As the author seems to be around: have you considered surface roughness of the sample as problem to transparency?

Add some water between the grid pattern and the chicken and also some on top of it - or even conduct the experiment under water and the results should be better. You might also freeze the meat before cutting to achieve much smoother surfaces in the first place?