Attention mechanism based multi-sequence MRI fusion improves prediction of response to neoadjuvant chemoradiotherapy in locally advanced rectal cancer

Table 1 Clinical characteristics of patients in two hospitals

Characteristics	SAHSYU (n = 267)			ZCH (n = 155)			P
Characteristics	Poor response (n = 147)	Good response (n = 120)	p	Poor response (n = 112)	Good response (n = 43)	p	P
Age, years	54.8 ± 12.4	53.0 ± 11.4	0.180	56.1 ± 10.9	56.6 ± 8.0	0.803	0.124
Sex			0.131			0.968	0.673
Male	117 (79.6%)	86 (71.7%)		83 (74.1%)	32 (74.4%)
Female	30 (20.4%)	34 (28.3%)		29 (25.9%)	11 (25.6%)
CEA			0.07			0.583	0.01
Positive	65 (44.2%)	40 (33.3%)		57 (50.9%)	24 (55.8%)
Negative	82 (55.8%)	80 (66.7%)		55 (49.1%)	19 (44.2%)
cT			0.113			0.038	< 0.001
T2	7 (4.8%)	14 (11.7%)		1 (1%)	0
T3	112 (76.2%)	84 (70.0%)		64 (57.1%)	34 (79.1%)
T4	28 (19%)	22 (18.3%)		47 (41.9%)	9 (20.9%)
cN			0.977			0.676	0.076
N0	35 (23.8%)	29 (24.2%)		17 (15.2%)	7 (16.3%)
N1	55 (37.4%)	46 (38.3%)		50 (44.6%)	22 (51.2%)
N2	57 (38.8%)	45 (37.5%)		45 (40.2%)	14 (32.5%)

Continuous data are given as mean ± standard deviation
P values for categorical variables were from chi-square test analysis
P values for continuous variables were from Mann-Whitney rank-sum test analysis
SAHSYU, Six Affiliated Hospital of Sun Yat-sen University; ZCH, Zhejiang Cancer Hospital; cT, clinical T stage; cN, clinical N stage

ISSN: 1748-717X