Artificial Intelligence for Surgical Safety: Automatic... : Annals of Surgery (original) (raw)

ORIGINAL ARTICLES

Automatic Assessment of the Critical View of Safety in Laparoscopic Cholecystectomy Using Deep Learning

Mascagni, Pietro MD∗,†; Vardazaryan, Armine MSc∗; Alapatt, Deepak MSc∗; Urade, Takeshi MD, PhD‡; Emre, Taha MSc∗; Fiorillo, Claudio MD†; Pessaux, Patrick MD, PhD‡,§,¶; Mutter, Didier MD, PhD§,¶; Marescaux, Jacques MD, FACS (Hon), FRCS (Hon), FJSES§; Costamagna, Guido MD, PhD†; Dallemagne, Bernard MD§,¶; Padoy, Nicolas PhD∗

∗ICube, University of Strasbourg, CNRS, IHU Strasbourg, France

†Fondazione Policlínico Universitario A. Gemelli IRCCS, Rome, Italy

‡IHU-Strasbourg, Institute of Image-Guided Surgery, Strasbourg, France

§Institute for Research against Digestive Cancer (IRCAD), Strasbourg, France

¶Department of Digestive and Endocrine Surgery, University of Strasbourg, Strasbourg, France.

[email protected].

P.M., A.V., and D.A. contributed equally and share first co-authorship.

This study is partially supported by an EAES Research Grant and by French State Funds managed by the Agence Nationale de la Recherche (ANR) through the Investissements d’Avenir Program under Grant ANR-11-LABX-0004 (Labex CAMI), Grant ANR-10-IDEX-0002-02 (Idex Unistra), and Grant ANR-10-IAHU-02 (IHU-Strasbourg), and by BPI France through Project CONDOR.

Authors’ Contributions: Pietro Mascagni, MD: Participated in the study conception and design; data acquisition, annotation and interpretation; manuscript drafting; and approved the final version of the manuscript. Armine Vardazaryan, MSc: Participated in the study design; data acquisition and interpretation; manuscript drafting; and approved the final version of the manuscript. Deepak Alapatt, MSc, Takeshi Urade, MD, PhD, Claudio Fiorillo, MD, Patrick Pessaux, MD, PhD, Didier Mutter, MD, PhD, Jacques Marescaux, MD, FACS, (Hon) FRCS, (Hon) FJSES, Guido Costamagna, MD, PhD: Participated in data interpretation; manuscript revision; and approved the final version of the manuscript. Taha Emre, MSc: Participated in the study design; data acquisition and interpretation; manuscript revision; and approved the final version of the manuscript. Bernard Dallemagne, MD: Participated in the study conception and design; data acquisition, annotation and interpretation; manuscript revision; and approved the final version of the manuscript. Nicolas Padoy, PhD: Supervised the whole project; participated in the study conception and design; data interpretation; manuscript revision; and approved the final version of the manuscript.

The authors declare no conflict of interest.

Abstract

Objective:

To develop a deep learning model to automatically segment hepatocystic anatomy and assess the criteria defining the critical view of safety (CVS) in laparoscopic cholecystectomy (LC).

Background:

Poor implementation and subjective interpretation of CVS contributes to the stable rates of bile duct injuries in LC. As CVS is assessed visually, this task can be automated by using computer vision, an area of artificial intelligence aimed at interpreting images.

Methods:

Still images from LC videos were annotated with CVS criteria and hepatocystic anatomy segmentation. A deep neural network comprising a segmentation model to highlight hepatocystic anatomy and a classification model to predict CVS criteria achievement was trained and tested using 5-fold cross validation. Intersection over union, average precision, and balanced accuracy were computed to evaluate the model performance versus the annotated ground truth.

Results:

A total of 2854 images from 201 LC videos were annotated and 402 images were further segmented. Mean intersection over union for segmentation was 66.6%. The model assessed the achievement of CVS criteria with a mean average precision and balanced accuracy of 71.9% and 71.4%, respectively.

Conclusions:

Deep learning algorithms can be trained to reliably segment hepatocystic anatomy and assess CVS criteria in still laparoscopic images. Surgical-technical partnerships should be encouraged to develop and evaluate deep learning models to improve surgical safety.