Ground-glass opacity (GGO) is a common CT imaging sign on high-resolution CT, which means the lesion is more likely to be malignant compared to common solid lung nodules. The automatic recognition of GGO CT imaging signs is of great importance for early diagnosis and possible cure of lung cancers. The present GGO recognition methods employ traditional low-level features and system performance improves slowly. Considering the high-performance of CNN model in computer vision field, we proposed an automatic recognition method of 3D GGO CT imaging signs through the fusion of hybrid resampling and layer-wise fine-tuning CNN models in this paper. Our hybrid resampling is performed on multi-views and multi-receptive fields, which reduces the risk of missing small or large GGOs by adopting representative sampling panels and processing GGOs with multiple scales simultaneously. The layer-wise fine-tuning strategy has the ability to obtain the optimal fine-tuning model. Multi-CNN models fusion strategy obtains better performance than any single trained model. We evaluated our method on the GGO nodule samples in publicly available LIDC-IDRI dataset of chest CT scans. The experimental results show that our method yields excellent results with 96.64% sensitivity, 71.43% specificity, and 0.83 F1 score. Our method is a promising approach to apply deep learning method to computer-aided analysis of specific CT imaging signs with insufficient labeled images.
Graphical abstract We proposed an automatic recognition method of 3D GGO CT imaging signs through the fusion of hybrid resampling and layer-wise fine-tuning CNN models in this paper. Our hybrid resampling reduces the risk of missing small or large GGOs by adopting representative sampling panels and processing GGOs with multiple scales simultaneously. The layer-wise fine-tuning strategy has ability to obtain the optimal fine-tuning model. Our method is a promising approach to apply deep learning method to computer-aided analysis of specific CT imaging signs with insufficient labeled images.
Facioscapulohumeral muscular dystrophy (FSHD) is the third most common inherited muscular dystrophy with markedly clinical variability and complex genetic cause. Several reports pertaining to the Caucasian population have confirmed that there are 4qA and 4qB variants of the 4qter subtelomere, and FSHD is uniquely associated with the 4qA variant. However, few data relevant to the Chinese population have been published. In present paper, detailed clinical and genetic re-evaluations were performed in members of four special families who had been initially diagnosed as atypical or asymptomatic FSHD based only on the D4Z4 repeat length analysis. The FSHD-sized D4Z4 repeats in the probands from families 1, 2 and 3 were identified as 4qB variants. These patients were further confirmed as limb-girdle muscular dystrophy (LGMD2) or myotonic dystrophy (DM1) by molecular analyses. Specifically, we identified a 4qB variant on chromosome 10 in the healthy members of the fourth FSHD family with complex D4Z4 rearrangements of two exchanged repeat arrays. For the first time, we demonstrated in the Chinese population that D4Z4 contractions on the 4qB variant do not cause FSHD and 4qB variant on chromosome 10 might also represent intermediate structures in the transition from 4q to 10q. Furthermore, our results emphasize that D4Z4 repeat length analysis alone is not sufficient for the diagnosis of FSHD, especially when used as an exclusion criterion. This analysis should be accompanied by 4qA/4qB variant determination and integrated chromosome assignments, especially in patients with obscure and unclassified myopathies similar to atypical forms of FSHD. 相似文献
