Evaluating global and local sequence alignment methods for comparing patient medical records

Table 3 Similarity scores of pairwise global sequence alignments

ID	Operation	Seed Patient 1 (N = 9)			Seed Patient 2 (N = 84)			Seed Patient 3 (N = 224)			Seed Patient 4 (N = 458)
ID	Operation	DTW	NWA	REF	DTW	NWA	REF	DTW	NWA	REF	DTW	NWA	REF
1	x	0.819	0.778	0.778	0.980	0.976	0.976	0.991	0.991	0.991	0.998	0.996	0.996
2	x x	0.597	0.556	0.556	0.976	0.952	0.952	0.987	0.982	0.982	0.994	0.991	0.991
3	u	0.852	0.852	0.852	0.988	0.988	0.988	0.996	0.996	0.996	0.996	0.996	0.996
4	u u	0.714	0.714	0.714	0.952	0.952	0.952	0.984	0.984	0.984	0.993	0.993	0.993
5	s	0.556	0.556	0.556	0.952	0.952	0.952	0.991	0.991	0.991	0.991	0.991	0.991
6	s s	0.286	0.286	0.175	0.905	0.905	0.905	0.988	0.988	0.988	0.983	0.983	0.983
7	x u	0.556	0.556	0.556	0.964	0.952	0.952	0.988	0.987	0.987	0.993	0.991	0.991
8	x s	0.611	0.556	0.556	0.929	0.929	0.929	0.978	0.973	0.973	0.989	0.987	0.987
9	u s	0.457	0.457	0.457	0.929	0.929	0.929	0.981	0.981	0.981	0.987	0.987	0.987
10	x u s	0.363	0.289	0.289	0.905	0.905	0.905	0.969	0.964	0.964	0.984	0.983	0.983
11	X	0.590	0.556	0.556	0.869	0.810	0.810	0.877	0.804	0.804	0.821	0.808	0.808
12	X X	0.179	0.111	0.111	0.702	0.667	0.667	0.708	0.625	0.625	0.657	0.633	0.633
13	U	0.667	0.667	0.667	0.821	0.821	0.821	0.832	0.832	0.832	0.831	0.831	0.831
14	U U	0.551	0.551	0.551	0.786	0.786	0.786	0.711	0.709	0.709	0.695	0.695	0.695
15	S	0.401	0.333	0.160	0.637	0.631	0.631	0.729	0.700	0.679	0.623	0.622	0.622
16	S S	0.078	0.000	−0.269	0.319	0.310	0.310	0.405	0.393	0.351	0.278	0.266	0.262
17	X U	0.185	0.185	0.185	0.702	0.676	0.676	0.679	0.668	0.668	0.716	0.704	0.704
18	X S	−0.204	−0.289	−0.333	0.539	0.530	0.530	0.577	0.552	0.495	0.509	0.501	0.474
19	U S	−0.204	−0.204	−0.204	0.526	0.518	0.518	0.689	0.685	0.664	0.646	0.640	0.636
20	X U S	−0.530	− 0.530	− 0.530	0.611	0.592	0.592	0.571	0.536	0.528	0.627	0.624	0.624

a. ID is the synthetic patient index. N is the number of daily events in a seed patient sequence
b. DTW, NWA and REF refer to as Dynamic Time Warping, Needleman-Wunsch Algorithm, and baseline reference, respectively
c. The lower case letters “x”, “u”, and “s” denote deleting, updating and switching a daily event, respectively. The upper case letters “X”, “U”, and “S” stand for deleting, updating and switching multi-day events (event block)

ISSN: 1472-6947