Animals and Perfusion System

The cerebral hypoxia-ischemia was simulated in vitro using a perfusion system, following previously published methods [18]. We used 3-week-old Sprague-Dawley rats (Shanghai Children’s Medical Center, Shanghai), as their brains are in the developmental stage [23, 24]. This study was approved by the Institutional Animal Care and Use Committee of Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of Medicine.

Rats were placed on ice, rapidly decapitated, and their brains were immediately immersed in ice-cold artificial slicing solution containing (mmol/L): 125 NaCl, 2.5 KCl, 1.25 NaH2PO4, 25 NaHCO3, 25 Glucose, 75 Sucrose, 1 MgCl2, 2 CaCl2, equilibrated with 95% O2/5% CO2, pH 7.4. Brain slices (400 μm) were prepared using a vibratome (VT1000S, Leica Microsystems, Mannheim, Germany). Only slices containing the corpus callosum—the primary white matter structure in the rat brain—were selected for perfusion experiments. The slices were incubated in oxygenated artificial cerebrospinal fluid (aCSF) at 31℃ for at least 1 h before further experimentation. Brain slices were transferred to the perfusion system and subjected to two distinct perfusion protocols: (1) Oxygenated glucose-aCSF (mmol/L): 125 NaCl, 2.5 KCl, 1.25 NaH2PO4, 25 NaHCO3, 10 Glucose, 1 MgCl2, 2 CaCl2, equilibrated with 95% O2/5% CO2, pH 7.4. (2) Hypoxic sucrose-aCSF (mmol/L): 125 NaCl, 2.5 KCl, 1.25 NaH2PO4, 25 NaHCO3, 10 Sucrose, 1 MgCl2, 2 CaCl2, equilibrated with 95% N2/5% CO2, pH 7.4. The temperature was maintained constant throughout all experimental procedures.

Oxygen-Glucose Deprivation

Oxygen-glucose deprivation (OGD) was used to simulate hypoxia-ischemia. This was achieved by hypoxic sucrose-aCSF, as described previously. After OGD, the brain slices were reperfused with oxygenated glucose-aCSF and equilibrated with 95% O2/5% CO2.

Perfusion Procedures

Three hypothermic protection strategies employed in clinical implementation: normothermia (31℃), moderate hypothermia (25℃), and deep hypothermia (18℃). Building on this, OGD was conducted under the following conditions: (1) 31℃ sham group; (2) OGD at 18℃, 25℃, and 31℃. Before OGD, the perfusion system was cooled to target temperatures for 5 min, followed by 60 min of OGD. Then the brain slices underwent 10 min of reperfusion with oxygenated glucose-aCSF, subsequently rewarmed to 31℃ for 30 min.

Weanling Rats DHCA Model

Male Sprague-Dawley rats (3 weeks old) were maintained under standard conditions with access to food/water. Following anesthesia induction with 2% isoflurane (RWD Life Science, R510-22-10, Shenzhen, China), tracheal intubation was performed using an 18-gauge cannula (Xi’an Xijing Medical Appliance, Xian, China) connected to a ventilator (RWD Instruments, Shenzhen, China). ECG electrodes (Philips Intellivue mp2, Amstelplein, Netherlands) were positioned on three limbs before triple disinfection of the chest with iodophor (60005-1, Yuyan Biology, Shanghai, China). A scalpel incision extended from the chest to the xiphoid process, with immediate hemostasis achieved using an electrosurgical unit (Huihan Technology, Suzhou, China) while avoiding major vessels. After thoracic cavity exposure via thymus excision (electrocautery) and sternal retractor placement (Tigergene Technology, Nanjing, China), the right common carotid artery (RCCA) was blunt-dissected (mosquito forceps FS027, Beyotime, Shanghai, China) and looped proximally/distally with 2-0 sutures. A 22-gauge catheter was advanced into the RCCA to the ascending aorta and secured with sutures. Similarly, the superior vena cava (SVC) was dissected, incised proximally, and cannulated with a custom venous catheter advanced to the right atrium. Heparin (500 U/kg, 1 mL) was administered pre-cannulation. Post-cannulation stability was confirmed by normal ECG and vital signs. The circuit connected to the RCCA catheter (pump flow: 1–2 mL/min) and the SVC catheter-reservoir. Initial water bath temperature was set at 31℃; after heart rate stabilization, pump flow increased to 80–100 mL/kg per min. The water tank temperature was gradually adjusted to 18℃ while continuously monitoring the rats’ rectal temperature. When the rectal temperature reached 18℃, the pump was stopped, and blood from the SVC was drained into the reservoir. Upon cessation of heart rate, drainage was halted, and the circuit was closed. Gradual rewarming was then initiated until a core temperature of 31℃ was restored, after which cardiopulmonary bypass (CPB) was discontinued.

Method of Disulfiram Administration

For the perfusion system, 10 mmol/L disulfiram was dissolved in DMSO and further diluted to 10 μmol/L in PBS [25]. For DHCA, rats were treated with DSF (50 mg/kg) formulated in corn oil or vehicle by intraperitoneal injection 4 h in advance of DHCA surgery [26].

Quantitative Real-Time PCR (qPCR)

Total RNA was extracted from frozen tissue specimens (−80℃) using Trizol reagent (R0016, Beyotime, Shanghai, China) following the manufacturer’s protocol. Briefly, tissues were homogenized in 1 mL Trizol, mixed with 200 μL chloroform, and incubated at room temperature for 5 min. After centrifugation (12,000 ×g, 4℃, 15 min), the aqueous phase was collected and combined with an equal volume of isopropanol to precipitate RNA. The RNA pellet was washed twice with 75% ethanol (12,000 ×g, 4℃, 15 min each) and resuspended in 20 μL nuclease-free water. cDNA was synthesized from the isolated RNA. Target genes (Nlrp3, Casp1, Gsdmd, Mbp, Il1β, and Il18) were amplified using KAPA HiFi HotStart ReadyMix (KK2601, KAPA Biosystems, Wilmington, USA) with gene-specific primers (Table S1). Gapdh expression was used as the endogenous control for normalization.

Western Blot

Tissues were homogenized in ice-cold lysis buffer consisting of 50 mmol/L Tris-HCl (pH 7.6), 150 mmol/L NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and protease inhibitors. Lysates were centrifuged (12,000 ×g, 15 min, 4℃) to remove debris, and protein concentrations were determined using the BCA assay (Thermo Fisher Scientific, Waltham, USA). For SDS-PAGE, protein extracts were denatured (65℃, 30 min) and resolved in Tris-glycine buffer. Proteins were transferred to PVDF membranes (162-0177, Bio-Rad, Berkeley, USA) and blocked with 3% BSA (BBI Life Sciences, A600332-0100) in TBS-T (20 mmol/L Tris, 150 mmol/L NaCl, 0.1% Tween 20, pH 7.6) for 2 h at room temperature. Primary antibodies were incubated overnight at 4℃ using the following: mouse anti-GSDMD (1:1000, Santa Cruz, Dallas, USA, Cat# sc-393656, RRID: AB_2728694), mouse anti-MBP (1:500, Millipore, Darmstadt, Germany, Cat# MAB382, RRID: AB_94971), rabbit anti-cleaved CASP1 (1:1000, Cell Signaling, Danvers, USA, Cat# 89332, RRID: AB_2923067), rabbit anti-NLRP3 (1:1000, Cell Signaling, Danvers, USA, Cat# 15101, RRID: AB_2722591), rabbit anti-IKBα (1:1000, Abcam, Cambridge, United Kingdom, Cat# ab32518, RRID: AB_733068), rabbit anti-p-IKBα (1:1000, Abcam, Cambridge, United Kingdom, Cat# ab133462, RRID: AB_2801653), rabbit anti-p65 (1:1000, Abcam, Cambridge, United Kingdom, Cat# ab16502, RRID: AB_2224674), rabbit anti-p-p65 (1:1000, Cell Signaling, Danvers, USA, Cat# 3033, RRID: AB_331284), rabbit anti-p-MLKL (1:1000, Abcam, Cambridge, United Kingdom, Cat# ab196436, RRID: AB_2687465), rabbit anti-cleaved CASP3 (1:1000, Cell Signaling, Danvers, USA, Cat# 9661, RRID: AB_2341188) and rabbit anti-β-actin (1:2000, Invitrogen, Waltham, USA, Cat# MA515739, RRID: AB_10979409). All antibodies used are detailed in Table S2. Membranes were washed with TBS-T and incubated with HRP-conjugated secondary antibodies (anti-rabbit CW0103, 1:5000; anti-mouse CW0102, 1:5000, Cwbio, Beijing, China) for 2 h at room temperature. Blots were developed using Clarity™ Western ECL Substrate (Bio-Rad, Berkeley, USA) and imaged on a Tanon Chemiluminescence system. Band intensities were quantified using ImageJ v1.30 (National Institutes of Health, USA).

Immunofluorescent Staining

Brain tissue sections (400 μm thick) were fixed overnight in 4% paraformaldehyde (PFA) at 4℃, embedded, and coronally sectioned at 35 μm using a Cryo-Vibratome to encompass the corpus callosum. After permeabilization with 0.3% Triton X-100/PBS for 13 min, sections were blocked for 2 h at room temperature with 10% donkey serum (w9030-05, Ruite Biotechnology, Guangzhou, China) in PBS-T (PBS + 0.1% Triton X-100). Primary antibodies were diluted in PBS with 5% donkey serum and incubated at 37℃ for 2 h, followed by overnight incubation at 4℃. Antibodies included: goat anti-Iba-1 (1:500, Wako, Tokyo, Japan, Cat# 011-27991, RRID: AB_2935833), rabbit anti-GSDMD (1:200, Abcam, Cambridge, United Kingdom, Cat# ab209845, RRID: AB_2783550), mouse anti-MBP (1:500, Millipore, Darmstadt, Germany, Cat# MAB382, RRID: AB_94971), rabbit anti-IL-18 (1:200, Abcam, Darmstadt, Germany, Cat# ab191152, RRID: AB_2737346), rabbit anti-S100β (1:500, Abcam, Darmstadt, Germany, Cat# ab52642, RRID: AB_882426), mouse anti-Olig2 (1:500, Millipore, Darmstadt, Germany, Cat# MABN50, RRID: AB_10807410), rabbit anti-GFAP (1:500, Millipore, Darmstadt, Germany, Cat# AB5804, RRID: AB_2109645), mouse anti-CD68 (1:500, Abcam, Cambridge, United Kingdom, Cat# ab955, RRID: AB_307338), mouse anti-Ki67 (1:500, Cell Signaling, Danvers, USA, Cat# 9449, RRID: AB_2797703) and mouse anti-APC (CC-1) (1:500, Millipore, Darmstadt, Germany, Cat# OP80, RRID: AB_2057371). After washing with PBS-T, sections were incubated with biotinylated secondary antibodies for 2 h at 37℃, including Donkey anti-Goat Alexa Fluor 647 (1:500, Invitrogen, Waltham, USA, Cat# A-21447, RRID: AB_2535864), Donkey anti-Rabbit Alexa Fluor 488 (1:500, Invitrogen, Waltham, USA, Cat# A-21206, RRID: AB_2535792), Donkey anti-Mouse Alexa Fluor 488 (1:500, Invitrogen, Waltham, USA, Cat# A-21202, RRID: AB_141607), Donkey anti-Rabbit Alexa Fluor 647 (1:500, Invitrogen, Waltham, USA, Cat# A-31573, RRID: AB_2536183), Donkey anti-Mouse Alexa Fluor 647 (1:500, Invitrogen, Waltham, USA, Cat# A-31571, RRID: AB_162542) and Donkey anti-Mouse Alexa Fluor 568 (1:500, Invitrogen, Waltham, USA, Cat# A10037, RRID: AB_11180865). Brain sections were counterstained with DAPI (1:1000, Cell Signaling, Danvers, USA, 4083S) for 10 min at room temperature before mounting with Fluoromount™ aqueous medium (13800, AQUA-MOUNT, Burlingame, USA). Images were acquired using a Leica TCS SP8 confocal microscope with HC PL APO CS2 objectives (×20/0.75 dry, ×40/1.30 oil, ×63/1.40 oil) and analyzed in ImageJ v1.30.

Enzyme-Linked Immunosorbent Assay (ELISA)

The aCSF from the perfusion system and the serum samples from DHCA rats were collected. Target protein concentrations were quantified using rat-specific ELISA kits (Amoy Lunchangshuo Biotech, Xiamen, China) according to the manufacturer’s instructions. The ELISA was performed following standard protocols with minor modifications. A 96-well microplate was coated overnight at 4℃ with the target antigen (or capture antibody) diluted in 0.1 mol/L carbonate-bicarbonate buffer (pH 9.6). The plate was then blocked with 5% bovine serum albumin in PBS for 1–2 h at room temperature to prevent nonspecific binding. After blocking, the plate was washed three times with PBS containing 0.05% Tween-20. Serial dilutions of standards and test samples were added to the wells and incubated for 2 h at 37℃. Following incubation, the plate was washed, and a biotinylated detection antibody was added for an additional 1 h at 37℃. After another washing step, streptavidin-horseradish peroxidase conjugate was applied and incubated for 30 min at room temperature. The reaction was developed using 3,3',5,5'-tetramethylbenzidine substrate, and the enzymatic reaction was stopped with 2 mol/L sulfuric acid. Absorbance was measured at 450 nm using a microplate reader, and sample concentrations were determined from a standard curve.

Hoechst 33342/Propidium Iodide (PI) Staining

Primary microglia underwent double-staining using a Hoechst 33342/PI Apoptosis Assay Kit (BL116A, Biosharp, Hefei, China) per manufacturer instructions. Cells were harvested and washed twice with ice-cold PBS. Following centrifugation at 300 × g for 5 min, the pellet was resuspended in PBS containing 5 μg/mL Hoechst 33342 and 2 μg/mL propidium iodide. The cell suspension was incubated for 15 min at 37℃ in the dark. After staining, cells were immediately analyzed by fluorescence microscopy (Eclipse Ti, Nikon, Tokyo, Japan) using UV excitation (350 nm) for Hoechst 33342 and green excitation (535 nm) for PI detection. Images were acquired on a Leica TCS SP8 confocal microscope with HC PL APO CS2 objectives.

Annexin V-FITC/Propidium Iodide (PI) Flow Cytometry

The tissue samples were digested into cells and then stained using the Annexin V-FITC Apoptosis Detection Kit (Millipore, Darmstadt, Germany). Briefly, 1×106 cells were harvested, washed twice with ice-cold PBS, and resuspended in 100 μL of 1× binding buffer. Subsequently, cells were stained with 5 μL Annexin V-FITC and 5 μL propidium iodide (PI, 50 μg/mL) for 15 min at room temperature in the dark. Before analysis, 400 μL of 1× binding buffer was added to each sample. Subsequent quantification was performed on a NovoCyte 2040R flow cytometer (ACEA Bioscience, San Diego, USA).

Primary Microglia Culture

During primary microglia cell culture, the brains of 24-hour-old neonatal rats were dissected, and meninges were removed in Hanks’ Balanced Salt Solution (HBSS). The rats’ tissues were transferred to digestion solution [(trypsin solution, 0.25% (Gibco, Baltimore, USA); Dnase I, 75 U/mL (Worthington, Lakewood, USA)] and incubated for 10 min in the culture incubator at 37℃. Cells were pelleted by centrifugation (1000 × g, 5 min) using a swinging bucket rotor and cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, Baltimore, USA) supplemented with 10% fetal bovine serum (FBS) in 75 cm2 flasks at 37℃ under 5% CO₂. The medium was replaced after 24 h and refreshed twice weekly. Microglia cell purification was performed as previously described [27]. Briefly, on day 21, cells were digested with 0.0625% trypsin-HBSS for 1–2 h at 37℃. The astrocyte-enriched supernatant was removed, and adherent microglia were maintained in DMEM. For maturation, purified microglia were cultured for 3 d before experimentation.

Transmission Electron Microscopy

Under a dissecting microscope, the corpus callosum was isolated and minced into 1–3 mm3 fragments before primary fixation in 2.5% glutaraldehyde (> 3 h). After 0.1 mol/L phosphate buffer washes, specimens underwent secondary fixation with 0.5% osmium tetroxide under agitation (4℃, 3 h), followed by additional phosphate buffer rinses. Tissues were progressively dehydrated through an ethanol series, infiltrated with propylene oxide, and embedded in Spurr’s resin for polymerization (70℃). Ultrathin sections (≈ 70 nm) prepared using UC7 ultramicrotome (Leica, Mannheim, Germany) were contrasted with uranyl acetate and lead citrate before imaging on Talos 120 transmission electron microscope (Thermo Fisher Scientific, Waltham, USA). The corpus callosum was microdissected into 1–3 mm3 fragments under a dissecting microscope and fixed in 2.5% glutaraldehyde (> 3 h). After washing with 0.1 mol/L phosphate buffer, samples were post-fixed in 0.5% osmium tetroxide (4℃, 3 h) with agitation, followed by additional buffer rinses. Tissues were dehydrated through a graded ethanol series, cleared in propylene oxide, and embedded in Spurr’s resin for polymerization (70℃). Ultrathin sections (≈ 70 nm) were cut using a UC7 ultramicrotome, stained with uranyl acetate and lead citrate, and imaged on a Talos 120 transmission electron microscope.

Single-cell Sequencing Analysis

The acquisition of corpus callosum tissue sections was performed as described previously. The corpus callosum tissue was subjected to single-cell RNA-seq at Shanghai Biotechnology Corporation (Seekgene, Beijing, China). Single-cell RNA Seq libraries were prepared using SeekOne® MM Single Cell 3’ library preparation kit (SeekGene). Briefly, an appropriate number of cells were loaded into the flow channel of the SeekOne® MM chip, which had 170,000 microwells and allowed to settle in the microwells by gravity. After removing the unsettled cells, sufficient Cell Barcoded Magnetic Beads (CBBs) were pipetted into the flow channel and also allowed to settle in microwells with the help of a magnetic field. Next, excess CBBs were rinsed out, and cells in the MM chip were lysed to release RNA, which was captured by the CBB in the same microwell. Then all CBBs were collected, and reverse transcription was performed at 37℃ for 30 min to label cDNA with the cell barcode on the beads. Further Exonuclease I treatment was performed to remove unused primer on CBBs. Subsequently, barcoded cDNA on the CBBs was hybridized with a random primer that had reads 2 SeqPrimer sequence on the 5’ end and could extend to form the second strand DNA with cell barcode on the 3’ end. The resulting second-strand DNA was denatured off the CBBs, purified, and amplified in a PCR reaction. The amplified cDNA product was then cleaned to remove unwanted fragments and added to full length sequencing adapter and sample index by indexed PCR. The indexed sequencing libraries were cleaned up with SPRI beads, quantified by quantitative PCR (KK4824, KAPA Biosystems, Wilmington, USA), and then sequenced on Illumina NovaSeq 6000 with PE150 read length or DNBSEQ-T7 platform with PE100 read length. The Seurat R package (version 3.2.0) was used for further inspection and data analysis [28]. The resulting filtered matrix consisted of ~60,000 cells. The matrix was normalized using the NormalizeData function, and variable features were identified using the FindVariableFeatures function with 2000 genes. The ScaleData function was used to center the gene expression. Next, principal component analysis (PCA) was performed, using the RunPCA function, to obtain the top 50 principal components (PCs). Clustering was conducted using the FindNeighbors and FindClusters functions using 20 PCs and a resolution parameter set to 0.3. Differential gene expression analysis was performed using PRESTO (Supplementary File 2). For visualization, the dimensionality of the data sets was reduced by UMAP. Cell populations were matched to cell types based on the expression of known marker genes and previously identified expression signatures [29]. Gene enrichment was analyzed using Metascape. Single-cell RNA-seq data are available in PRJNA1312850.

Rat Cytokine Array Panel A

Collect corpus callosum tissue from both the sham and OGD groups. Add RIPA lysis buffer, centrifuge (12,000 × g, 15 min, 4℃), and take the supernatant. Sonicate for 20 s. Equilibrate the array membrane at room temperature for 30 min, then wash with 1× Wash Buffer for 5 min. Add 300 μL of diluted sample (containing an equal amount of total protein, recommended ≥ 1 mg/mL) to each membrane, and incubate at room temperature overnight at 4℃. Wash the membrane three times with 1× Wash Buffer, 10 min each time. Add biotin-labeled antibody mixture and incubate at room temperature for 2 h. Wash the membrane three times with 1× Wash Buffer, 10 min each time. Add streptavidin-HRP and incubate at room temperature for 30 min. Wash the membrane three times with 1× Wash Buffer, 10 min each time. Cover the membrane with ECL chemiluminescence reagent and incubate in the dark for 1 min. Then the membranes were visualized using ClarityTM Western ECL Substrate (Bio-Rad, Hercules, USA) and imaged on a Tanon Chemiluminescence system. Equivalence of protein loading was corrected by probing for the standard value. Band densities were quantified using ImageJ v1.30 (National Institutes of Health, USA).

Magnetic Resonance Imaging (MRI)

MRI was performed at Shanghai Children’s Medical Center using a 3.0T scanner (Siemens) with a 64-channel infant head coil and cardiorespiratory monitoring. Postoperative pediatric patients (Table S3) were scanned. Anatomical imaging included T1-weighted (repetition time [TR], 2400 milliseconds (ms), echo time [TE], 2.22 ms, slice thickness, 0.80 mm, field of view (FOV), 256 mm, matrix, 320 × 320) and T2-weighted (TR, 3200 ms, TE, 563 ms, slice thickness, 0.8 mm, FOV, 256 mm) sequences. Resting-state MRI (rs-fMRI) acquired 420 volumes over 336 s using an echo-planar imaging (EPI) sequence (TR, 800 ms, TE, 37 ms, FOV, 208 mm, matrix, 104 × 104, slice thickness, 2mm, 72 slices, 144 volumes). Subjects maintained relaxed wakefulness with head immobilization. Diffusion tensor imaging (DTI) used a single-shot EPI sequence (TR, 4200 ms, TE, 89 ms, FOV, 240 mm, matrix, 140 × 140, slice thickness, 2mm, motion-probing gradient in 102 diffusion-encoding directions with a diffusion weighting of 3000 s/mm2 [b value] and 10 times of non-diffusion weighted image). White matter injury (WMI) was graded according to established criteria for congenital heart disease based on imaging characteristics defined in prior literature [7, 30], assessed by neurologists.

DTI and Atlas-Based Analysis

Diffusion tensor imaging (DTI) analysis was performed using FSL (www.fmrib.ox.ac.uk/fsl) [31] following established methods [32]. Preprocessing included eddy current correction, linear registration of diffusion-weighted volumes to the non-diffusion volume (b0) [33, 34], and brain masking [35]. The diffusion tensor model was fitted voxelwise to compute fractional anisotropy (FA) maps, where FA quantifies regional water diffusion directionality in white matter and increases with microstructural maturation [36]. The JHU (Johns Hopkins University) atlas was registered to each child’s individual diffusion-weighted imaging (DWI) space using the ANTs toolkit (https://github.com/ANTsX/ANTs) through a multi-step process. First, the individual’s DWI image was linearly registered to their corresponding T2-weighted structural image. Second, this T2 structural image was nonlinearly registered to a pediatric standard template (the UNC-BCP 4D Infant Brain Volumetric Atlas 2 months). Subsequently, the resulting linear and nonlinear transformation fields were combined to compute their composite inverse transformation field. Finally, the JHU-ICBM-labels-2 mm atlas [37] was warped into the individual child’s space using this composite inverse transformation field and interpolated, enabling atlas-based parcellation for extracting fractional anisotropy (FA) metric values from the participant’s FA image.

Tractography Methodology

Fiber tracking employed DSI Studio using generalized q-sampling imaging (sampling length ratio = 1.25) to reconstruct orientation distribution functions. Quantitative anisotropy (QA) values guided automated tract identification per published protocols [38].

Ethical Approval

The ethical approval was reviewed and approved by the Institutional Animal Protection and Use Committee of Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of Medicine (SCMC-LAWEC-2022-613). Ethical approval for research involving human participants was granted by Shanghai Children’s Medical Center, Shanghai Jiao Tong University School of Medicine (approval number: SCMCIRB-K2025167-1).

Statistical Analysis

All statistical analysis was run in GraphPad Prism 9.0. The graphs were created in GraphPad Prism 9.0. Data are presented as means ± SEM, and the error bars represent SEM for each set of data to be compared. All data were first assessed for normality using the Shapiro-Wilk test and for homogeneity of variances using the F test. Based on the outcome of these tests, appropriate parametric or non-parametric tests were chosen for group comparisons. Specifically, for comparisons between two groups that followed a normal distribution, an unpaired two-tailed t test was used if variances were equal, or Welch’s corrected t test was applied if variances were unequal. For data that did not meet the normality assumption, the Mann-Whitney test was used. Statistical comparisons among three or more groups were performed using a one-way analysis of variance (ANOVA). Researchers were blinded to the groups or samples during the experiments. Regions with FA reduction (DHCA vs. CPB: ΔFA < 0) were mapped to the WMI probability area. The statistical methods were not used to pre-determine sample size or to randomize. Statistical significance was set at *P < 0.05, **P < 0.01, ***P < 0.001.