Related Experiment Video
Updated: Jun 13, 2026
07:43
Endoscopic Endonasal Trans-sphenoidal Approach: Minimally Invasive Surgery for Pituitary Adenomas
Published on: January 17, 2018
PituPhase65: An endoscopic pituitary surgery dataset for surgical phase recognition
Ángela González-Cebrián1, Alba Centeno López1, Igor Paredes2,3,4
1Computer Science and Engineering Department, Universidad Carlos III de Madrid, Madrid, Spain.
Scientific Data
|June 11, 2026
Summary
This study introduces PituPhase65, a new dataset for surgical phase recognition in endoscopic pituitary surgery. This resource aims to improve AI model training for this complex procedure.
Area of Science:
- Neurosurgery
- Medical Imaging
- Artificial Intelligence
Background:
- Surgical Phase Recognition (SPR) offers real-time feedback to surgical teams.
- Standardized surgeries like cholecystectomy benefit from SPR, but endoscopic pituitary surgery presents challenges due to workflow variability.
- Developing AI models for pituitary surgery requires robust, labeled datasets.
Purpose of the Study:
- To address the need for labeled data in endoscopic pituitary surgery.
- To facilitate the development of AI models for Surgical Phase Recognition (SPR) in this domain.
- To introduce the PituPhase65 dataset.
Main Methods:
- Creation of the PituPhase65 dataset, comprising 65 endoscopic pituitary surgeries.
- Data collection from Hospital Universitario 12 de Octubre (Madrid, Spain).
- Labeling of surgeries into eight distinct phases, from initial setup to dural closure.
Main Results:
- PituPhase65 is a publicly available dataset.
- The dataset includes 65 surgeries with an average duration of 90 minutes.
- Surgeries are meticulously labeled into eight phases: out of patient, middle turbinate resection, nasoseptal flap preparation, ethmoidectomy and sphenoidal sinus opening, sellar opening, dural opening, tumor resection, and dural closure.
Conclusions:
- The PituPhase65 dataset is crucial for advancing SPR in endoscopic pituitary surgery.
- Availability of this dataset will aid in training AI models and reducing the learning curve for residents.
- This resource supports improved operating room communication and automated surgical reporting.