Table 2.

Clustering details.

Clustering order	Silhouette score	Mean dot product	Single-patient clusters
2	0.50	0.74 ± 0.086	No
3	0.44	0.74 ± 0.081	No
4	0.49	0.73 ± 0.080	No
5	0.54	0.68 ± 0.067	1
6	0.52	0.68 ± 0.065	2
7	0.55	0.67 ± 0.062	3
8	0.61	0.68 ± 0.063	3
9	0.63	0.67 ± 0.065	4
10	0.67	0.68 ± 0.059	4
11	0.71	0.69 ± 0.063	5

For each clustering order, the optimal solution found with Matlab k-means function is reported, with the Silhouette score, the mean dot product with other clusters, and the number of single-patient clusters.