Fig. 1. SBL algorithm combined with iterative-shrinking permissible region strategy
Fig. 2. (a) Model of non-homogeneous cylinder phantom; (b) surface optical information
Fig. 3. Reconstruction results of simulation experiments. (a)-(c) Stereograms of reconstruction results with IVTCG, StOMP, and SBL; (d)-(f) two-dimensional cross-section views with the three algorithms at z=15 mm
Fig. 4. Results of preliminary experiment 1. (a) Pseudocolor images collected by IVIS system (first column represents results of experimental group, while second column represents results of control group); (b) quantification analysis results of Fig. 4(a)
Fig. 5. Results of preliminary experiment 2. (a) Pseudocolor images collected by IVIS system; (b) quantification analysis results of Fig. 5(a)
Fig. 6. Geometric structure diagrams of (a) cubic and (b) cylindrical phantom; single-views of (c) cubic and (d) cylindrical phantoms collected by IVIS system
Fig. 7. Reconstruction results of cubic physical phantom experiment. (a)-(c) Stereograms of reconstruction results with IVTCG, StOMP, and SBL; (d)-(f) two-dimensional cross-section views of three algorithms at z=1 mm
Fig. 8. Reconstruction results of cylindrical physical phantom experiment. (a)-(c) Stereograms of reconstruction results with IVTCG, StOMP, and SBL; (d)-(f) two-dimensional cross-section views of three algorithms at z=1 mm
Organ | /mm-1 | /mm-1 | g |
---|
Muscle | 0.0052 | 10.800 | 0.90 | Heart | 0.0083 | 6.733 | 0.85 | Lung | 0.0133 | 19.700 | 0.90 | Liver | 0.0329 | 7.000 | 0.90 | Bone | 0.0060 | 60.090 | 0.90 |
|
Table 1. Optical parameters of different regions of non-homogeneous cylinder phantom
Algorithms | Actual source central position /mm | Reconstructed source central position /mm | /mm | fDice | Time /s |
---|
IVTCG | (2,5,15) | (1.72,4.11,12.36) | 2.81 | 0 | 1.38 | StOMP | (2,5,15) | (2.66,5.92,15.54) | 1.26 | 0.33 | 0.74 | SBL | (2,5,15) | (2.09,5.35,15.51) | 0.63 | 0.50 | 0.58 |
|
Table 2. Results of three reconstruction algorithms in simulation experiment
Noise level /% | Actual source central position /mm | Reconstructed source central position /mm | /mm | fDice | Time /s |
---|
10 | (2,5,15) | (2.08,5.38,15.51) | 0.64 | 0.50 | 0.71 | 20 | (2,5,15) | (2.07,5.40,15.52) | 0.66 | 0.45 | 0.68 | 30 | (2,5,15) | (2.19,5.57,14.60) | 0.72 | 0.40 | 0.74 | 40 | (2,5,15) | (2.38,5.36,15.51) | 0.73 | 0.40 | 0.70 |
|
Table 3. Reconstruction results of SBL algorithms at different noise levels
Algorithm | Actual source central position /mm | Reconstructed source central position /mm | /mm | fDice | Time /s |
---|
IVTCG | (6.25,0,1.00) | (5.23,0.73,-1.40) | 2.71 | 0 | 4.45 | StOMP | (6.25,0,1.00) | (5.54,1.48,0.71) | 1.66 | 0.13 | 2.82 | SBL | (6.25,0,1.00) | (6.08,-0.26,0.42) | 0.66 | 0.50 | 1.97 |
|
Table 4. Results of three reconstruction algorithms in cubic physical phantom experiment
Algorithm | Actual source central position /mm | Reconstructed source central position /mm | /mm | fDice | Time /s |
---|
IVTCG | (6.25,0,1.00) | (5.62,-0.22,-2.01) | 3.08 | 0 | 4.72 | StOMP | (6.25,0,1.00) | (5.19,-0.54,0.56) | 1.27 | 0.29 | 2.98 | SBL | (6.25,0,1.00) | (6.59,-0.07,0.47) | 0.63 | 0.50 | 1.87 |
|
Table 5. Results of three reconstruction algorithms in cylindrical physical phantom experiment