Design of Metaheuristic Optimization-Based Vascular Segmentation Techniques for Photoacoustic Images

Table 3

Comparative results analysis of proposed versus existing techniques.

Methods	Dice coefficient	IoU	Sensitivity	Accuracy
Threshold segmentation	71.18	56.49	61.74	97.80
Region growing	64.80	49.57	52.36	97.46
Maximum entropy	50.87	35.53	95.85	91.22
K-means clustering	75.51	61.13	71.22	97.99
DCNN-FCN	68.67	52.40	70.62	96.83
DCNN-UNet	82.38	70.37	83.89	98.09
DCNN-Hy-net	83.91	72.65	89.14	98.28
SEF-CS	84.92	73.82	96.66	98.46
SEF-EO	85.04	73.39	96.76	98.55
SEF-HS	85.27	73.69	97.23	98.62
MLOT-CS	86.55	73.83	97.44	98.67
MLOT-EO	87.07	73.97	97.63	98.64
MLOT-HS	87.09	74.30	98.22	98.71