Automatic classification of fetal heart rate based on a multi-scale LSTM network.

Introduction: Fetal heart rate monitoring during labor can aid healthcare professionals in identifying alterations in the heart rate pattern. However, discrepancies in guidelines and obstetrician expertise present challenges in interpreting fetal heart rate, including failure to acknowledge findings or misinterpretation. Artificial intelligence has the potential to support obstetricians in diagnosing abnormal fetal heart rates. Methods: Employ preprocessing techniques to mitigate the effects of missing signals and artifacts on the model, utilize data augmentation methods to address data imbalance. Introduce a multi-scale long short-term memory neural network trained with a variety of time-scale data for automatically classifying fetal heart rate. Carried out experimental on both single and multi-scale models. Results: The results indicate that multi-scale LSTM models outperform regular LSTM models in various performance metrics. Specifically, in the single models tested, the model with a sampling rate of 10 exhibited the highest classification accuracy. The model achieves an accuracy of 85.73%, a specificity of 85.32%, and a precision of 85.53% on CTU-UHB dataset. Furthermore, the area under the receiver operating curve of 0.918 suggests that our model demonstrates a high level of credibility. Discussion: Compared to previous research, our methodology exhibits superior performance across various evaluation metrics. By incorporating alternative sampling rates into the model, we observed improvements in all performance indicators, including ACC (85.73% vs. 83.28%), SP (85.32% vs. 82.47%), PR (85.53% vs. 82.84%), recall (86.13% vs. 84.09%), F1-score (85.79% vs. 83.42%), and AUC(0.9180 vs. 0.8667). The limitations of this research include the limited consideration of pregnant women's clinical characteristics and disregard the potential impact of varying gestational weeks.

[Protein gene product 9.5-immunoactive nerve fibers and its clinical significance in endometriotic peritoneal lesions].

OBJECTIVE: To investigate the association between distribution of protein gene product (PGP) 9.5-immunoactive nerve fibers in peritoneal endometriotic lesions and disease-associated pain symptoms. METHODS: Thirty two peritoneal endometriotic lesions from patients with endometriosis (16 cases with pain and 16 cases without pain) and matched with 20 peritoneal tissues from patients with uterine leiomyoma without endometriosis were stained immunohistochemically for PGP9.5-immunoactive nerve fibers. RESULTS: The positive rate and density of PGP9.5-immunoreactive nerve fibers in peritoneal endometriotic leision were 62% (10/16) and (3.8+/-1.7)/mm2 in endometriosis patients with pain, which were significantly higher than 19% (3/16) and (1.7+/-0.5)/mm2 in endometriosis patients without pain (P<0.05) and 25% (5/20) and (1.3+/-0.6)/mm2 in peritoneal tissues in women without endometriosis (P<0.05). However, no differences were found between endometriosis patients without pain and women without endometriosis (P>0.05). Moreover, the density of PGP9.5-immunoreactive nerve fibers in peritoneal lesions in endometriosis patients with pain was positively correlated with the severity of pain (r=0.855, P<0.05). In addition, the density of PGP9.5-immunoreactive nerve fibers in peritoneal lesions was statistically higher in endometriosis patients with chronic pelvic pain and (or) dysmenorrhea than those in endometriosis patients with other type of pain (P<0.05), which was not associated with active lesion, site and staging (P>0.05). CONCLUSION: It suggested that PGP9.5-immunoreactive nerve fibers might confer the mechanism of pelvic pain with endometriosis.

[Distribution of nerve fibers in endometrium and its clinical significance in adenomyosis].

OBJECTIVE: To investigate nerve fibers distribution in endometrium of adenomyosis and their relationship with dysmenorrhea. METHODS: Endometrial tissue was sampled from 74 hysterectomy specimens including 32 cases with adenomyosis and 42 cases with uterine fibroids. Two-step Envision immunohistochemical staining was used to detect distribution of nerve fibers in endometrium. Highly specific polyclonal rabbit anti-protein gene product 9.5 (PGP9.5) and monoclonal mouse anti-neurofilament protein (NF) were used to demonstrate both myelinated and unmyelinated nerve fibers in endometrium in women with adenomyosis and uterine fibroids. RESULTS: The positive rate of PGP9.5 immunoreactive nerve fibers in the functional layer of endometrium of pain patients were with 64% (14/22) in adenomyosis and 67% (10/15) in uterine fibroids. And their density were 0.6 (0 - 9.4)/mm(2) and 0.6 (0 - 6.0)/mm(2) without reaching statistical difference (P > 0.05). No expression of NF could be detected in the functional layer of endometrium of adenomyosis and uterine fibroids. There were no PGP9.5 immunoreactive nerve fibers in the functional layer of endometrium in non-pain women with adenomyosis and uterine fibroids. Moreover, No NF immunoreactive nerve fibers in the functional layer of endometrium were shown in non-pain patients with adenomyosis and uterine fibroids. PGP9.5 immunoreactive nerve fibers and the nerve density in the basal layer of endometrium were 64% (14/22), 1.1 (0 - 12.0)/mm(2) in pain adenomyosis and 50% (5/10), 0.6 (0 - 3.0)/mm(2) in non-pain adenomyosis. NF immunoreactive nerve fibers and the density in the basal layer of endometrium were 23% (5/22), (0 - 0.6)/mm(2) in pain adenomyosis and 20% (2/10), (0 - 1.0)/mm(2) in non-pain adenomyosis. PGP9.5 immunoreactive nerve fibers and the nerve density in the basal layer of endometrium were 80% (12/15) and 1.6 (0 - 10.0)/mm(2) in pain fibroids and 44% (12/27), 0 (0 - 5.0)/mm(2) in non-pain fibroids. NF immunoreactive nerve fibers and the nerve density in the basal layer of endometrium were 40% (6/15), 0 (0 - 0.4)/mm(2) in pain fibroids and 15% (4/27), 0 (0 - 1.0)/mm(2) in non-pain fibroids. There was no statistical different PGP9.5 and NF immunoreactive nerve fibers distribution in basal layer of endometrium between pain adenomyosis and pain fibroids or between non-pain adenomyosis and non-pain fibroids (all P > 0.05). However, PGP9.5 immunoreactive nerve fibers density in basal layer of endometrium was higher in pain adenomyosis and fibroids when compared with non-pain adenomyosis and fibroids (P < 0.05). CONCLUSIONS: PGP9.5 immunoreactive nerve fibers might confer the occurrence of pelvic pain, however, NF immunoreactive nerve fibers may not involved in the pathogenesis of pain.

[Associations of metabolism of lipid, calcium and phosphate in endometriosis].

OBJECTIVE: To investigate the metabolism of lipid, calcium and phosphorus in women with endometriosis. METHODS: Clinical data of 223 patients with endometriosis and 200 patients without endometriosis were retrospectively analyzed. Electrochemiluminoimmunoassay was used to detect the levels of serum luteinizing hormone (LH), follicle-stimulating hormone (FSH), testosterone (T), estradiol (E2) and progesterone (P), and photoelectric colorimetry was used to determine the concentrations of serum triglyceride (TG), total cholesterol (TC), low density lipoprotein (LDL), high density lipoprotein (HDL), calcium and phosphorus in the patients with and without endometriosis. RESULTS: The levels of serum P were 2.0 nmol/L and 1.8 nmol/L in endometriosis patients with age < or = 35 years old (case group A) and > 35 years old (case group B), and 1.6 nmol/L and 1.2 nmol/L in patients without endometriosis at age < or = 35 years old (control group A) and > 35 years old (control group B), respectively. A significant difference was found between case group A and control group A, and between case group B and control group B. The levels of serum LH was significantly higher in case group B (7.2 U/L) than that in control group B (6.1 U/L), however, there was no significant difference between case group A (7.0 U/L) and control group A (6.5 U/L). Also no significant difference in serum FSH, T and E2 levels was found between case group A (respectively, 6.2 U/L, 1.1 nmol/L and 269 pmol/L) and control group A (respectively, 6.3 U/L, 1.1 nmol/L and 264 pmol/L), also between case group B (respectively, 6.6 U/L, 1.0 nmol/L and 345 pmol/L) and control group B (6.6 IU/L, 0.9 nmol/L and 279 pmol/L). The concentrations of serum TG in case group A and B (0.71 mmol/L and 0.72 mmol/L) were significantly lower than in control group A and B (0.92 mmol/L and 1.08 mmol/L), respectively. The concentrations of serum LDL in case group A and B [(2.2 +/- 0.5) mmol/L and (2.4 +/- 0.6) mmol/L]were also significantly lower than in control group A and B [(2.4 +/- 0.7) mmol/L and (2.62 +/- 0.63) mmol/L], respectively. However, the concentrations of serum HDL in case group A and B [(1.62 +/- 0.31) mmol/L and (1.53 +/- 0.32) mmol/L] were significantly higher than in control group A and B [(1.48 +/- 0.21) mmol/L and (1.37 +/- 0.22) mmol/L] , respectively. In addition, the concentrations of serum TC were not significantly different between case group A and control group A [(4.2 +/- 0.7) mmol/L and (4.29 +/- 0.71) mmol/L], and between case group B and control group B [(4.4 +/- 0.8) mmol/L and (4.5 +/- 0.7) mmol/L]. The levels of serum phosphorus in case group A and B [(1.01 +/- 0.22) mmol/L and (0.89 +/- 0.18 mmol/L] were significantly lower than in control group A and B [1.23 +/- 0.24 mmol/L and (1.10 +/- 0.13) mmol/L], respectively. But the levels of serum calcium had no significant difference between case group A and control group A [(2.33 +/- 0.23) mmol/L and (2.41 +/- 0.12) mmol/L], and between case group B and control group B [(2.40 +/- 0.28) mmol/L and (2.42 +/- 0.20) mmol/L]. CONCLUSION: The abnormal metabolism of lipid and phosphorus, and the higher levels of serum P may playing a role in the pathogenesis of endometriosis.