Surgical data science - from concepts toward clinical translation

@article{MaierHein2022SurgicalDS,
  title={Surgical data science - from concepts toward clinical translation},
  author={Lena Maier-Hein and Matthias Eisenmann and Duygu Sarikaya and Keno Marz and Toby Collins and Anand Malpani and Johannes Fallert and Hubertus Feu{\ss}ner and Stamatia Giannarou and Pietro Mascagni and Hirenkumar Nakawala and Adrian E. Park and Carla M. Pugh and Danail Stoyanov and S. Swaroop Vedula and Kevin Cleary and Gabor Fichtinger and Germain Forestier and Bernard Gibaud and Teodor P. Grantcharov and Makoto Hashizume and Doreen Heckmann-Notzel and Hannes G Kenngott and Ron Kikinis and Lars Mundermann and Nassir Navab and Sinan Onogur and Raphael Sznitman and Russell H. Taylor and Minu Dietlinde Tizabi and Martin Wagner and Gregory Hager and Thomas Neumuth and Nicolas Padoy and Justin William Collins and Ines Gockel and Jan Goedeke and Daniel A. Hashimoto and Luc Joyeux and Kyle Lam and Daniel Richard Leff and Amin Madani and Hani J. Marcus and Ozanan R. Meireles and Alexander Seitel and Dogu Teber and Frank Uckert and Beat P. Muller-Stich and Pierre Jannin and Stefanie Speidel},
  journal={Medical image analysis},
  year={2022},
  volume={76},
  pages={
          102306
        }
}
The importance of machine learning in autonomous actions for surgical decision making
TLDR
Basic ML concepts are introduced as enablers for autonomous actions in surgery, highlighting examples for such actions along the surgical treatment path, by means of ML methods.
Reducing prediction volatility in the surgical workflow recognition of endoscopic pituitary surgery
TLDR
The results confirm the established methods generalise to endoscopic pituitary surgery, and show simple temporal smoothing not only reduces prediction volatility, but actively improves performance.
Fluorescence angiography classification in colorectal surgery - A preliminary report
TLDR
An automated classification of intraoperative fluorescence angiography with a high degree of accuracy is possible and allows automated decision boundary identiﷁcation, which will enable surgeons to standardise the technique of fferent surgeons.
Image-Guided Interventional Robotics: Lost in Translation?
TLDR
This article aims to embrace the state of the art, provide a comprehensive narrative of the situation described, support future system developers, and facilitate the translation from scientific research to applied clinical technology development.
Breaking down the silos of artificial intelligence in surgery: glossary of terms
TLDR
A glossary with definitions of the commonly used AI terms in surgery to improve their understanding by surgeons is provided to provide a glossary of AI terms used by residents and attending consultant surgeons.
Artificial Intelligence Methods and Artificial Intelligence-Enabled Metrics for Surgical Education: A Multidisciplinary Consensus.
TLDR
The Delphi panel consensus provides a specific, bold, and forward-looking roadmap for AI methods and AI-enabled metrics for surgical education.
On the Pitfalls of Batch Normalization for End-to-End Video Learning: A Study on Surgical Workflow Analysis
TLDR
It is argued that BN’s properties create major obstacles for training CNNs and temporal models end to end in video tasks, and it is shown that even simple, endto-end CNN-LSTMs can outperform the state of the art when CNNs without BN are used.
Federated Cycling (FedCy): Semi-supervised Federated Learning of Surgical Phases
TLDR
FedCy is proposed, a federated semi-supervised learning (FSSL) method that combines FL and self-super supervised learning to exploit a decentralized dataset of both labeled and unlabeled videos, thereby improving performance on the task of surgical phase recognition.
Data-centric multi-task surgical phase estimation with sparse scene segmentation
TLDR
A simple multi-task encoder is proposed that effectively fuses both streams, when available, based on their importance and jointly optimise them for performing accurate phase prediction and it is shown that with a small fraction of scene segmentation annotations, a relatively simple model can obtain comparable results than previous state-of-the-art and more complex architectures when evaluated in similar settings.
Endoscopic Imaging Technology Today
TLDR
This survey gives an overview of modern endoscopic technology’s state of the art and some trends for upcoming developments were collected.
...
...

References

SHOWING 1-10 OF 427 REFERENCES
Surgical Data Science: A Consensus Perspective
TLDR
This document summarizes the key findings of the 1st workshop on Surgical Data Science, which brought together researchers working on diverse topics in surgical data science in order to discuss existing challenges, potential standards and new research directions in the field.
The Value of Surgical Data—Impact on the Future of the Surgical Field
TLDR
The authors attempt to characterize the various ways in which surgical data are collected and shared and conclude that an attempt to outline best practices should be undertaken that benefits all stakeholders.
Deep learning for healthcare: review, opportunities and challenges
TLDR
It is suggested that deep learning approaches could be the vehicle for translating big biomedical data into improved human health and develop holistic and meaningful interpretable architectures to bridge deep learning models and human interpretability.
Causality matters in medical imaging
TLDR
The authors show how causal reasoning can shed new light on scarcity of high-quality annotated data and mismatch between the development dataset and the target environment, and show step-by-step recommendations for future studies.
CAI4CAI: The Rise of Contextual Artificial Intelligence in Computer-Assisted Interventions
TLDR
Central challenges being addressed in CAI4CAI include how to integrate the ensemble of prior knowledge and instantaneous sensory information from experts, sensors, and actuators; how to create and communicate a faithful and actionable shared representation of the surgery among a mixed human–AI actor team.
Toward knowledge-based liver surgery: holistic information processing for surgical decision support
TLDR
A formalized dynamic patient model that incorporates all the heterogeneous data acquired for a specific patient in the whole course of disease treatment and a technical infrastructure that enables storing, accessing, and processing of heterogeneousData to support clinical decision making are presented.
High-performance medicine: the convergence of human and artificial intelligence
  • E. Topol
  • Medicine, Computer Science
    Nature Medicine
  • 2019
TLDR
Over time, marked improvements in accuracy, productivity, and workflow will likely be actualized, but whether that will be used to improve the patient–doctor relationship or facilitate its erosion remains to be seen.
Lapgyn4: a dataset for 4 automatic content analysis problems in the domain of laparoscopic gynecology
TLDR
Four datasets extracted from gynecologic, laparoscopic interventions are published with the intend on encouraging research in the field of post-surgical automatic media analysis, designed with the following use cases in mind: medical image retrieval based on a query image, detection of instrument counts, surgical actions and anatomical structures.
Uncertainty-Aware Organ Classification for Surgical Data Science Applications in Laparoscopy
TLDR
This paper significantly enhances the state of art in automatic labeling of endoscopic videos by introducing the use of the confidence metric, and by being the first study to use MI data for in vivo laparoscopic tissue classification.
Toward a standard ontology of surgical process models
TLDR
The OntoSPM Collaborative Action has been in operation for 24 months and its main result is a modular ontology, undergoing constant updates and extensions, based on the experts’ suggestions, making it ready for wider public use.
...
...