fertility and Development Preprint at https://doi.org/10.1071/RD21282 (2021).
Dumesic, D. A., Meldrum, D. R., Katz-Jaffe, M. G., Krisher, R. L. & Schoolcraft, W. B. Oocyte environment: Follicular fluid and cumulus cells are critical for oocyte health. Fertility and Sterility vol. 103 303–316 Preprint at https://doi.org/10.1016/j.fertnstert.2014.11.015 (2015).
Macaulay, A. D. et al. Cumulus cell transcripts transit to the bovine oocyte in preparation for maturation. Biol Reprod (2016) doi:10.1095/biolreprod.114.127571.
Eisenbach, M. Sperm chemotaxis. Rev Reprod 4, 56–66 (1999).
Kidder, G. M. & Mhawi, A. A. Gap junctions and ovarian folliculogenesis. Reproduction 123, 613–620 (2002).
van Soom, A., Tanghe, S., de Pauw, I., Maes, D. & de Kruif, A. Function of the cumulus oophorus before and during mammalian fertilization. Reprod Domest Anim 37, 144–151 (2002).
Chen, L., Russell, P. T. & Larsen, W. J. Functional significance of cumulus expansion in the mouse: Roles for the preovulatory synthesis of hyaluronic acid within the cumulus mass. Mol Reprod Dev (1993) doi:10.1002/mrd.1080340114.
Eppig, J. J., O’Brien, M. & Wigglesworth, K. Mammalian oocyte growth and development in vitro. Mol Reprod Dev 44, 260–273 (1996).
Yokoo, M., Kimura, N., Abe, H. & Sato, E. Influence of hyaluronan accumulation during cumulus expansion on in vitro porcine oocyte maturation. Zygote 16, 309–314 (2008).
Kimura, N., Konno, Y., Miyoshi, K., Matsumoto, H. & Sato, E. Expression of Hyaluronan Synthases and CD44 Messenger RNAs in Porcine Cumulus-Oocyte Complexes During In Vitro Maturation. Biol Reprod 66, 707–717 (2002).
Han, Z. bin et al. Interactive effects of granulosa cell apoptosis, follicle size, cumulus-oocyte complex morphology, and cumulus expansion on the developmental competence of goat oocytes: a study using the well-in-drop culture system. Reproduction 132, 749–758 (2006).
Nevoral, J. et al. Cumulus cell expansion, its role in oocyte biology and perspectives of measurement: A review. Scientia Agriculturae Bohemica (2014) doi:10.1515/sab-2015-0002.
Rose-Hellekant, T. A., Libersky-Williamson, E. A. & Bavister, B. D. Energy substrates and amino acids provided during in vitro maturation of bovine oocytes alter acquisition of developmental competence. Zygote 6, 285–294 (1998).
Pan, Y. et al. Estrogen improves the development of yak (Bos grunniens) oocytes by targeting cumulus expansion and levels of oocyte-secreted factors during in vitro maturation. PLoS One (2020) doi:10.1371/journal.pone.0239151.
Demiray, S. B. et al. Differential gene expression analysis of human cumulus cells. Clin Exp Reprod Med (2019) doi:10.5653/cerm.2019.46.2.76.
Sutton-McDowall, M. L., Gilchrist, R. B. & Thompson, J. G. Cumulus expansion and glucose utilisation by bovine cumulus-oocyte complexes during in vitro maturation: The influence of glucosamine and follicle-stimulating hormone. Reproduction 128, 313–319 (2004).
Choi, Y. H., Carnevale, E. M., Seidel, G. E. & Squires, E. L. Effects of gonadotropins on bovine oocytes matured in TCM-199. Theriogenology 56, 661–670 (2001).
Leibfried-Rutledge, M. L., Critser, E. S. & First, N. L. Effects of fetal calf serum and bovine serum albumin on in vitro maturation and fertilization of bovine and hamster cumulus-oocyte complexes. Biol Reprod 35, 850–857 (1986).
Hess, K. A., Chen, L. & Larsen, W. J. Inter-α-inhibitor binding to hyaluronan in the cumulus extracellular matrix is required for optimal ovulation and development of mouse oocytes. Biol Reprod (1999) doi:10.1095/biolreprod61.2.436.
Eppig, J. J. Role of Serum in FSH Stimulated Cumulus Expansion by Mouse Oocyte-Cumulus Cell Complexes in vitro. Biol Reprod 22, 629–633 (1980).
Furnus, C. C., De Matos, D. G. & Moses, D. F. Cumulus expansion during in vitro maturation of bovine oocytes: Relationship with intracellular glutathione level and its role on subsequent embryo development. Mol Reprod Dev 51, 76–83 (1998).
Vanderhyden, B. C., Caron, P. J., Buccione, R. & Eppig, J. J. Developmental pattern of the secretion of cumulus expansion-enabling factor by mouse oocytes and the role of oocytes in promoting granulosa cell differentiation. Dev Biol (1990) doi:10.1016/0012-1606(90)90081-S.
Azari-Dolatabad, N. et al. Follicular fluid during individual oocyte maturation enhances cumulus expansion and improves embryo development and quality in a dose-specific manner. Theriogenology 166, 38–45 (2021).
Machado, M. F. et al. Fibroblast growth factor 17 and bone morphogenetic protein 15 enhance cumulus expansion and improve quality of in vitro-produced embryos in cattle. Theriogenology 84, 390–398 (2015).
Downs, S. M. Specificity of Epidermal Growth Factor Action on Maturation of the Murine Oocyte and Cumulus Oophorus in Vitrol. Biol Reprod 41, 371–379 (1989).
Hunter, A. G. & Moor, R. M. Stage-Dependent Effects of Inhibiting Ribonucleic Acids and Protein Synthesis on Meiotic Maturation of Bovine Oocytes In Vitro. J Dairy Sci 70, 1646–1651 (1987).
Kobayashi, K., Yamashita, S. & Hoshi, H. Influence of epidermal growth factor and transforming growth factor-alpha on in vitro maturation of cumulus cell-enclosed bovine oocytes in a defined medium. J Reprod Fertil 100, 439–446 (1994).
Romero-Aguirregomezcorta, J., Santa, Á. P., García-Vázquez, F. A., Coy, P. & Matás, C. Nitric oxide synthase (NOS) inhibition during porcine in vitro maturation modifies oocyte protein S-nitrosylation and in vitro fertilization. PLoS One 9, (2014).
Eppig, J. J. Role of Serum in FSH Stimulated Cumulus Expansion by Mouse Oocyte-Cumulus Cell Complexes in vitro. Biol Reprod 22, 629–633 (1980).
Azari-Dolatabad, N. et al. Oocyte developmental capacity is influenced by intrinsic ovarian factors in a bovine model for individual embryo production. Anim Reprod Sci 249, 107185 (2023).
Kim, K. G. Book Review: Deep Learning. Healthc Inform Res (2016) doi:10.4258/hir.2016.22.4.351.
Bui, T. T. H. et al. Cytoplasmic movement profiles of mouse surrounding nucleolus and not-surrounding nucleolus antral oocytes during meiotic resumption. Mol Reprod Dev (2017) doi:10.1002/mrd.22788.
Blank, C. et al. Prediction of implantation after blastocyst transfer in in vitro fertilization: a machine-learning perspective. Fertil Steril (2019) doi:10.1016/j.fertnstert.2018.10.030.
Curchoe, C. L. & Bormann, C. L. Artificial intelligence and machine learning for human reproduction and embryology presented at ASRM and ESHRE 2018. Journal of Assisted Reproduction and Genetics Preprint at https://doi.org/10.1007/s10815-019-01408-x (2019).
Carrell, D. T., Middleton, R. G., Peterson, C. M., Jones, K. P. & Urry, R. L. Role of the cumulus in the selection of morphologically normal sperm and induction of the acrosome reaction during human in vitro fertilization. Syst Biol Reprod Med (1993) doi:10.3109/01485019308988391.
Abràmoff, M. D., Magalhães, P. J. & Ram, S. J. Image processing with imageJ. Biophotonics International Preprint at https://doi.org/10.1201/9781420005615.ax4 (2004).
Athanasiou, G. et al. Detecting the Area of Bovine Cumulus Oocyte Complexes Using Deep Learning and Semantic Segmentation. Artificial Intelligence Research and Development 0, (2022).
Ronneberger, O., Fischer, P. & Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation.
Gutman, D. et al. Skin Lesion Analysis toward Melanoma Detection: A Challenge at the International Symposium on Biomedical Imaging (ISBI) 2016, hosted by the International Skin Imaging Collaboration (ISIC).
Kohavi, R. A Study of Cross-Validation and Bootstrap for Accuracy Estimation and Model Selection.
Dice, L. R. Measures of the Amount of Ecologic Association Between Species. Ecology 26, 297–302 (1945).
Vallat, R. Pingouin: statistics in Python. J Open Source Softw (2018) doi:10.21105/joss.01026.
Landis, J. R. & Koch, G. G. The Measurement of Observer Agreement for Categorical Data. Biometrics 33, 159 (1977).
Lorenzo, P. L., Illera, M. J., Illera, J. C. & Illera, M. Enhancement of cumulus expansion and nuclear maturation during bovine oocyte maturation in vitro by the addition of epidermal growth factor and insulin-like growth factor I. J Reprod Fertil (1994) doi:10.1530/jrf.0.1010697.
Rouhollahi Varnosfaderani, S. et al. Importance of the GDF9 signaling pathway on cumulus cell expansion and oocyte competency in sheep. Theriogenology (2013) doi:10.1016/j.theriogenology.2013.05.009.
Alkrad, J. A., Mrestani, Y., Stroehl, D., Wartewig, S. & Neubert, R. Characterization of enzymatically digested hyaluronic acid using NMR, Raman, IR, and UV-Vis spectroscopies. J Pharm Biomed Anal (2003) doi:10.1016/S0731-7085(02)00682-9.
Raimundo, J. M. & Cabrita, P. Artificial intelligence at assisted reproductive technology. in Procedia Computer Science (2021). doi:10.1016/j.procs.2021.01.189.
Uyar, A., Bener, A. & Ciray, H. N. Predictive Modeling of Implantation Outcome in an In Vitro Fertilization Setting: An Application of Machine Learning Methods. MEDICAL DECISION MAKING (2015).
Huang, B. et al. Using deep learning to predict the outcome of live birth from more than 10,000 embryo data. BMC Pregnancy Childbirth (2022) doi:10.1186/s12884-021-04373-5.
Cao, Q. et al. Identification of viable embryos using deep learning for medical image. in ACM International Conference Proceeding Series (2018). doi:10.1145/3309129.3309143.
Louis, C. M. et al. Review of computer vision application in in vitro fertilization: the application of deep learning-based computer vision technology in the world of IVF. Journal of Assisted Reproduction and Genetics Preprint at https://doi.org/10.1007/s10815-021-02123-2 (2021).
Raef, B. & Ferdousi, R. A review of machine learning approaches in assisted reproductive technologies. Acta Informatica Medica Preprint at https://doi.org/10.5455/aim.2019.27.205-211 (2019).
Smits, K., Hoogewijs, M., Woelders, H., Daels, P. & van Soom, A. Breeding or assisted reproduction? Relevance of the horse model applied to the conservation of endangered equids. Reproduction in Domestic Animals 47, (2012).
Comizzoli, P., Mermillod, P. & Mauget, R. Reproductive biotechnologies for endangered mammalian species. Reprod Nutr Dev (2000) doi:10.1051/rnd:2000113.
De Matos, D. G. et al. Leukemia inhibitory factor induces cumulus expansion in immature human and mouse oocytes and improves mouse two-cell rate and delivery rates when it is present during mouse in vitro oocyte maturation. Fertil Steril (2008) doi:10.1016/j.fertnstert.2007.10.061.

No competing interests reported.

MeasuringcumulusexpansionSupplementarydata.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Measuring cumulus expansion of mammalian oocytes: comparing the reliability of methods and how artificial intelligence could automate the measurement

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials And Methods

2.1 Media and reagents

2.2 Oocyte collection and in vitro maturation

2.3 Image acquisition

2.4 Cumulus expansion measurement

2.4.1 Area

2.4.2 3-distance

2.4.3 Scoring

2.6 AIxpansion: a deep learning model for cumulus expansion measurement

2.7 Statistical analyses

3. Results

3.2 The area method resulted in the highest inter-observer and intra- observer agreement

3.2.1 Inter-observer agreement

3.2.2 Overall inter-observer agreement

3.2.3 Intra-observer agreement

3.3 AIxpansion automatically measures cumulus expansion based on the COCs area

3.3.1 AIxpansion processor capacity

3.3.2 Alxpansion performs similar to human observers

4. Discussion

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1