Development and validation of a deep learning model based on cascade mask regional convolutional neural network to noninvasively and accurately identify human round spermatids

TL;DR

A deep learning model was developed to noninvasively identify human round spermatids, which could help improve fertility treatments.

Contribution

A novel deep learning model using cascade mask R-CNN was developed for noninvasive and accurate identification of human round spermatids.

Findings

The model achieved a mean average precision (mAP) of over 0.80 in test datasets.

All cells selected by the model expressed PRM1 and/or PNA, confirming its accuracy.

The model's noninvasive approach is suitable for clinical application in human round spermatid injection.

Abstract

•Deep learning model was built by analyzing images of sorted human round spermatids (hRSs) by flow cytometric analysis.•Expression of PRM1and/or PNA (RSs markers) was observed in all cells selected by our model.•Results of double-blind test proved accuracy and effectiveness of our model for identifying hRSs.•Our model solved the most difficult technological problem of noninvasively and accurately identifying hRSs.•Our model will promote widely clinical application of human round spermatid injection technique.

Deep learning model was built by analyzing images of sorted human round spermatids (hRSs) by flow cytometric analysis.

Expression of PRM1and/or PNA (RSs markers) was observed in all cells selected by our model.

Results of double-blind test proved accuracy and effectiveness of our model for identifying hRSs.

Our model solved the most difficult technological problem of noninvasively and accurately identifying hRSs.

Our model will promote widely clinical application of human round spermatid injection technique.

The difficulty of identifying human round spermatids (hRSs) has impeded applications of the human round spermatid injection (ROSI) technique. RSs can be accurately screened through flow cytometric analysis utilizing the Hoechst fluorescence profile reflecting DNA, but this method is not suitable for isolating hRSs due to the toxicity associated with Hoechst staining.

To evaluate the capacity of a deep learning model grounded in a cascade mask region-based convolutional neural network (R-CNN) for the noninvasive and accurate identification of hRSs.

In this study, we presented the development and validation of a deep learning model for identifying hRSs through the analysis of 3457 optical light microscope images of sorted hRSs obtained via flow cytometric analysis. The model’s accuracy and specificity were evaluated by calculating the mean average precision (mAP). Furthermore, a double-blind experiment was conducted to access the reliability of the proposed model in accurately identifying hRSs. It detected the expression of protamine (PRM1) and/or peanut lectin (PNA), which are established markers for RSs.

Our deep learning-based model demonstrated a high precision, achieving a mAP of over 0.80 for isolating hRSs in test datasets. The expression of PRM1 and/or PNA was observed in all cells noninvasively selected by our AI model during an independent double-blind test. This phenomenon confirmed the accuracy and effectiveness of the proposed model. The model’s capability for noninvasive and accurate isolation of hRSs among spermatogenic cells highlighted its robustness and generalizability for clinical applications.

The deep learning AI model based on a cascade R-CNN has the ability to accurately identify hRSs among spermatogenic cells. The application of this noninvasive method, which requires no additional procedures in clinical practice, is able to facilitate the widespread implementation of ROSI technique. Therefore, it can provide patients with spermatogenic arrest the opportunity to become biological fathers.

Figures