How can I train (manually labeled) x-ray image data using CNN (labeled data in .json format)?

I have manually labeled some x-ray image data using dataturks.com I downloaded the file which contain the labeled information. But, the downloaded file is in .json format. Using these .json files, how can I train data in CNN? Can MATLAB understands .json file like .mat file?

[Kevin Chng]

Do you mind to attach your jason file hereby?

or you may try code below on how to decode jason in MATLAB

fname = 'meta_Electronics.json'; 
fid = fopen(fname); 
raw = fread(fid,inf); 
str = char(raw'); 
fclose(fid); 
val = jsondecode(str);

[Fahad Parvez Mahdi]

Thank you Kevin, but the code is not running. The following error is showing. Error using jsondecode JSON syntax error at line 2, column 1 (character 3689): extra text.

I have attached the json file with my question.

Answer 1

[Kevin Chng]

Hi Fahad Parvez Mahdi,

Sorry for my late reply, It is because there are so many web link in your json file. Therefore, you have to decode them one by one.

Please refer to the code below, it will help you decode your json file. There are 50 web links.

fname = 'Dental X-Ray Image Dataset 1.json'; 
fid = fopen(fname); 
raw = fread(fid,inf); 
str = char(raw'); 
fclose(fid); 
k = strfind(str,'con');
k=k-3
k(1:2:end)=k(1:2:end)+1
w=1;
for i=1:2:length(k)
val(w) = jsondecode(str(k(i):k(i+1)));
w=w+1;
end

if you have any inquiries, you are always welcome.

[Fahad Parvez Mahdi]

Hi Kevin,

Thank you very much for the code. I can now see the data stored inside the file. Still long way to go. I will knock back you if I need any further assistance from you. Take care.

[Kevin Chng]

Sure. No worries. ^^