Algorithm 1.

Transmural segmentation based on 2 diffusion analyses (A,B)

$$\begin{array}{l}{\mathrm{\Omega }}_{\mathit{epi}}:={\mathrm{\Omega }}_A^{+}\\ \equiv \{\mathbf{x}:{\varphi }_A(\mathbf{x})>{\varphi }_{+}\}\\ {\mathrm{\Omega }}_M:=({\mathrm{\Omega }}_A^{-}\cap {\mathrm{\Omega }}_B^{\pm })\cup {\mathrm{\Omega }}_A^{\pm }\\ \equiv \{\mathbf{x}:({\varphi }_A(\mathbf{x})<{\varphi }_{-}\wedge {\varphi }_{-}<{\varphi }_B(\mathbf{x})<{\varphi }_{+})\vee {\varphi }_{-}<{\varphi }_A(\mathbf{x})<{\varphi }_{+}\}\\ {\mathrm{\Omega }}_{\mathit{endo}}:={\mathrm{\Omega }}_A^{-}\cap ({\mathrm{\Omega }}_B^{-}\cup {\mathrm{\Omega }}_B^{+})\\ \equiv \{\mathbf{x}:{\varphi }_A(\mathbf{x})<{\varphi }_{-}\wedge ({\varphi }_B(\mathbf{x})<{\varphi }_{-}\vee {\varphi }_B(\mathbf{x})>{\varphi }_{+})\}\end{array}$$