Development and Validation of a Deep Learning Predictive Model Combining Clinical and Radiomic Features for Short-Term Postoperative Facial Nerve Function in Acoustic Neuroma Patients

Objective This study aims to construct and validate a predictable deep learning model associated with clinical data and multi-sequence magnetic resonance imaging(MRI)for short-term postoperative facial nerve function in patients with acoustic neuroma.Methods A total of 110 patients with acoustic neuroma who underwent surgery through the retrosigmoid sinus approach were included.Clinical data and raw features from four MRI sequences(T1-weighted,T2-weighted,T1-weighted contrast enhancement,and T2-weighted-Flair images)were analyzed.Spearman correlation analysis along with least absolute shrinkage and selection operator regression were used to screen combined clinical and radiomic features.Nomogram,machine learning,and convolutional neural network(CNN)models were constructed to predict the prognosis of facial nerve function on the seventh day after surgery.Receiver operating characteristic(ROC)curve and decision curve analysis(DCA)were used to evaluate model performance.A total of 1050 radiomic parameters were extracted,from which 13 radiomic and 3 clinical features were selected.Results The CNN model performed best among all prediction models in the test set with an area under the curve(AUC)of 0.89(95%CI,0.84–0.91).Conclusion CNN modeling that combines clinical and multi-sequence MRI radiomic features provides excellent performance for predicting short-term facial nerve function after surgery in patients with acoustic neuroma.As such,CNN modeling may serve as a potential decision-making tool for neurosurgery.

FGF-2 is required to prevent astrogliosis in the facial nucleus after facial nerve injury and mechanical stimulation of denervated vibrissal muscles

Recently,we have shown that manual stimulation of paralyzed vibrissal muscles after facial-facial anastomosis reduced the poly-innervation of neuromuscular junctions and restored vibrissal whisking.Using gene knock outs,we found a differential dependence of manual stimulation effects on growth factors.Thus,insulin-like growth factor-1 and brain-derived neurotrophic factor are required to underpin manual stimulation-mediated improvements,whereas FGF-2 is not.The lack of dependence on FGF-2 in mediating these peripheral effects prompted us to look centrally,i.e.within the facial nucleus where increased astrogliosis after facial-facial anastomosis follows ＂synaptic stripping＂.We measured the intensity of Cy3-fluorescence after immunostaining for glial fibrillary acidic protein（GFAP） as an indirect indicator of synaptic coverage of axotomized neurons in the facial nucleus of mice lacking FGF-2（FGF-2^（-/-） mice）.There was no difference in GFAP-Cy3-fluorescence（pixel number,gray value range17-103） between intact wildtype mice（2.12± 0.37×10~7） and their intact FGF-2^（-/-） counterparts（2.12±0.27×10~7） nor after facial-facial anastomosis ＋handling（wildtype：4.06±0.32×10~7;FGF-2^（-/-）：4.39±0.17×10~7）.However,after facial-facial anastomosis,GFAP-Cy3-fluorescence remained elevated in FGF-2^（-/-）-animals（4.54±0.12×10~7）,whereas manual otimulation reduced the intensity of GFAP-immunofluorescence in wild type mice to values that were not significantly different from intact mice（2.63±0.39×10）.We conclude that FGF-2 is not required to underpin the beneficial effects of manual stimulation at the neuro-muscular junction,but it is required to minimize astrogliosis in the brainstem and,by implication,restore synaptic coverage of recovering facial motoneurons.

Electrical response grading versus House-Brackmannscale for evaluation of facial nerve injury after Bell's palsy: a comparative study

OBJECTIVE： There are no convenient techniques to evaluate the degree of facial nerve injury during a course of acupuncture treatment for Bell＇s palsy. Our previous studies found that observing the electrical response of specific facial muscles provided reasonable correlation with the prognosis of electroacupuncture treatment. Hence, we used the new method to evaluate the degree of facial nerve injury in patients with Bell＇s palsy in comparison with the House-Brackmann scale. The relationship between therapeutic effects and prognosis was analyzed to explore an objective method for evaluating Bell＇s palsy. METHODS： The facial nerve function of 68 patients with Bell＇s palsy was assessed with both electrical response grading and the House-Brackmann scale before treatment. Then differences in evaluation results of the two methods were compared. All enrolled patients received electroacupuncture treatment with disperse-dense wave at 1/100 Hz for 4 weeks. After treatment, correlation analysis was conducted to find the relationship between electrical response and therapeutic effects or prognosis. RESULTS： Checking consistency between electrical response grading and House-Brackmann scale： Kappa value 0.028 （P = 0.578）. Correlation analysis： the two methods were correlated with the prognosis, and electrical response grading （rER = 0.789） was better than the House-Brackmann scale （rHB = 0.423）. CONCLUSION： Electrical response grading is superior to the House-Brackmann scale in efficacy and reliability, and can conveniently assess the degree of facial nerve injury. The House-Brackmann scale is suitable for the patients with mild facial nerve injury, but its evaluation quality for severe facial nerve injury is poor.