results from subtask 4.2.2

23
Results from Subtask 4.2.2 Martin Høy Oddvin Sørheim Matforsk AS, Norway

Upload: kyle

Post on 16-Mar-2016

45 views

Category:

Documents


1 download

DESCRIPTION

Results from Subtask 4.2.2. Martin Høy Oddvin Sørheim Matforsk AS, Norway. Outline. Latest results Planned activities to finish the subtask Achievement of scientific goals List of publications. Changes at Matforsk. Vegard Segtnan has left Matforsk - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Results from Subtask 4.2.2

Results from Subtask 4.2.2

Martin HøyOddvin Sørheim

Matforsk AS, Norway

Page 2: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 2

Outline

• Latest results• Planned activities to finish the subtask• Achievement of scientific goals• List of publications

Page 3: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 3

Changes at Matforsk

• Vegard Segtnan has left Matforsk• Martin Høy and Oddvin Sørheim have taken over Vegards

responsibilities wrt. TrueFood

• Matforsk has been merged with other Norwegian institutes– Akvaforsk (Ås)– Fiskeriforskning (Tromsø and Bergen)– NorConserv (Stavanger)

• The name of the new corporation is Nofima• The new name of Matforsk will be “Nofima Food”

Page 4: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 4

Previous results

• A model for predicting fat content from NIR imaging has been developed (RMSECV = 1.0 %)

• Have developed models for predicting salt from– Only NIR imaging (RMSECV = 0.56

%)– Only CT imaging (RMSECV = 0.40

%)– Combination of NIR and CT (RMSECV = 0.34 %)

Page 5: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 5

Experimental design

• Total: 90 fillets: 45 salmon from 3 different weight classes (and fat content classes)

• 10 fillets in each design point• Right and left fillets from the same

fish were always subjected to different salting times

• 30 fillets were taken out of the experiment raw (after NIR and CT)

• 30 fillets were taken out of the experiment after salting (after NIR and CT)

• 30 fillets were smoked and analysed with CT

Salting time

Fillet size/Fat content

Salting time

Fillet size/Fat content

Page 6: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 6

Experimental logistics

Page 7: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 7

Sampling strategy

A

DE

B

CT (salt) NMR (fat reference) Salt reference FT-IR NIR (fat)

D

A

E

BC C

A

DE

B

CT (salt) NMR (fat reference) Salt reference FT-IR NIR (fat)

D

A

E

BC C

Page 8: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 8

Fat modelling: NIR interactance imaging

Page 9: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 9

15 NIR wavelengths

Page 10: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 10

NIR fat prediction results

Page 11: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 11

Fat prediction imagesFisk: 16 FettFisk: 17.2034% Share: 20.2078

10 20 30 40 50 60

50

100

150

200

250

0

5

10

15

20

25

30

35

40

45

50Fisk: 20 FettFisk: 19.8969% Share: 23.6285

10 20 30 40 50 60

50

100

150

200

250

3000

5

10

15

20

25

30

35

40

45

50

Page 12: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 12

CT salt prediction results (CT only)

Page 13: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 13

CT salt results with fat included

Page 14: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 14

Saltcontent raw: 17-L, position 1

50 100 150 200 250 300 350

20

40

60

80

1000

5

10

15

Saltcontent after salting: 17-L, position 1

50 100 150 200 250 300

20

40

60

80

1000

5

10

15

Saltcontent after salting and smoking: 17-L, position 1

50 100 150 200 250 300 350

20

40

60

80

0

5

10

15

Salt prediction images

Tail

NaCl (w/w %)

0

2

4

6

8

10

12

14

16

Page 15: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 15

Distribution of fat and salt vs. position

54321

25

20

15

10

5

Position

Mea

n fa

t con

tent

(%

)

A

54321

5.5

5.0

4.5

4.0

3.5

3.0

2.5

2.0

Position

Mea

n N

aCl c

onte

nt (

%)

B

Page 16: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 16

Fat versus salt content, all samples

Predicted fat

Salt

3025201510

7

6

5

4

3

2

1

Plugs from all 5 positions

R=-0.17

Page 17: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 17

Fat versus salt content, back samples

Predicted fat

Salt

20.017.515.012.510.07.55.0

6

5

4

3

2

3 plugs from back

R=-0.78

Page 18: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 18

Fat versus salt content, belly samples

Predicted fat

"Sal

t"

3028262422201816

7

6

5

4

3

2

Plugs from belly

R=-0.06

Page 19: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 19

Results from salting experiment

3210 32104

2

04

2

0

Position

Salting technique

Salt level

12345

Position

0123

techniqueSalting

Interaction Plot for NaClData Means

Salting levels:0: No salt1: Low2: Medium3: High

Salting technique:0: No salt1: Dry salting2: Brine salting3: Brine injection

Position:1-3: Head to tail4-5: Belly

Page 20: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 20

Results from salting (continued)

• Position vs technique: Clear differences between positions for dry salting and brine salting. No differences between positions for brine injection

• Technique vs level: Smaller differences between salt levels for dry salting than compared to brine salting and brine injection. Generally higher levels for brine salting than for brine injection.

Page 21: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 21

Planned activities (4.2.4)

• Two new experiments will be performed in Jan/Feb 2009• Focus will be on reducing overall salt content• We will use two salting techniques, dry salting and brine

injection, with different salt levels• NaCl will be partially replaced by KCl• Analysis to be performed:

– NIR fat distribution measurements– CT salt distribution measurements– NaCl reference, microbial count, texture

• Aim: Protocol for production of fillets with reduced salt content

Page 22: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 22

Planned activities (continued)

• The experiment will be performed in collaboration with WP1• The produced fillets will be tested by a sensory panel and/or by

consumers

• Aim: Identification of sensory properties / consumer preference / acceptance

Page 23: Results from Subtask 4.2.2

20.10.2008 Results from 4.2.2 - Meeting in Paris 23

List of publications

• Fat distributional analysis in salmon fillets using non-contact near infrared interactance imaging: A sampling and calibration strategy, Vegard. H. Segtnan, Martin Høy, Frank Lundby, Bjørg Narum and Jens Petter Wold. Submitted to Journal of Near Infrared Spectroscopy.

• Non-contact salt and fat distributional analysis in salted and smoked salmon fillets using X-ray computed tomography and NIR interactance imaging, Vegard. H. Segtnan, Martin Høy, Oddvin Sørheim, Achim Kohler, Frank Lundby, Jens Petter Wold and Ragni Ofstad. Submitted to Journal of Agricultural and Food Chemistry.