Additional Table 1.

Summary of 12 DTI and structural MRI studies in amblyopia, some of which combined with task-based or resting-state functional MRI

Studies	Modality	Analysis method	Amblyopiceye and type	Patients number and age*	Ways and proportion of treatment received	Controls number and age*	Main findings
Liang et al., 2019	3DT1	SBM(ROI)	AA: 12OS, 8OD	9M/11F, 25.2±9.9 y	All had undergone treatment while the details were NA	8M/12F, 26.7±7.4 y	Thinner cortical thickness in the bilateral V1, left V2, ventral V3, V4, and V5/MT+ were found in AA than HC. The cortical thicknesses in bilateral V1 negatively correlated with the difference value of the binocular refractive power in patients withAA.
Mendola et al.,2005	3DT1	VBM (whole brain, ROI)	PediatricAA: 6OS, 10OD; Adult AA: 5OS, 4OD; PediatricSA: 5OS, 5OD; 8ES,2EX; Adult SA: 9OS, 4OD; 9ES, 4EX	Pediatric AA: 16, 9.8±1.2y; Adult AA: 9, 28.9±6.5y; Pediatric SA: 10, 9.7±1.5y; Adult SA: 13, 24.8±5.5y	Surgery: pediatric SA 5/10, adult SA 9/13. Patch: pediatric SA 9/10, adult of SA 10/13, pediatric AA 14/15, adult AA 8/9	17children, 10.2±1.5y; 10 adults, 25.2±3.8y	Pediatric AA and SA showed lower GMV in bilateral calcarine and paracalcarine cortex, medial parieto-occipital junction (but only in the right hemisphere for SA), lateral parieto-occipital junction, and ventral temporal cortex (wider reduction in AA than SA) than their respective pediatric HC. Adult AA and SA both showed lower GMV in calcarine cortex than their respective adult HC, the areas of which is larger inAA than SA.
Su et al., 2022	3DT1	VBM (whole brain)	SA: 5ES, 11EX	SA: 5M/11F; 24.4±6.0 y	Noophthalmic surgery	5M/11F; 25.7±5.9 y	Smaller GMV were found in the right superior temporal gyrus, and the posterior and anterior areas of cerebellum in patients with SA, and the mean GMV values in these areas were negatively correlated with the angle of strabismus in SA.
Barnes et al., 2010	3DT1, tfMRI	VBM and brain activation (ROI)	Adult SA: 7OS, 9OD; 14ES, 2EX	10M/6F, 37.9±13.6 (20-69) y	Patch11/16 Surgery 3/16 Glasses 5/16	11, 34±5 y	SA had less gray matter density in the lateral geniculate nucleus than HC, which was correlated with the brain functional activation difference in V1 between AME and fellow eye.
Lu et al., 2020	3DT1, rfMRI	VBM and ReHo (whole brain)	PediatricAA: 16OS	10M/6F, 6.9±2.3 y	Untreated	8M/6F, 8.1±2.2 y	Lower ReHo in rightSTG and MFG and larger GMV in the right lobules 4 and 5 of the cerebellum and fusiform gyrus (which was negatively correlated with theReHo of MFG in patients) were foundin pediatric AA than controls.
Lin et al., 2012	3DT1, rfMRI	VBM and ReHo (whole brain)	AA: 7OS, 5OD, 1OU	5M/8F, 22.3±7.2 (17-43) y	NA	8M/13F, 23.5±2.1 y	In AA, there was no GMV alteration but lower ReHo in the right precuneus, left medial prefrontal cortex, inferior frontal gyrus, and cerebellum, as well as higher ReHo in the right middle occipital gyrus, left STG, and left fusiform gyrus compared with HC.
Qi et al., 2016	3DT1, DTI	SBM (whole brain), FA (ROI)	PediatricAA: 8OS, 7OD	9M/6F, 8.0±1.3 (6.5-10.3) y	NA, but excluding patients who underwent long-term amblyopiatreatment	8M/7F, 7.9±0.8 y	AA showed thicker cortical thickness in V1 but thinner extrastraite cortex of V2, lateral occipitotemporal gyrus, occipital lobe, inferior parietal lobe, and superior occipital cortex than HC. The FA in the bilateral optic radiations was correlated with the thinner cortex of V2, occipital gyrus, lateral occipitotemporal gyrus, lunate cortex.
Li et al., 2013b	3DT1, DTI	VBM and FA (whole brain)	PediatricAA: 7OS, 8OD	10M/5F, (6-15) y	Untreated	15, (5-14) y	Compared to HC, AA showed reduced GMV in the left inferior occipital gyrus, increased GMV in the lingual gyrus, lower FA in the optic radiation, andhigher FA in the rightprecuneus and left middle occipital area in patients.
Li et al., 2015	DTI	FA (whole brain, ROI)	PediatricAA: 8OS, 12OD	15M/5F, 9.6±2.9 (5-14) y	Glasses 5/20	9M/9F, 10.8±2.1 (5-14) y	Smaller FA values was found in the right optic radiation, left ILF/IFO and right SLF of patients than HC, which were correlated with visual acuity of amblyopes.
Allen et al., 2018	DTI	FA and MD (ROI)	AA: 6OS, 2OD; SA: 4OS, 1OD;	AA: 4M/4F, (13-37) y; SA: 4M/1F, (14-40) y	allhad undergone treatment while the details were NA	13, 21.0±5.5 (14-34) y	AA showed lower FA values in the optic nerve and optic tract and greater MD in the optic radiation than controls. No differences between SA and AA nor between SA and controls.
Duan et al., 2015	DTI	FA and MD (ROI)	Adult SA: 13ES, 3EX	8M/8F, 45.3±13.3 (20-67) y	Surgery 11/16	15M/17F, 45.0±13.8y	Higher MD values were found in the left optic radiation, anterior frontal corpus callosum, right vertical occipital fasciculus, and left ILF of SA than HC.
Gupta et al., 2016	DTI, tfMRI	FA and Brain activation (ROI)	PediatricSA: 5OS, 4OD	4M/5F, 8±1.8 (6-10) y	Base line: untreated Patch:5	4M/6F, 8.8±2 (7-13) y	Under a flickering checkerboard task, less activation in the left occipital lobe evoked to the amblyopic eye was found in SA before the treatment than HC. The activation in striate cortex and exstrastriate cortex were increased after the treatment in patients. The FA of V1 and V5 in SA were significantly lower thancontrols, but thatwas nodifference after occlusion therapy.

*Age represented as mean ± standard deviation or range. 3DT1: Three-dimensional T1-weighted imaging; AA: anisometropic amblyopia; DTI: diffusion tensor imaging; ES: esotropia; EX: exotropia; F: female; FA: fractional anisotropy; GMV: gray matter volume; HC: healthy controls; ILF/IFO: inferior longitudinal fasciculus/inferior fronto-occipital fasciculus; M, male; MD, mean diffusivity; MFG: middle frontal gyrus; MRI: magnetic resonance imaging; MT+: middle temporal complex; NA: not applicable; OD: right eye; OS: left eye; OU: both eyes; ReHo: regional homogeneity; ROI: region of interest; rs/tfMRI: resting-state/task-based functional MRI; SA: strabismic amblyopia; SBM: surface-based morphometry; SLF: superior longitudinal fasciculus; STG: superior temporal gyrus; VBM: voxel-based morphometry; WMV: white matter volume; y: years