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

Abstract

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.

References

[1] KoikeH, MorikawaM, IshimaruH, et al.. Quantitative Chemical Exchange Saturation Transfer Imaging of Amide Proton Transfer Differentiates between Cerebellopontine Angle Schwannoma and Meningioma: Preliminary Results. IJMS, 2022, 23(17): 10187

[2] EgizA, NautiyalH, AlaladeAF, et al.. Evaluating growth trends of residual sporadic vestibular schwannomas: a systematic review and meta-analysis. J Neurooncol, 2022, 159(1): 135-150

[3] YaşargilMG. The internal acoustic meatus. J Neurosurg, 2002, 97(5): 1014-1017

[4] JungGS, MontibellerGR, FragaG de, et al.. Facial Nerve Adherence in Vestibular Schwannomas: Classification and Radiological Predictors. J Neurol Surg B Skull Base, 2020, 82(4): 456-460

[5] FishmanZ, KissA, ZukerRM, et al.. Measuring 3D facial displacement of increasing smile expressions. J Plast Reconstr Aesthet Surg, 2022, 75(11): 4273-4280

[6] Bali ZU, Tuluy Y, Özkaya Ünsal M, et al. The evaluation of the effect of free gracilis muscle transfer on cheek tone and oral competence in long-standing facial paralysis of patients by using Blasco index. Microsurgery, Online ahead of print. October 19, 2022:micr.30976

[7] KhanNR, ElarjaniT, JamshidiAM, et al.. Microsurgical Management of Vestibular Schwannoma (Acoustic Neuroma): Facial Nerve Outcomes, Radiographic Analysis, Complications, and Long-Term Follow-Up in a Series of 420 Surgeries. World Neurosurg, 2022, 168: e297-e308

[8] OsthuesM, KuttenreichAM, VolkGF, et al.. Continual rehabilitation motivation of patients with postparalytic facial nerve syndrome. Eur Arch Otorhinolaryngol, 2022, 279(1): 481-491

[9] PattinsonR, PooleHM, ShorthouseO, et al.. Exploring beliefs and distress in patients with facial palsies. Psychol Health Med, 2022, 27(4): 788-802

[10] CrossT, SheardCE, GarrudP, et al.. Impact of facial paralysis on patients with acoustic neuroma. Laryngoscope, 2000, 110(9): 1539-1542

[11] HuangX, XuJ, XuM, et al.. Functional outcome and complications after the microsurgical removal of giant vestibular schwannomas via the retrosigmoid approach: a retrospective review of 16-year experience in a single hospital. BMC Neurol, 2017, 17(1): 18

[12] HirutaR, SatoT, ItakuraT, et al.. Intraoperative transcranial facial motor evoked potential monitoring in surgery of cerebellopontine angle tumors predicts early and late postoperative facial nerve function. Clin Neurophysiol, 2021, 132(4): 864-871

[13] MastronardiL, GazzeriR, BarbieriFR, et al.. Postoperative Functional Preservation of Facial Nerve in Cystic Vestibular Schwannoma. World Neurosurg, 2020, 143: e36-e43

[14] TorresR, NguyenY, VanierA, et al.. Multivariate Analysis of Factors Influencing Facial Nerve Outcome following Microsurgical Resection of Vestibular Schwannoma. Otolaryngol Head Neck Surg, 2017, 156(3): 525-533

[15] TahaI, HyvärinenA, RantaA, et al.. Facial nerve function and hearing after microsurgical removal of sporadic vestibular schwannomas in a population-based cohort. Acta Neurochir, 2020, 162(1): 43-54

[16] SalehE, PiccirilloE, MigliorelliA, et al.. Wait and Scan Management of Intra-canalicular Vestibular Schwannomas: Analysis of Growth and Hearing Outcome. Otol Neurotol, 2022, 43(6): 676-684

[17] CarlstromLP, Muñoz-CasabellaA, PerryA, et al.. Dramatic Growth of a Vestibular Schwannoma After 16 Years of Postradiosurgery Stability in Association With Exposure to Tyrosine Kinase Inhibitors. Otol Neurotol, 2021, 42(10): e1609-e1613

[18] YangHC, WuCC, LeeCC, et al.. Prediction of pseudoprogression and long-term outcome of vestibular schwannoma after Gamma knife radiosurgery based on preradiosurgical MR radiomics. Radiother Oncol, 2021, 155: 123-130

[19] LiuJ, SunD, ChenL, et al.. Radiomics Analysis of Dynamic Contrast-Enhanced Magnetic Resonance Imaging for the Prediction of Sentinel Lymph Node Metastasis in Breast Cancer. Front Oncol, 2019, 9: 980

[20] ZhangZ, LiX, SunH. Development of machine learning models integrating PET/CT radiomic and immunohistochemical pathomic features for treatment strategy choice of cervical cancer with negative pelvic lymph node by mediating COX-2 expression. Front Physiol, 2022, 13: 994304

[21] ZhuF, ZhuZ, ZhangY, et al.. Severity detection of COVID-19 infection with machine learning of clinical records and CT images. Technol Health Care, 2022, 30(6): 1299-1314

[22] ZengQ, LiH, ZhuY, et al.. Development and validation of a predictive model combining clinical, radiomics, and deep transfer learning features for lymph node metastasis in early gastric cancer. Front Med, 2022, 9: 986437

[23] YangH, YanS, LiJ, et al.. Prediction of acute versus chronic osteoporotic vertebral fracture using radiomics-clinical model on CT. Eur J Radiol, 2022, 149: 110197

[24] LangenhuizenPPJH, ZingerS, LeenstraS, et al.. Radiomics-Based Prediction of Long-Term Treatment Response of Vestibular Schwannomas Following Stereotactic Radiosurgery. Otol Neurotol, 2020, 41(10): e1321-e1327

[25] LeeWK, WuCC, LeeCC, et al.. Combining analysis of multi-parametric MR images into a convolutional neural network: Precise target delineation for vestibular schwannoma treatment planning. Artif Intell Med, 2020, 107: 101911

[26] BoublataL, BelahrecheM, OuchtatiR, et al.. Facial Nerve Function and Quality of Resection in Large and Giant Vestibular Schwannomas Surgery Operated By Retrosigmoid Transmeatal Approach in Semi-sitting Position with Intraoperative Facial Nerve Monitoring. World Neuros, 2017, 103: 231-240

[27] LinkMJ, DriscollCLW, FengY, et al.. Retrosigmoid Approach for Resection of Large Cystic Vestibular Schwannoma. J Neurol Surg B, 2019, 80(S03): S285-S285

[28] MoonKS, JungS, SeoSK, et al.. Cystic vestibular schwannomas: a possible role of matrix metalloproteinase-2 in cyst development and unfavorable surgical outcome. JNS, 2007, 106(5): 866-871

[29] LeeCC, LeeWK, WuCC, et al.. Applying artificial intelligence to longitudinal imaging analysis of vestibular schwannoma following radiosurgery. Sci Rep, 2021, 11(1): 3106

[30] MastronardiL, CampioneA, BoccacciF, et al.. Koos grade IV vestibular schwannomas: considerations on a consecutive series of 60 cases—searching for the balance between preservation of function and maximal tumor removal. Neurosurg Rev, 2021, 44(6): 3349-3358

[31] LiuC, ShenY, HanD, et al.. Analysis of Related Factors Affecting Facial Nerve Function after Acoustic Neuroma Surgery. Evid Based Complement Alternat Med, 2022, 2022: 1-6

[32] ElsayedM, JiaH, HochetB, et al.. Intraoperative facial nerve electromyography parameters to optimize postoperative facial nerve outcome in patients with large unilateral vestibular schwannoma. Acta Neurochir, 2021, 163(8): 2209-2217

[33] George-JonesNA, ChkheidzeR, MooreS, et al.. MRI Texture Features are Associated with Vestibular Schwannoma Histology. Laryngoscope, 2021, 131(6): E2000-E2006

[34] ZhangR, WeiY, ShiF, et al.. The diagnostic and prognostic value of radiomics and deep learning technologies for patients with solid pulmonary nodules in chest CT images. BMC Cancer, 2022, 22(1): 1118

[35] LinQ, WuHJ, SongQS, et al.. CT-based radiomics in predicting pathological response in non-small cell lung cancer patients receiving neoadjuvant immunotherapy. Front Oncol, 2022, 12: 937277

RIGHTS & PERMISSIONS

Huazhong University of Science and Technology