Abstract
Objective
To investigate the feasibility of a deep learning method based on a UNETR model for
fully automatic segmentation of the cochlea in temporal bone CT images.
Methods
The normal temporal bone CTs of 77 patients were used in 3D U-Net and UNETR model
automatic cochlear segmentation. Tests were performed on two types of CT datasets
and cochlear deformity datasets.
Results
Through training the UNETR model, when batch_size=1, the Dice coefficient of the normal
cochlear test set was 0.92, which was higher than that of the 3D U-Net model; on the
GE 256 CT, SE-DS CT and Cochlear Deformity CT dataset tests, the Dice coefficients
were 0.91, 0.93, 0 93, respectively.
Conclusion
According to the anatomical characteristics of the temporal bone, the use of the UNETR
model can achieve fully automatic segmentation of the cochlea and obtain an accuracy
close to manual segmentation. This method is feasible and has high accuracy.
Keywords
Abbreviations:
CT (computed tomography), IP (Incomplete partition), SNHL (sensorineural hearing loss), CIs (cochlear implants), MRI (magnetic resonance imaging), AI (artificial intelligence), DL (deep learning)To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to Auris Nasus LarynxAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- Pediatric sensorineural hearing loss, part 1: practical aspects for neuroradiologists.AJNR Am J Neuroradiol. 2012; 33: 211-217
- Audiologic and radiologic findings in cochlear hypoplasia.Auris Nasus Larynx. 2017; 44: 655-663
- Unpartitioned versus incompletely partitioned cochleae: radiologic differentiation.Otol Neurotol. 2004; 25 (discussion 529): 520-529
- A new classification for cochleovestibular malformations.Laryngoscope. 2002; 112: 2230-2241
- Cochlear implants in adults and children.NIH Consens Statement. 1995; 13: 1-30
- The mondini dysplasia-from early diagnosis to cochlear implant[J].Acta Otolaryngol. 1983; 95: 627-631
- Clinical outcomes following cochlear implantation in children with inner ear anomalies.Int J Pediatr Otorhinolaryngol. 2017; 93: 1-6
- Real-time intraoperative computed tomography to assist cochlear implant placement in the malformed inner ear.Otol Neurotol. 2009; 30: 23-26
- Cochlear length determination using cone beam computed tomography in a clinical setting.Hear Res. 2014; 316: 65-72
- Impact of electrode insertion depth on intracochlear trauma.Otolaryngol Head Neck Surg. 2006; 135: 374-382
- Cochlear implantation in inner ear malformations–a review article.Cochlear Implant Int. 2010; 11: 4-41
- Sensorineural and conductive hearing loss associated with lateral semicircular canal malformation.Laryngoscope. 2000; 110: 1673-1679
- Robotic middle ear access for cochlear implantation: first in man.PLoS One. 2019; 14e0220543
- Instrument flight to the inner ear.Sci Robot. 2017; 2: 1-12
- Robotic cochlear implant surgery: imaging-based evaluation of feasibility in clinical routine.Front Surg. 2021; 8742219
- Multi-scale deep learning framework for cochlea localization, segmentation and analysis on clinical ultra-high-resolution CT images.Comput Methods Programs Biomed. 2020; 191105387
- A 3D deep supervised densely network for small organs of human temporal bone segmentation in CT images.Neural Netw. 2020; 124: 75-85
- Toward an automatic preoperative pipeline for image-guided temporal bone surgery.Int J Comput Assist Radiol Surg. 2019; 14: 967-976
- Retrospective in silico evaluation of optimized preoperative planning for temporal bone surgery.Int J Comput Assist Radiol Surg. 2020; 15: 1825-1833
- Automated measurement of hydrops ratio from MRI in patients with Meniere's disease using CNN-based segmentation.Sci Rep. 2020; 10: 7003
- Automatic detection of the inner ears in head CT images using deep convolutional neural networks.Proc SPIE Int Soc Opt Eng. 2018; 105741057427
- PWD-3DNet: a deep learning-based fully-automated segmentation of multiple structures on temporal bone CT scans.IEEE Trans Image Process. 2021; 30: 739-753
- Automatic semicircular canal segmentation of CT volumes using improved 3D U-net with attention mechanism.Comput Intell Neurosci. 2021; 20219654059
- Automatic segmentation of temporal bone structures from clinical conventional CT using a CNN approach.Int J Med Robot. 2021; 17: e2229
- Fully automated segmentation in temporal bone CT with neural network: a preliminary assessment study.BMC Med Imaging. 2021; 21: 166
- Automatic segmentation of inner ear on CT-scan using auto-context convolutional neural network.Sci Rep. 2021; 11: 4406
- Deep learning for the fully automated segmentation of the inner ear on MRI.Sci Rep. 2021; 11: 2885
- 3D U-Net: learning dense volumetric segmentation from sparse annotation.(ed.)in: Ourselin S Joskowicz L Sabuncu M Unal G Wells W Medical image computing and computer-assisted intervention, lecture notes in computer science. Springer, Cham, Switzerland2016: 424-432
Ali H, Yucheng T, Vishwesh N, Dong Y, Andriy M, Bennett L, et al. UNETR: transformers for 3D medical image segmentation. arXiv preprint arXiv:2103.10504, 2021.
- Classification and current management of inner ear malformations.Balkan Med J. 2017; 34: 397-411
- Augmented reality for inner ear procedures: visualization of the cochlear central axis in microscopic videos.Int J Comput Assist Radiol Surg. 2020; 15: 1703-1711
Article info
Publication history
Published online: August 12, 2022
Accepted:
June 30,
2022
Received:
April 24,
2022
Identification
Copyright
© 2022 Japanese Society of Otorhinolaryngology-Head and Neck Surgery, Inc. Published by Elsevier B.V. All rights reserved.
