1. 15th International Conference on Bioinformatics & Bioengineering (BIBE)
Nov-02-04, 2015, Belgrade, Serbia
A Two-Level Competitive Fuzzy Cognitive Map for Modelling Soft Tissue Knee Injuries
PhD Candidate Antigoni P. Anninou
Laboratory for Automation and Robotics
Department of Electrical and Computer Engineering
University of Patras

2. Authors
Anninou P. Antigoni, Ph.D Candidate, Laboratory of Automation and Robotics, Department of Electrical and Computer Engineering, University of Patras
Groumpos P. Peter, Professor, Laboratory of Automation and Robotics, Department of Electrical and Computer Engineering, University of Patras
Gkliatis Ioannis, Assistant Professor, Orthopaedic Clinic, Department of Medicine, University of Patras, Greece
Poulios Panagiotis, Orthopaedic Clinic, Department of Medicine, University of Patras,Greece

3. Layout Aim Soft Tissue Knee Injuries Fuzzy Cognitive Maps
Competitive Fuzzy Cognitive Maps
Modelling Knee Injuries
Case Studies
Conclusions
Future Research

4. Aim
Propose a novel approach of modelling soft tissue knee injuries using a two-level Competitive Fuzzy Cognitive Map

5. Soft Tissue Knee Injury
Most common and clinically challenging musculoskeletal disorder met in the emergency department
Problem:
Many and complex parameters
Impossible an accurate diagnosis

6. Solution
A modeling tool that can handle all these challenges and at the same time be able to infer a decision
A special type of Fuzzy Cognitive Maps has been introduced for Medical Diagnosis systems, with advanced capabilities, the Competitive Fuzzy Cognitive Map (CFCM)

7. Fuzzy Cognitive Maps (FCM)
Modeling method for describing particular domains
Fyzzy-graph structures for representing causal reasoning

8. Fuzzy Cognitive Maps
Nodes: Represent the system’s concepts or variables
Arrows: Interconnection between nodes. Show the cause-effect relationship between them.
W: Weight between two nodes:
W>0 positive causality
W<0 negative causality
W=0 no relationship

9. Fuzzy Cognitive Maps
The value of each concept at every simulation step is calculated, computing the influence of the interconnected concepts to the specific concept, by applying the following calculation rule:

10. Fuzzy Cognitive Maps
Ai(k+1) : the value of the concept Ci at the iteration step k+1
Ai(k): the value of the concept Cj at the iteration step k
Wij : the weight of interconnection from concept Ci to concept Cj
k1: the influence of the interconnected concepts in the configuration of the new value of the concept Ai
k2: the proportion of the contribution of the previous value of the concept in the computation of the new value
f : the sigmoid function

11. Competitive Fuzzy Cognitive Maps (CFCM)
The output nodes of a FCM used in decision- making, in many cases, must "compete" against each other in order for only one of them to dominate and be considered the correct decision
In order to achieve this "competition", the interaction of each of these nodes with the others should have a very high negative weight. This implies that the higher the value of a given node, the lower the value of competing nodes

12. Modelling Knee Injuries Using CFCM
Mechanisms of injury: The manner in which a physical injury occurred
Clinical Tests: Clinical examination results – Confirm a diagnosis
Two-Level CFCM

13. Possible Mechanisms of Injury

14. Possible Symptoms

15. All Possible Diagnoses

16. CFCM Model of Mechanisms of Injury and Diagnoses

17. CFCM Model of Symptoms and Diagnoses

18. 1st Case
A 22 year old male, victim of traffic road accident with motorcycle, recalls landing on his right leg with the knee flexed, in valgus moment, unable to ambulate after injury with sense of instability
Symptoms: C1 (acute effusion), C4 (effusion generalized), C5 (extension lag), C6 (flexion contracture), C7 (knee joint unstable in extension), C8 (knee joint unstable in flexion), C12 (tenderness), C13 (unable to bear weight joint), C14 (sense of instability).
Mechanism of injury: M7 (Flexion, valgus and external rotation)

19. 1st Case-1st Level D1 (Meniscus Injury): 0.38 D2 (ACL Injury): 0.44
D3 (PCL Injury): 0.15
D4 (MCL/PMCC Injury) : 0.44
D5 (LCL): 0.15
D6 (PLCC Injury): 0.21
D7 (Patella Dislocation/Extensor Apparatus Injury): 0.15

20. 1st Case-2nd Level Diagnoses: ACL injury (D2)  0.9
Medial Collateral Ligament(D4)  0.95
Lateral Collateral Ligament (D5)  0.95
Posterior Lateral Complex (D6)  0.83
Patella Dislocation (D7)  0.95

21. Clinical Tests - Diagnoses

22. 1st Case – Final Diagnoses
Positive Clinical Tests: T5, T13, T14, T15, T17, T18
Possible Diagnoses based on tests: D2, D3, D5, D6
Diagnoses from CFCM Model: D2, D4, D5, D6, D7
Final Diagnoses:
ACL injury D2,
Lateral Collateral Ligament D5,
Posterior Lateral Complex D6

23. 2nd Case
A 34 year old man, during a martial art session, heard a ”pop” and he had a pivoting moment with foot planted on mat, he was unable to ambulate after injury and had a sense of instability
Symptoms: C1 (acute effusion), C4 (effusion generalized), C5 (extension lag), C6 (flexion contracture), C7 (knee joint unstable in extension), C13 (unable to bear weight joint), C14 (sense of instability), C15 (”Poping” Sound during Injury)
Mechanism of injury: M7 (Flexion, valgus and external rotation)

24. 2nd Case-2nd Level Diagnoses: Meniscus Injury (D1)  0.85
ACL injury (D2)  0.97
Patella Dislocation (D7)  0.96

25. 2nd Case – Final Diagnoses
Positive Clinical Tests: T5, T7, T11
Possible Diagnoses based on tests: D2, D3, D4, D6
Diagnoses from CFCM Model: D1, D2, D7
Final Diagnoses:
ACL injury D2

26. 3rd Case
A 20 year old woman suffered forced valgus injury of knee during a tackling maneuver in Mixed Martial Arts (MMA) match. She couldn’t bear weight, heard a ”pop”, with acute effusion of the knee joint, and had a sense of instability
Symptoms: C1 (acute effusion), C4 (effusion generalized), C5 (extension lag), C13 (unable to bear weight joint), C14 (sense of instability), C15 (”Poping” Sound during Injury)
Mechanism of injury: M5 (Flexion, valgus and external rotation)

27. 3rd Case – 1st Level D1 (Meniscus Injury): 0.39 D2 (ACL Injury): 0.34
D3 (PCL Injury): 0.29
D4 (MCL/PMCC Injury) : 0.41
D5 (LCL): 0.16
D6 (PLCC Injury): 0.18
D7 (Patella Dislocation/Extensor Apparatus Injury): 0.16

28. 3rd Case – 2nd Level Diagnoses: Meniscus Injury (D1)  0.86
ACL injury (D2)  0.81
Medial Collateral Ligament(D4)  0.84
Patella Dislocation (D7)  0.91

29. 3st Case – Final Diagnoses
Positive Clinical Tests: T2, T5, T6, T7, T11
Possible Diagnoses based on tests: D1, D2, D3, D4, D6
Diagnoses from CFCM Model: D1, D2, D4, D7
Final Diagnoses:
Meniscus Injury D1
ACL injury D2,
MCL/PMCC Injury D4

30. Conclusions A new CFCM model for knee injuries
Simple
Real-time
Fast
Reliable
Satisfactory results for three real cases
Helps physicians for a first and accurate diagnosis without having to use modern imaging techniques (MRI) in every single patient

31. Future Research A more comprehensive CFCM model for knee injuries
Clinical use in real-time medical diagnosis
New simulations with more clinical data
Use the CFCM model in other medical problems

32. Thank you for your attention
Antigoni P. Anninou
Peter P. Groumpos
Ioannis Gkliatis
Panagiotis Poulios