Fourier　ptychographic　microscopy　is　a　new　computational　imaging　method　developed　in　recent　years,　which　can　obtain　the　complex　amplitude　information　of　the　sample　simultaneously　maintaining　the　large　field-of-view　and　improving　the　imaging　resolution.　Meanwhile,　the　diffraction　tomography　technique　can　realize　the　three-dimensional　(3D)　refractive　index　imaging　of　the　object.　So,　the　combination　of　Fourier　ptychographic　microscopy　imaging　and　the　diffraction　tomography　algorithm　can　obtain　high-resolution　3D　complex　refractive　index　distribution　of　thick　samples.　In　visible　light,　recent　studies　have　shown　that　diffraction　tomography　reconstruction　algorithms　based　on　the　multi-scattering　forward　propagation　model　can　overcome　the　limitations　of　the　Born/Rytov　approximation　and　reduce　the　impact　of　multi-scattering　effects　on　the　reconstructed　refractive　indices　of　3D　objects.　In　this　paper,　we　propose　a　method　to　accurately　describe　the　image　acquisition　process　of　the　FPM　system　using　a　multiple　scattering　forward　propagation　model,　and　then　combine　iterative　optimization　methods　to　obtain　high-quality　diffraction　tomography　reconstruction　results.　The　multiple-sphere　sample　is　employed　to　do　the　simulation,　and　the　results　show　that　the　method　can　effectively　solve　the　multiple　scattering　problem　inside　complex　structure　samples　and　obtain　better　reconstructed　results　than　traditional　diffraction　tomography　reconstruction　algorithms　for　samples　that　do　not　meet　the　Born/Rytov　approximation　requirements.　The　method　provides　a　new　perspective　on　image　recovery　methods　for　Fourier　ptychographic　microscopy　imaging　systems.　©　2023　SPIE.