hi community,
i am currently using compression algorithmes such as JPEG and JPEG2000 to compress medical images stored in dicom files, but whene i extract the image i found that most of them are stored in a 16bits coding and using only 12bits or 10bits.
my problem is that whene i try to use lossy compression it does not allow me to because it sas that i cant use lossy for more than 8bits,
i did convert the image to 8bit resolution and i did my compression but i am still wondering why they use 12bits resolution in medical images ?
The value of a pixel of an image is stored in a binary format, i.e., 0 or 1. The maximum range of values that can be assigned to a pixel of an image depends on the number of allocated bits.
- 1 bit would allow only two possible values (2^1) per pixel, i.e., 0 or 1
- 8 bits would allow 256 (2^8) possible values per pixel, i.e., between 0 and 255 inclusive. For example, an arbitrary pixel can have value, 10010110
- 12 bits would allow 4096 (2^12) possible values per pixel, i.e., between 0 and 4095 inclusive.
Medical image, i.e., CT scan image has 4096 levels of grey. It means we need 12 bits to store CT without the loss of any information. For example, an arbitrary pixel of a CT image can have a value, 111001011011. Note that we used 12 bits to store a value of a pixel.
From scanner you can get -1000HU up to 4096HU, 8 bits is only 256 different numbers and it is not enouth to store such dynamic range
- Badges
- Science topic
- Similar topics
- Computer Science
- Images