Visualization of fMRI data

This is the supplemental website in support of the final project
Reading the Noisy Mind for CSE 546 (Autumn 2016),
by Avi Levy and Peter Lin.

Go to: word embedding | training data visualization | report | poster | source code.


We have used the data from the paper by Mitchell et. al. (Science 2008):
Predicting Human Brain Activity Associated with the Meanings of Nouns.

There are 60 words in total. We learn words from brain (fMRI) images.

Table of pairwise cosine distances for the learned feature embedding (click to enlarge):

We visualize an fMRI scan for each word (2D cross section + 3D view).


Go to top

refrigerator

   

1



barn

   

2



bell

   

3



key

   

4



window

   

5



screwdriver

   

6



church

   

7



desk

   

8



cow

   

9



knife

   

10



train

   

11



beetle

   

12



tomato

   

13



foot

   

14



eye

   

15



dresser

   

16



spoon

   

17



truck

   

18



apartment

   

19



pants

   

20



ant

   

21



saw

   

22



coat

   

23



closet

   

24



dress

   

25



igloo

   

26



skirt

   

27



corn

   

28



airplane

   

29



car

   

30



bear

   

31



celery

   

32



butterfly

   

33



shirt

   

34



arm

   

35



arch

   

36



pliers

   

37



glass

   

38



house

   

39



cup

   

40



horse

   

41



bottle

   

42



chisel

   

43



cat

   

44



bee

   

45



lettuce

   

46



bicycle

   

47



hand

   

48



chair

   

49



hammer

   

50



door

   

51



leg

   

52



carrot

   

53



watch

   

54



table

   

55



chimney

   

56



bed

   

57



fly

   

58



dog

   

59



telephone

   

60



server: mathweb01 © 2016