ExaMode at the Artificial Ontellingence in Healthcare Workshop

Svetla Boytcheva (Sirma AI – @Ontotext) participated at the stakeholders’ workshop “Artificial Intelligence in Healthcare: paving the way with standardization (27 October 2020 )  #AIStandards4Health organized by #CEN & #CENELEC – European Standardization.

There, she discussed challenges, opportunities, and standardization AI solutions for the healthcare sector, all central topics of the ExaMode project.

Papers accepted for the SPIE conference on Digital Pathology

Two papers from Radboudumc have been accepted for oral presentation at the SPIE conference on Digital Pathology, which will be held in February 2021 in San Diego. The abstracts are available here.

Neural image compression for non-small cell lung cancer subtype classification in H&E stained whole-slide images
Paper 11603-1
Author(s): Witali Aswolinskiy, David Tellez, Gabriel Raya, Lieke van der Woude, Monika Looijen-Salamon, Jeroen van der Laak, Katrien Grünberg, Francesco Ciompi, Radboud Univ. Medical Ctr. (Netherlands)

This paper addresses the problem of lung cancer classification, which is one of the topics of ExaMode, and uses whole-slide image compression, which is a technology that is being developed in the project.

Few-shot weakly supervised detection and retrieval in histopathology whole-slide images
Paper 11603-20
Author(s): Mart van Rijthoven, Maschenka Balkenhol, Radboud Univ. Medical Ctr. (Netherlands); Manfredo Atzori, Haute Ecole Spécialisée de Suisse Occidentale (Switzerland); Peter Bult, Radboud Univ. Medical Ctr. (Netherlands); Jeroen van der Laak, Radboud Univ. Medical Ctr. (Netherlands), Linköping Univ. (Sweden); Francesco Ciompi, Radboud Univ. Medical Ctr. (Netherlands)

The second paper contains work from Deliverable D4.1, and it is in collaboration with HES-SO.

People’s choice award at the MICCAI 2020 LABELS Workshop

Congratulations to Sebastian Otálora, Niccolò Marini, Henning Müller, and Manfredo Atzori who received the “People’s choice award” at the MICCAI 2020 LABELS Workshop (Large-scale Annotation of Biomedical data and Expert Label Synthesis).
The article entitled “The Semi-weakly Supervised Learning for Prostate Cancer Image Classification with Teacher-Student Deep Convolutional Networks” is available at the following link: https://link.springer.com/chapter/10.1007%2F978-3-030-61166-8_21

Virtum prototipe published by MicroscopeIT

MicroscopeIT has just finished the first prototype of Virtum, a digital histopathology-dedicated application developed as one of the prototypes in the ExaMode project.
This innovative solution is based on MicroscopeIT’s proprietary Virtum Technology which includes, inter alia, pre-developed modules for storage, management, sharing, processing, and annotating image data, such as Whole Slide Images (WSIs) and other types of biomedical data.

Virtum HP is a digital histopathology-dedicated cloud-based application with virtually unlimited storage and computational resources and can be used to build customized image-analysis applications with integrated AI models.

For more information, you can look at the YouTube video or contact directly MicroscopeIT

ExaMode Periodic Report

September 9 2020 the ExaMode team went through the Periodic Report of the project for month 18. The report showed the results of the team and the high-quality results achieved so far. The collaboration among the partners is being fruitful, driving technical partners to solve practical clinical needs with a variety of advanced methods, expertise, and knowledge from different domains. All deliverables were accepted.

Sirma AI (Ontotext) ExaMode Webinar

Sirma AI (Ontotext)  will deliver a webinar for Examode through BDVA.

The webinar, scheduled on the 2nd of July 2020, 2 PM CEST, is free but for the live session is required a registration.

The webinar will be recorded and will be available as a free video for those who are unable to attend the live session.

The agenda of the webinar includes:

  1. Objectives of the ExaMode project
  2. Knowledge management of diagnosis-related medical data using advanced text analytical technologies and knowledge graphs.
  3. Demonstration of services:
    • Advanced text analytics for semantic data normalization of Electronic Health Record (EHR) extracts – clinical synopsis
    • Semantic data fusion of extracted results with a referential knowledge graph built from relevant ontologies and thesauri (Mondo Disease Ontology, Disease Ontology, UMLS, SNOMED-CT, and others)
    • Visual graph analytics and exploration of semantically normalized cases in the context of the referential knowledge graph
    • Graph similarity search for the identification of similar medical cases
    The solution is implemented on top of Ontotext’s GraphDB, a highly scalable RDF triple store.
  4. Discussion