1 Purpose of the PigLame model Herd Pig Info Info To estimate - - PDF document

1

Bayesian network Case examples

Advanced Herd Managem ent October 2nd 2 00 7 Tina Birk Jensen

Outline 1 . PigLam e m odel

Background of the m odel The qualitative part of the m odel The quantitative part of the m odel Elicitation of probabilities Output variables Exam ple of use ( Esthauge)

2 . Mycoplasm a net

Background of the m odel The m odel Consequences of Mycoplasm a

Background

Leg disorders in finishers: An economical problem for farmers due to:

• Increased work load
• Cost of treatments
• Reduced productivity
• Risk of condemnations
• A negative impact on animal welfare

Background

Causes of leg disorders 1 . I nfectious Mycoplasma hyosynoviae, Erysipelothrix rhusiopathiae, Haemophilus parasuis, Streptococcus suis 2 . Environm ental Fracture, lesion to the claw wall, lesion to the claw sole 3 . I nherited Osteochondrosis manifesta, osteochondrosis dissecans

Control strategies against leg disorders will depend on the cause category

Leg disorder Pig Info Leg disorder Herd Info

2

Strategy 2 Pig Info Leg disorder Strategy 1 Herd Info Strategy 3

Purpose of the PigLame model To estimate probability distributions of different manageable causes of leg disorders in finisher herds PigLame model

Bayesian Netw orks

• A set of variables and directed edges betw een

variables

• Each variable has a finite set of m utually exclusive

states

• The variables and edges form a directed acyclic graph
• To each variable A w ith parents B1….Bn, there is

attached the probability table P( A| B1,….,Bn)

Bayesian network

A B ( ) ( ) ( )

) ( ) | ( ... ) ( ) | ( ) ( ) | ( | |

2 2 1 1 n n i i i

A P A B P A P A B P A P A B P A P A B P B A P + + + =

Bayes Theorem :

Bayesian network

W hy use Bayesian Netw ork approach for the PigLam e m odel?

• Static m odel
• Express the biological variation and represent

uncertainty

• I nform ation can flow in the opposite direction of the

causality

• Evidence about som e input variables can tell us

som ething about variables that are not observable ”Hypothesis variables”

Object-oriented Bayesian network The object diagram for the PigLam e m odel

3

Qualitative structure of the PigLame model

• Tw o classes: Herd class and pig class
• Background for the qualitative structure is

based on evidence from published literature

Qualitative structure of the PigLame model

Pen den Floor Straw Environmental Infectious Inherited Herd size Produc- tion Weight Breed Feed strat Purchase

Herd class Pig class

Environmental Infectious Inherited

The quantitative part of the PigLame model The probabilities in the model are based on:

• Literature
• Expert opinions

The quantitative part of the model

Probabilities elicited from the literature How to get the probabilities into the model Example:

OCM OCD Meat Percent

4

The quantitative part of the model

Probabilities elicited from the literature

I ncrease in the LMP of one percentage point OR= 1.03 ( OCM) OR= 1.05 ( OCD) Form ula:

Assum e:

P( OCM) = 0.5 P( OCD) = 0.2

OR Pno Pno OR pno Pno Pyes ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ − + ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ − = 1 1 1 The quantitative part of the model

0.537 OR5 = 1.035 = 1.16 P(OCM| 5 percent) 0.530 OR4 = 1.034 = 1.13 P(OCM| 4 percent) 0.522 OR3= 1.033 = 1.09 P(OCM| 3 percent) 0.515 OR2 = 1.032 = 1.06 P(OCM| 2 percent) 0.507 OR1 = 1.031 = 1.03 P(OCM| 1 percent) 0.5 OR0 = 1.030 = 1 P(OCM| 0 percent) Probability Odds Ratio Level 0.242 OR5 = 1.055 = 1.28 P(OCD| 5 percent) 0.234 OR4 = 1.054 = 1.22 P(OCD| 4 percent) 0.225 OR3= 1.053 = 1.16 P(OCD| 3 percent) 0.216 OR2 = 1.052 = 1.10 P(OCD| 2 percent) 0.208 OR1 = 1.051 = 1.05 P(OCD| 1 percent) 0.2 OR0 = 1.050 = 1 P(OCD| 0 percent) Probability Odds Ratio Level

Exam ple: P( Fracture| fully slatted floors)

Consider 100 pigs examined individually at a herd visit. The herd has fully slatted floors in the pens. How often do you, during the examination expect to find a pig with a fracture?

Exam ple: P( Fracture| fully slatted floors) Always (almost) 15 25 50 75 85 100 Usually Often As often as not Sometimes Once in a while (Almost) never

Consider 100 pigs examined individually at a herd visit. The herd has fully slatted floors in the pens. How often do you, during the examination expect to find a pig with a fracture?

Exam ple: P( Fracture| fully slatted floors)

The quantitative part of the PigLame model From the literature

• 46 conditional probabilities

From experts

• > 150 conditional probabilities
• 6 experts
• Not randomly distributed

5

The quantitative part of the model

The ”output” nodes

• I nfectious, environm ental and inherited express the

m agnitude of the cause-category at herd level

• Considered as a continuous variable
• Need to m ake a discretization of the node
• Logit m ethod
• The nodes have 1 0 states

Pig class

Environmental Infectious Inherited

The quantitative part of the model

Exam ple:

The ”output” nodes Inherited OCD OCM

The quantitative part of the model

The ”output” nodes

( 1 ) I nherited= Logit( P( OCM| I nherited) ) + 0 C ( 2) I nherited= Logit( P( OCD| I nherited) ) + C Then w e get: ( 1’) P( OCM| I nherited) = Logit -1( I nherited – 0C) ( 2’) P( OCD| I nherited) = Logit -1( I nherited – C)

The quantitative part of the model

The ”output” nodes

Assum e: P( OCM| I nherited) = 0 .5 P( OCD| I nherited) = 0.2 Then: ( 1’) P( OCM| I nherited) = Logit -1( I nherited – 0C ) = 0 .5 ( 2’) P( OCD| I nherited) = Logit -1( I nherited – C) = 0.2 W hich gives: C= - 1.39 I nherited ( OCM) = 0 I nherited ( OCD) = 1.39

The quantitative part of the model

The ”output” nodes

1.39 0.85 0.41

• 0.41
• 0.85
• 1.39
• 2.20

I nherited

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

P( OCM| I nherited)

2.77 2.23 1.79 1.39 0.98 0.54 0.81

I nherited

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

P( OCD| I nherited)

6

Perspectives

• Probability distributions for the three cause-

categories of leg disorders can serve as inputs for future econom ic calculations

• First step in developing an econom ic m odel for

leg disorders in finisher herds

Mycoplasma net

Outline Background of the model The model Consequences of Mycoplasma

Literature: Otto, L. and Kristensen, C.S., 2 0 0 4 : A biological netw ork describing infection w ith Mycoplasm a hyopnem oniae in sw ine herds. Preventive Veterinary Medicine, 6 6 , 1 4 1 -1 6 1 .

Mycoplasma hyopneumonia

• Cause enzootic pneumonia in finishers
• Important in the intensive pig production
• Can exist latent in the herd
• Outbreak: 50-70 percent of the finishers have lung

lesions

• A well documented disease

Mycoplasma net

Purpose To develop a tool for evaluating the economic consequences of different control strategies against Mycoplasma The decisions are often based on intuition from farmers, advisors and vets

Mycoplasma net

Building the m odel: Biological m odel:

• Based on biological knowledge

Econom ic m odel:

• Economic risk due to the biological variation

Risk factors Disease level Productivity Economics Constibution margin

7

Risk factors Disease level Productivity Economics Constibution margin Control stragegy Diagnostics Mycoplasma net

I nput Risk factors:

• Herd size
• Production type
• Purchase policy
• Season
• Region

Mycoplasma net

I nput Diagnostics:

• Clinical examination
• Serology examination
• Postmortem examination

Determines the disease level with more precision

Mycoplasma net

I nput Control strategy: (Any action with the aim to change the level of disease) Short term strategy: Medication Vaccination Long term strategy: Change in mangement Change in production system

Mycoplasma net

Output: Probability distribution:

• Disease level
• Production outcome
• Economical outcome

8

The states of the variables

States: 1-1000 / 1001-3000 / 3001-5000 / 5001- States: Yes / no States: Sept. / Dec. / Mar. / Jun. States: Zero / low / middle / high States: Zero / low / middle / high States: 0-1% / 1-10% / 10-20% / 20%- States: 0g / -30g / -60g / -90g States: 0 / 0-0.15 / 0.15-0.30 / 0.30-0.45 States: 0% / ½% / 1% / 1½% Risk factors Herd size All in all out Seson Diagnostics Clinical examination Serologic examination Postmortem examination Productivity outcome Change in daily weight gain Change in feed conversion rate Change in mortality

Text Label

The economic part

Contribution margin = TR – VC TR: The income from the slaughterhouse VC: Weaner price Cost of feed Dead pigs Labor cost Control strategy Diagnosis

Different scenarios High-risk herd:

Large herd size Continuous production Buying piglets from many other herds

Low-risk herd:

Small herd size Managing all in all out production Not buying piglets from other herds Econom ic losses due to Mycoplasm a: High- risk herd Econom ic loss due to Mycoplasm a: Low - risk herd

Effect of medication and vaccination

9

Perspectives: Mycoplasma net

• Not all needed inform ation can be found in the

literature

• Describes the steady state betw een risk factors

and the severity of Mycoplasm a

• Often it does not pays off to control just one

disease

• I t is im portant to consider several diseases and

their interactions