Clutter suppression in ultrasound: performance evaluation and review of low-rank and sparse matrix decomposition methods

Table 6 The algorithms with good results in RF phantom experiments

Group	Abbreviation	Time	CNR	Group	Abbreviation	Time	CNR
TTD	3WD\(\bullet\)	1.826	2.393	RPCA	Lag-SPCP-SPG	31.118	2.626
TTD	ADMM\(\bullet\)	4.009	2.478	MC	LMaFit	0.260	2.651
RPCA	ALM	5.456	2.672	MC	LRGeomCG	0.817	2.640
RPCA	APG\(\bullet\)	5.585	2.747	RPCA	LSADM\(\bullet\)	1.443	2.747
RPCA	APG-PARTIAL\(\bullet\)	4.690	2.747	TTD	MAMR	2.263	2.681
RPCA	AS-RPCA	2.697	2.732	RPCA	MoG-RPCA	9.156	2.777
RPCA	DECOLOR	10.266	4.895	NMF	nmfLS2	0.219	2.672
NMF	Deep-Semi-NMF	0.150	2.672	RPCA	NSA1\(\bullet\)	1.549	2.746
NMF	DRMF\(\bullet\)	2.723	2.754	RPCA	NSA2\(\bullet\)	1.656	2.746
RPCA	DUAL\(\bullet\)	215	2.746	MC	OptSpace\(\bullet\)	7.020	2.526
LRR	EALM	0.351	2.672	MC	OR1MP\(\bullet\)	0.089	2.627
RPCA	EALM\(\bullet\)	37.360	2.744	TD	OSTD\(\bullet\)	70.747	1.799
RPCA	flip-SPCP-max-QN\(\bullet\)	119	2.768	RPCA	PCP\(\bullet\)	26.791	2.745
RPCA	flip-SPCP-sum-SPG\(\bullet\)	431	2.768	NMF	PNMF	16.963	2.684
RPCA	FPCP	0.108	2.672	RPCA	PRMF	1.573	2.623
RPCA	FW-T\(\circ\)	0.591	2.578	RPCA	R2PCP	2.241	2.703
RPCA	GA	0.031	3.652	RPCA	RegL1-ALM	4.745	2.774
RPCA	GM	0.155	2.775	TTD	RMAMR	9.728	2.545
RPCA	GoDec	0.097	2.674	LRR	ROSL	0.421	2.715
ST	GRASTA\(\circ\)	1.394	1.207	MC	RPCA-GD	6.215	2.622
RPCA	GreGoDec	0.237	2.821	TD	RSTD\(\circ\)	91.200	1.636
TD	HoRPCA-S-NCX	70.134	2.777	MC	ScGrassMC	4.093	2.567
TD	HoSVD	0.497	2.672	NMF	Semi-NMF	1.267	2.295
RPCA	IALM	0.796	2.748	RPCA	SSGoDec	1.496	2.736
LRR	IALM\(\bullet\)	2.003	2.672	MC	SVP\(\bullet\)	3.235	2.471
RPCA	IALM-BLWS\(\bullet\)	2.474	2.748	RPCA	TFOCS-EC	9.815	2.188
MC	IALM-MC	7.764	2.419	RPCA	TFOCS-IC	9.568	2.197
RPCA	L1F	2.680	0.837	TD	Tucker-ADAL	267	2.617
RPCA	Lag-SPCP-QN	15.766	2.684	TD	Tucker-ALS	0.123	2.672

The algorithms with \(\circ\) are the 3 new algorithms work on processed data, which are defective on original data. The algorithms with \(\bullet\) are sensitive to structured peak pixels and work after logarithmic processing

ISSN: 1475-925X