A Hybrid Compression Method for Medical Images Based on Region of Interest Using Artificial Neural Networks

Table 2

Summary of the performance measures for the proposed compression method at .

Category	This study ()				Ayoobkhan et al. [7]
	ROI	NROI	Entire image
	PSNR (dB)	PSNR (dB)	CR	PSNR (dB)	CR	PSNR (dB)
Neck	50.84	47.41	10.57	48.32	6.28	45.54
Chest	50.81	47.1	8.25	47.91	6.86	46.42
Front skull view	50.24	47.77	9.98	49.45	8.02	41.64
Right skull view	48.44	47.07	8.4	48.16	8.11	40.29
Left skull view	48.46	46.28	9.64	46.64	8.19	41.43
Pelvic girdle	51.2	48.4	9.96	50.65	8.55	42.81
Radius and ulna bones	49.57	46.54	9.59	47.46	8.62	42.04
Pelvic girdle and back bone	47.15	45.52	8.94	46.85	9.07	39.5
Hand x-ray	47.48	45.57	10.31	45.99	9.18	39.25
Leg	48.77	45.81	9.92	47.2	9.35	38.4
Left mammogram	48.19	46.7	9.84	47.36	14	36.16
Right mammogram	47.7	46.73	10.77	47.53	14.1	35.64
Average	49.07	46.74	9.68	47.79	9.19	40.76