Research Article
[Retracted] An Automated Deep Learning Model for the Cerebellum Segmentation from Fetal Brain Images
Table 1
Overview of reviewed literatures.
| Authors | Reference | Year | Article title | Proposed approach | Type of images | Training images | Testing images | Limitation |
| Singh et al. | [18] | 2021 | 2D fetal US brain images cerebellum semantic segmentation | ResU-Net-c | US | 588 | 146 | The model does not segment the cerebellum automatically | Zhao et al. | [19] | 2022 | Optimized DL method-based automated segmentation of 3D fetal brain images | 3D U-Net | MR | 65 | 41 | Not all the evaluation metrics are computed | Hesse et al. | [20] | 2022 | Fetal brain’s subcortical segmentation in 3D US images by DL methods | CNN | US | 215 | 20 | Segmentation robustness requires more improvement | Fidon et al. | [21] | 2021 | DRO-based abnormal brain of fetal segmentation using 3D MR images | nnU-Net-DRO | MR | 116 | 26 | The process of segmentation may take more time to complete | Kim et al. | [22] | 2019 | Fetal-head biometry automatic evaluation by ML | DL-based method | US | 102 | 70 | The method is employed only on smaller datasets | Khalili et al. | [23] | 2019 | Automatic segmentation of fetal brain tissue by CNN | CNN | MR | 32 | 94 | The error rate of developed method was not measured | Avisdris et al. | [24] | 2021 | Fetal brain automated linear measures by DNN | DNN | MR | 121 | 33 | The linear measurements take more time | Dumast et al. | [25] | 2022 | Segmentation of fetal brain tissue from synthetic MR images | FaBiaN | MR | 17 | 11 | The developed model is not applicable for large datasets | Rackerseder et al. | [26] | 2019 | 3D brain US fully automatic segmentation | DeepVNet | US | N/A | N/A | The quantitative evaluation takes more time | Venturini et al. | [27] | 2019 | Structural semantic segmentation using multitask CNN from 3D US images | Multitask CNN | US | 480 | 48 | The presented approach is challenging |
|
|