a-novel-dual-modal-deep-learning-approach-for-real-time-removal

A Novel Dual-Modal Deep Learning Approach for Real-Time Removal of Hepatic Fluorescence in Indocyanine Green-Guided Laparoscopic Cholecystectomy

Indocyanine green-guided laparoscopic cholecystectomy (ICGLC) improves biliary visualization but is often hindered by hepatic fluorescence contamination, ...

Chinese Academy of Medical Sciences & Peking Union Medical College

Laparoscopic pancreatic surgery remains highly challenging due to the complexity of the pancreas and surrounding vascular structures, with risk of injuring critical blood vessels such as the Superior Mesenteric Vein (SMV)-Portal Vein (PV) axis and splenic vein. Here, we evaluated the High Resolution Network (HRNet)-Full Convolutional Network (FCN) model for its ability to accurately identify vascular contours and improve surgical safety. Using 12,694 images from 126 laparoscopic distal pancreatectomy (LDP) videos and 35,986 images from 138 Whipple procedure videos, the model demonstrated robust performance, achieving a mean Dice coefficient of 0.754, a recall of 85.00%, and a precision of 91.10%. By combining datasets from LDP and Whipple procedures, the model showed strong generalization across different surgical contexts and achieved real-time processing speeds of 11 frames per second during surgery process. These findings highlight HRNet-FCN's potential to recognize anatomical landmarks, enhance surgical precision, reduce complications, and improve laparoscopic pancreatic outcomes.

Deep learning HRNet FCN for blood vessel identification in laparoscopic pancreatic surgery.

Despite the rapid progress of robotic surgery, intelligent object recognition and autonomous computer-assisted interventions continues to face limitations. Algorithms designed for the identification and localization of surgical tool tips are pivotal in augmenting the interactive collaboration between the robotic surgical system and its operator. In the present study, our aim to formulate a strategy proficient in recognizing and pinpointing commonly used surgical tools in robotic surgery.

Development and validation of a surgical tool recognition and localization strategy in robotic surgeries.

Laparoscopic operation holds multiple advantages as a minimal invasive method of surgical treatment. Vascular-related manipulations, including identification and dissection of vascular structures and control of bleeding, are highly experience-based and demand a tortuous learning curve. With the rapid development of artificial intelligence (AI) in the entire diagnosis and treatment process of diseases, data-driven AI models have shown promising potentials in both education and real-time aiding in surgery. However, there is no dedicated dataset existing for developing vascular identification models in laparoscopic settings.

Construction and validation of a high-precision annotated dataset for developing intelligent critical vein recognition models in laparoscopic pancreatic surgery.

The integration of artificial intelligence (AI) into surgical navigation represents a pivotal advancement in modern operative medicine. In endoscopic thyroidectomy, safeguarding the recurrent laryngeal nerve (RLN) is of critical importance due to its vulnerability to iatrogenic injury, which affects 3-8% of cases and can lead to serious complications such as vocal cord paralysis. However, existing intraoperative nerve monitoring (IONM) technologies are limited by high costs, operator dependence, and discontinuous signal acquisition. To address the lack of large-scale, annotated datasets essential for training robust deep learning models in real-world surgical settings, we present ThyRLN-PUMCH, the first comprehensive in vivo dataset dedicated to RLN identification in endoscopic thyroid surgery. This dataset comprises 18,178 pixel-level annotated frames from 28 clinically diverse surgical cases. Annotations were performed and validated by board-certified endocrine surgeons through a multi-stage quality control process. We benchmarked two segmentation models to verify their practicability and proved the dataset's capacity to support high-precision RLN segmentation tasks. ThyRLN-PUMCH fills a critical gap in AI assisted head and neck surgery by offering temporally continuous, clinically representative images and annotations. It provides a robust foundation for developing AI-based intraoperative navigation tools aimed at enhancing surgical safety, education, and efficiency in head and neck surgery.

Article	Year
Deep learning HRNet FCN for blood vessel identification in laparoscopic pancreatic surgery. NPJ digital medicine\| PubMed ID: 40312536	2025
Development and validation of a surgical tool recognition and localization strategy in robotic surgeries. Journal of robotic surgery\| PubMed ID: 41175306	2025
Construction and validation of a high-precision annotated dataset for developing intelligent critical vein recognition models in laparoscopic pancreatic surgery. Frontiers in surgery\| PubMed ID: 41704439	2026
A High-Quality Endoscopic Image Dataset with Annotated Recurrent Laryngeal Nerve for AI-Assisted Thyroid Surgery. Scientific data\| PubMed ID: 41932923	2026

Ruohan Cui

Biography

JoVE Journal Publications

Other Publications