PKIN features a comprehensive set of compartment models as listed below. The number of tissue compartments ranges from 1 to 3. There are different variants of the same model structure, so that prior information can be entered easily, and that coupling of physiologic parameters across region is possible.
For instance, the 2tissue compartment model has the standard parameters K_{1}, k_{2}, k_{3}, k_{4} . An equivalent description is by the parameters K_{1}, K_{1}/k_{2}, k_{3}, k_{4} . The advantage is, that the K_{1}/k_{2}, represents the distribution volume of the nondisplacement compartment in tissue (free and nonspecifically bound tracer), which can often be assumed to be the same across different tissues. Therefore, K_{1}/k_{2} can be included as a common parameter in a coupled fit, hereby reducing the number of fitted parameters and thus potentially improving the identifiability of all.
Model Name 
Description 
Most basic compartment model with the plasma compartment and one tissue compartment. 

Compartment model with the plasma compartment and two sequential tissue compartments. Often used for receptor studies. 

The 2Tissue compartment model including the plasma glucose and the lumped constant to calculate the metabolic rate of glucose. 

The same model as the 2Tissue compartment model, except that K_{1}/k_{2 }is used as a model parameter instead of k_{2}. This facilitates coupled fitting. 

The same models as the 2Tissue compartment model, except that K_{1}/k_{2} and the specific or total distribution volumes are used as model parameters instead of k_{2} and k_{4}. This facilitates coupled fitting and the easy generation of synthetic model curves. 

2Tissue Compartment model solved by the Linear Least Squares method. 

2Tissue compartment model with an efflux from the last compartment. 

Nonlinear 2Tissue compartment model for receptor tracer studies accounting for the saturation of receptor sites. 

Same model as above, but using K_{1}/k_{2} as fitting parameter instead of k_{2}. 

2Tissue compartment model with an additional endothelial trapping compartment. 

Compartment model which separates free tracer in tissue from nonspecific binding. 

As above, but using but using K_{1}/k_{2} and the specific distribution volume as fitting parameters. 

Model with three sequential compartments originally developed for FDG uptake in skeletal muscle. 

Specific model for dynamic H_{2}^{15}O PET Data with implicit deconvolution of the input curve dispersion 

Extends the 2 tissue compartment model by a metabolite compartment with a second input curve of labeled metabolites from the plasma. 

3Tissue Compartments, Metabolites, K1/k2 3Tissue Compartments, Metabolites, K1/k2, Vs

As above, but using but using K_{1}/k_{2} and the specific distribution volume as fitting parameters. 
Two parallel 2compartment models for authentic ligand and metabolites, linked by a transfer constant between the nonspecific compartments. 

During a single imaging study three injections are applied: hot ligand first, then cold ligand for displacement, then a mixture of cold & hot ligand. The individual receptor parameters can be estimated. 