1.Automatic sleep staging algorithm for stochastic depth residual networks based on transfer learning.
Yunzhi TIAN ; Qiang ZHOU ; Wan LI
Journal of Biomedical Engineering 2023;40(2):286-294
The existing automatic sleep staging algorithms have the problems of too many model parameters and long training time, which in turn results in poor sleep staging efficiency. Using a single channel electroencephalogram (EEG) signal, this paper proposed an automatic sleep staging algorithm for stochastic depth residual networks based on transfer learning (TL-SDResNet). Firstly, a total of 30 single-channel (Fpz-Cz) EEG signals from 16 individuals were selected, and after preserving the effective sleep segments, the raw EEG signals were pre-processed using Butterworth filter and continuous wavelet transform to obtain two-dimensional images containing its time-frequency joint features as the input data for the staging model. Then, a ResNet50 pre-trained model trained on a publicly available dataset, the sleep database extension stored in European data format (Sleep-EDFx) was constructed, using a stochastic depth strategy and modifying the output layer to optimize the model structure. Finally, transfer learning was applied to the human sleep process throughout the night. The algorithm in this paper achieved a model staging accuracy of 87.95% after conducting several experiments. Experiments show that TL-SDResNet50 can accomplish fast training of a small amount of EEG data, and the overall effect is better than other staging algorithms and classical algorithms in recent years, which has certain practical value.
Humans
;
Sleep Stages
;
Algorithms
;
Sleep
;
Wavelet Analysis
;
Electroencephalography/methods*
;
Machine Learning
2.Synchronous occurrence of digestive tract cancer and gastrointestinal stromal tumor
Changqing LU ; Yunzhi SHEN ; Bo TIAN
Chinese Journal of Digestive Endoscopy 1996;0(06):-
Objective To investigate the clinicopathological and immunohistochemical expression in 8 cases with synchronous occurrence of digestive tract cancer and GIST. Methods Clinical and pathologic data of 8 out of 70 cases of GIST coexisted with gastrointestinal cancer were recorded, and immunohistochemical stain ( Envision method) was used to detect the expressions of CD117, CD34 ,Vimenlin etc. Results In 8 cases of GIST (4 benign, 3 borderline, 1 malignant) coexisted with GI carcinoma, of which 6 cases featured the simultaneous occurrence of GIST and adenocarcinomas (5 cases in slomach and 1 case in colon) , whereas the other 2 cases associated with squamus cell carcinoma of esophagus. In 4 cases, tumors arose from the same organ, in the other 4 cases they developed from different sites of digestive tract. In all of the cases two kinds of neoplasms were developed in different areas without any connection, preoperative endoscopy and biopsy showed adenocarcinomas and squamous cell carcinoma, but histologic and clinical diagnosis of coexisted tumors were not achieved in any case. All GIST expressed Vimentin and CD34; 7 cases expressed CD117; 2 cases were focally positive for S-100 protein and 1 each for SMA and Desmin. Conclusion The simultaneous occurrence of epithelial tumor and GIST in digestive tract is not less than that usually expected, in clinical practice we should pay much attention on the diagnosis of this entity. Endoscopic biopsy conjugated with profound inspection during operation and pathological examination of specimen postoperatively can improve the possibility of early diagnosis.

Result Analysis
Print
Save
E-mail