Due　to　the　high　scattering　and　low　absorption　characteristics　of　near-infrared　light　in　the　biological　medium,　the　image　quality　and　spatial　resolution　obtained　by　the　Diffuse　Optical　Tomographic　(DOT)　cannot　meet　the　medical　diagnosis　request.　In　this　paper,　we　introduce　a　reconstruction　approach　that　the　gray　value　of　Magnetic　Resonance　Image　(MRI)　can　be　encoded　into　the　image　reconstruction　process　as　prior　information　to　avoid　the　user　intervention　of　image　segmentation　and　to　improve　the　quality　of　the　reconstructed　images.　Consider　to　the　sparsity　of　the　interest　region　in　background,　L1　norm　are　adopted　as　the　regularization　term.　Relative　to　the　no　prior　and　the　L2-norm　regularization,　the　results　show　improvements　in　reconstructed　accuracy　and　contrast.