Which reconstruction technique is known for generating substantial streak and noise, even outside of the patient's body?

Filtered Back Projection (FBP) is a commonly used reconstruction technique in positron emission tomography (PET) and other imaging modalities. This technique is known for its simplicity and speed, making it a popular choice in clinical settings. However, one of the significant limitations of FBP is its tendency to generate substantial streak artifacts and noise, particularly in areas with high attenuation or limited data.

The presence of streaks occurs because FBP essentially relies on the assumption of uniform noise and error distribution across the image. When this assumption doesn't hold—such as with the uneven distribution of radioactive tracers or varying densities in the patient's body—artifacts become prominent. Additionally, FBP does not effectively handle the presence of low-count data, which often results in increased noise that can extend beyond the confines of the imaged object.

These artifacts can lead to misinterpretation of anatomical structures and functional data, affecting diagnostic accuracy. While other reconstruction methods such as Maximum Likelihood Expectation Maximization (MLEM) or Ordered Subset Expectation Maximization (OSEM) are designed to mitigate such issues by utilizing more sophisticated algorithms for dealing with statistical properties of the data, FBP remains less effective in managing streak and noise artifacts.

This characteristic makes Filtered Back Projection less