Clinical decision system for GIST cancer - Presentation

0 %
100 %
Information about Clinical decision system for GIST cancer - Presentation

Published on September 22, 2015

Author: Cognitum

Source: slideshare.net

1. 1 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Clinical decision support system for GIST Cancer Alessandro Seganti Data Engineer

2. 2 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Who we are : Cognitum  Fluent Editor  Ontorion Server  Ontorion Controlled Natural Language (OCNL)  Ask Your Data Anything (ADA)

3. 3 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Introduction Medical background, approach and challenges

4. 4 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Medical background  GIST cancer  Little known, High error rate  Well specified recommendations  Manage the long clinical history of the patient  Clinical trial

5. 5 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Requirements  Improve efficiency of treatment and quality of the data  No automatic decisions but recommendations  Use standard recommendations depending on the patient history  Keep the patient clinical history so that it can be reused

6. 6 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Next  Application architecture  Knowledge base  Reasoning  Live presentation

7. 7 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Application architecture

8. 8 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Application Architecture

9. 9 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. General idea Domain knowledge UI knowledge UI Generated knowledge

10. 10 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Knowledge base

11. 11 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Knowledge Modeling  How to model the long clinical history of a patient with GIST cancer.  Close collaboration with the domain expert  OCNL to write the ontologies that can be understood by the domain expert  SWRL for Rules.  Integrity Constraint rules for UI constraints.  Annotations for flexibility.

12. 12 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Histopathological report Patient basic information Surgery Follow up Medical Oncologist consultation Systemic therapy start Systemic therapy continue Systemic therapy stop Clinical trial

13. 13 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Automatic UI Generation

14. 14 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. UI generation - Structure Form Section Field

15. 15 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. UI generation - Fields Every patient must have-date (some datetime value).

16. 16 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. UI generation - Fields Every patient must have-age (some integer value). Every patient must have-sex a sex. Something is a sex if-and-only-if-it is either Male or Female.

17. 17 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Reasoning

18. 18 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Challenges  Reason efficiently and in a scalable way (independently on the number of patients).  Use knowledge dependent on time and reason over it.  Show recommendations based on the patient history.

19. 19 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Reasoning  Ontorion Server to store the knowledge base in a scalable way.  Modularization based on unique patient namespace.  Time dependent variables treated by having one instance at each point of time.  Only the latest instance is used for reasoning.  Reasoning using OWL-RL+ profile.

20. 20 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Reasoning: Transitions  If some condition on form-1 then form-1 is followed by another-form.  If some condition on section-1 inside form-1 then form-1 contains section-2.  SWRL statements.  Possible to use SWRL built-ins.

21. 21 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Recommendations  Recommendations coming from the ESMO guidelines.  Recommendations coming from good practices.  Modelled using SWRL rules.  Depending on the patient clinical history  Another rule is used to decide in which form the recommendation should be shown.

22. 22 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Live demonstration

23. 23 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Live Demonstration  Cdss  Other example of applications using Semantic Forms

24. 24 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Conclusions  Modeled medical knowledge base and applied it to build a complex web application supporting physicians in the daily work.  Innovative approach allowing for scalable reasoning independently on the number of patients.  Currently used in the Oncological center in Warsaw experimentally.  Currently working to apply the same approach to other type of cancer.

25. 25 The company, product and service names used in this web site are for identification purposes only. © Cognitum 2014. All trademarks and registered trademarks are the property of their respective owners. Thank you! Alessandro Seganti Data Engineer a.seganti@cognitum.eu Cognitum Sp. z o.o. Wal Miedzeszynski 630, Warsaw, Poland http://www.cognitum.eu/, office@cognitum.eu Windows Azure Circle Partner MEMBER

Add a comment

Related pages

Epidemiology, classification, clinical presentation ...

... clinical presentation, ... Implications for the SDHB Expression Based Classification of Gists. J Cancer 2011; ... Smarter Decisions, Better Care.
Read more

Gastrointestinal Stromal Tumors Clinical Presentation ...

Gastrointestinal Stromal Tumors Clinical Presentation. ... tumors for a proposed clinical staging system. ... (GIST). Br J Cancer ...
Read more

Gastrointestinal Stromal Tumors Treatment (PDQ®)—Health ...

... the treatment of Gastrointestinal Stromal Tumors. ... Cancers by Body Location/System. ... clinical presentation of patients with GIST varies ...
Read more

Gastrointestinal Stromal Tumors: Practice Essentials ...

Gastrointestinal stromal tumors (GISTs) ... on the clinical presentation and ... gastrointestinal stromal tumors. Clin Cancer ...
Read more

Gastrointestinal Stromal Tumours. GIST information | Patient

Gastrointestinal stromal tumours (GISTs) ... cancers. However GISTs are the most common ... staging systems for GISTs is ongoing with the ...
Read more

Adjuvant and neoadjuvant imatinib for gastrointestinal ...

Adjuvant and neoadjuvant imatinib for gastrointestinal stromal tumors. ... clinical presentation, ... staging system for GIST was developed by the ...
Read more

GIST Support International - GIST Support International ...

GIST Support International ... 2011 ASCO GIST Presentations; ... GIST Support International has compiled current clinical trials for GIST cancer, ...
Read more

Gist | LinkedIn

View 19302 Gist posts, presentations, experts, and more. Get the professional knowledge you need on LinkedIn. ... Management Services Manager at Gist ...
Read more