Table S1. Ability of traditional PET: probability-based, and textural features in differentiating tumors with tumor regression score 4 from those with other scores.

Classification	Index	AUC	P value	95% CI
				LB	UB
Classical PET feature	SUVmax	0.55	0.446	0.436	0.665
	Mean	0.579	0.234	0.465	0.693
	Median	0.578	0.239	0.461	0.695
	Variance	0.585	0.197	0.471	0.7
	Std. Dev.	0.585	0.197	0.471	0.7
	Skewness	0.4	0.13	0.287	0.512
	Kurtosis	0.353	0.027	0.238	0.469
	25th percentile	0.582	0.217	0.466	0.697
	75th percentile	0.579	0.23	0.465	0.694
	Peak	0.553	0.422	0.441	MTV 0.665
	MTV	0.321	0.007	0.202	0.439
	TLGmax	0.374	0.057	0.259	0.489
	TLGmean	0.388	0.09	0.273	0.503
	TLGpeak	0.383	0.077	0.269	0.497
	Total	0.388	0.09	0.273	0.503
Probability based feature	Entropy	0.32	0.007	0.202	0.438
	Energy	0.679	0.007	0.562	0.797
	DiversityD2	0.321	0.007	0.203	0.438
	DiversityD3	0.321	0.007	0.204	0.439
	DiversityD4	0.324	0.008	0.206	0.442
Gray Level Co-occurrence Matrix (GLCM)	Autocorrelation	0.676	0.008	0.555	0.798
	ContrastG	0.647	0.026	0.51	0.785
	Correlation	0.375	0.059	0.236	0.514
	Cluster prominence	0.544	0.51	0.406	0.681
	Cluster shade	0.445	0.406	0.328	0.561
	Dissimilarity	0.651	0.022	0.518	0.785
	Energy	0.603	0.12	0.473	0.733
	Entropy	0.38	0.071	0.256	0.505
	Homogeneitym	0.314	0.005	0.209	0.419
	Homogeneityp	0.322	0.007	0.215	0.429
	Maximum probability	0.565	0.329	0.436	0.693
	Sum of squares variance	0.678	0.007	0.555	0.802
	Sum average	0.682	0.006	0.562	0.802
	Sum variance	0.682	0.006	0.561	0.803
	Sum entropy	0.415	0.2	0.292	0.538
	Difference variance	0.647	0.026	0.51	0.785
	Difference entropy	0.505	0.944	0.373	0.636
	Information measure of correlation 1	0.299	0.002	0.178	0.42
	Information measure of correlation 2	0.683	0.006	0.563	0.803
	Inverse difference normalized	0.316	0.005	0.209	0.422
	Inverse difference moment normalized	0.331	0.011	0.218	0.445
Gray-Level Run Length Matrix (GLRLM)	SRE	0.686	0.005	0.569	0.804
	LRE	0.32	0.006	0.216	0.423
	GLNUr	0.292	0.002	0.176	0.407
	RP	0.357	0.03	0.243	0.47
	RLNU	0.324	0.008	0.204	0.443
	LGRE	0.314	0.005	0.188	0.441
	HGRE	0.676	0.008	0.558	0.795
	SRLGE	0.316	0.005	0.188	0.443
	SRHGE	0.683	0.006	0.564	0.803
	LRLGE	0.285	0.001	0.159	0.411
	LRHGE	0.645	0.028	0.523	0.767
Neighborhood Gray-Level Different Matrix (NGLDM)	Coarseness	0.708	0.002	0.598	0.819
	ContrastN	0.687	0.005	0.566	0.808
	Busyness	0.285	0.001	0.171	0.399
	Complexity	0.683	0.006	0.559	0.807
	Strength	0.658	0.017	0.539	0.777
Gray-Level Zone Length Matrix (GLSZM)	SZE	0.638	0.038	0.509	0.766
	LZE	0.305	0.003	0.191	0.419
	GLNUz	0.319	0.006	0.199	0.439
	ZP	0.466	0.611	0.354	0.579
	ZLNU	0.356	0.03	0.233	0.48
	LGZE	0.389	0.093	0.26	0.518
	HGZE	0.652	0.022	0.534	0.769
	SZLGE	0.496	0.948	0.362	0.63
	SZHGE	0.613	0.087	0.492	0.734
	LZLGE	0.257	<0.001	0.138	0.376
	LZHGE	0.49	0.877	0.376	0.604

MTV, metabolic tumor volume; TLG, total lesion glycolysis; SRE, short-run emphasis; LRE, long-run emphasis; LGRE, low gray-level run emphasis; HGRE, high gray-level run emphasis; SRLGE, short-run low gray-level emphasis; SRHGE, short-run high gray-level emphasis; LRLGE = long-run low gray-level emphasis; LRHGE, long-run high gray-level emphasis; GLNUr, gray-level nonuniformity for run; RLNU, run-length nonuniformity; RP, run percentage; SZE, short-zone emphasis; LZE, long-zone emphasis; LGZE, low gray-level zone emphasis; HGZE, high gray-level zone emphasis; SZLGE, short-zone low gray-level emphasis; SZHGE, short-zone high gray-level emphasis; LZLGE, long-zone low gray-level emphasis; LZHGE, long-zone high gray-level emphasis; GLNUz, gray-level nonuniformity for zone; ZLNU, zone length nonuniformity; ZP, zone percentage. Definition of 25^th percentile: The 25^th percentile is a measurement of relative standing within SUVs of an MTV, indicating that 25% of all SUVs are below the MTV. The same model is applied for the 75^th percentile. Definition of peak: The average of SUV_max and SUVs of 26 adjacent voxels. Total: Sum of all SUVs within the MTV. $\mathrm{Diversit}{y}_{\mathrm{Dn}}={({\displaystyle {\sum }_{i=1}})}^{\frac{1}{(1-n)}}$ where P_i indicating the occurrence probability of discretized SUVs within MTV assign to bin.