Fakultät Informatik und Mathematik
Regensburg Center for Artificial Intelligence
Regensburg Center of Biomedical Engineering
Regensburg Center of Health Sciences and Technology

Prof. Dr. rer nat. Christoph Palm

Deep Learning for computer-assisted early detection of esophageal cancer from endoscopic images

Cooperation partners

  • Regensburg Medical Image Computing, Ostbayerische Technische Hochschule Regensburg
    • Prof. Dr. Christoph Palm
    • Robert Mendel
    • David Rauber
  • III. Medical Clinic, Klinikum Augsburg
    • Prof. Dr. Helmut Messmann
    • Dr. Alanna Ebigbo
    • Dr. Andreas Probst
  • Department of Computing, São Paulo State UniversitySão Paulo, Brazil
    • Prof. João Papa, PhD
    • Luis Antonio de Souza Jr.

Publications

Projekt Barrett (Prof. Palm)

2020 | 2019 | 2018 | 2017

2020

Alanna Ebigbo, Robert Mendel, Andreas Probst, Johannes Manzeneder, Friederike Prinz, Luis Antonio de Souza, João P. Papa, Christoph Palm, Helmut Messmann Real-time use of artificial intelligence in the evaluation of cancer in Barrett’s oesophagus

2019

Alanna Ebigbo, Christoph Palm, Andreas Probst, Robert Mendel, Johannes Manzeneder, Friederike Prinz, Luis Antonio de Souza, João P. Papa, Peter Siersema, Helmut Messmann A technical review of artificial intelligence as applied to gastrointestinal endoscopy: clarifying the terminology
Alanna Ebigbo, Robert Mendel, Andreas Probst, Johannes Manzeneder, Luis Antonio de Souza, João P. Papa, Christoph Palm, Helmut Messmann Artificial Intelligence in Early Barrett's Cancer: The Segmentation Task
Leandro A. Passos, Luis Antonio de Souza, Robert Mendel, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Christoph Palm, João P. Papa Barrett’s esophagus analysis using infinity Restricted Boltzmann Machines
Alanna Ebigbo, Robert Mendel, Andreas Probst, Johannes Manzeneder, Luis Antonio de Souza, João P. Papa, Christoph Palm, Helmut Messmann Computer-aided diagnosis using deep learning in the evaluation of early oesophageal adenocarcinoma
Luis Antonio de Souza, Luis Claudio Sugi Afonso, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Robert Mendel, Christian Hook, Christoph Palm, João P. Papa Learning visual representations with optimum-path forest and its applications to Barrett's esophagus and adenocarcinoma diagnosis

2018

Luis Antonio de Souza, Christoph Palm, Robert Mendel, Christian Hook, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Silke Weber, João P. Papa A survey on Barrett's esophagus analysis using machine learning
Luis Antonio de Souza, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Joao P. Papa, Robert Mendel, Christoph Palm Barrett's Esophagus Identification Using Color Co-occurrence Matrices

2017

Luis Antonio de Souza, Christian Hook, João P. Papa, Christoph Palm Barrett's Esophagus Analysis Using SURF Features
Luis Antonio de Souza, Luis Claudio Sugi Afonso, Christoph Palm, João P. Papa Barrett's Esophagus Identification Using Optimum-Path Forest
Robert Mendel, Alanna Ebigbo, Andreas Probst, Helmut Messmann, Christoph Palm Barrett’s Esophagus Analysis Using Convolutional Neural Networks

 

Project description

Motivation

Reflux is an inflammation of the esophagus, which is caused by an increased reflux of acid stomach contents into the esophagus. Chronic reflux is the main cause of the Barrett's esophagus, a lesion of the mucous membrane with an increased risk of developing esophageal cancer. The survival chances of affected patients are considered poor, as the disease is usually diagnosed at a late stage. If a standard drug treatment of reflux is not successful, an endoscopic examination may be indicated to detect treatable symptoms as early as possible. However, this is not unproblematic, because many reflux patients are endoscopically negative, i.e. mucous membrane lesions are not visible despite the presence of disease (low sensitivity of the examination). The significance in case of a pathological finding is relatively high (high specificity of the examination).

Goals and procedure

Machine learning methods are increasingly being used in diagnostic imaging procedures. With the help of deep learning approaches, the physician should be supported in reliably detecting reflux-related mucous membrane damage, in particular (pre-)carcinogenic lesions, when evaluating endoscopic images. Based on the machine evaluation of endoscopic images, conclusions on the severity of a possible disease should be drawn. Through the use of Deep Learning, a quality in the diagnostic evaluation of medical images has been achieved several times in recent years that not only reaches the medical "gold standard", but even exceeds it. This means that physician and computer meet at eye level, so that in the future, for example, the computer could at least be established as a second assessor.