Axonal loss is the primary cause of permanent neurological disability in spinal cord injury.^1,2 Diffusion tensor imaging (DTI) derived fractional anisotropy (FA) has been widely sued to non-invasively and longitudinally assess the severity of SCI. However, co-existing axonal injury/loss, demyelination, and inflammation significantly interfere with the accuracy of FA assessed SCI severity. In this study, we applied diffusion basis spectrum imaging (DBSI)^3,4 to detect, distinguish, and quantify axonal pathologies in a mouse model of SCI in vivo, followed by immunohistochemistry (IHC) validation. Our results support that DBSI may serve as outcome measure to quantitatively reflect irreversible damage in SCI.Animal model: A contusive SCI at T9 vertebral level was performed in eleven of seventeen 10-week-old female C57BL/6 mice. The remaining six mice (sham) underwent laminectomy without contusion. The injury group received contusion injury at 0.1 m/s with 0.8 mm impact displacement using 1.3-mm diameter rounded tip. After impact, the site was closed in layers followed by subcutaneous administration of Baytril (2.5mg/kg) and lactated Ringers (5mL). Standard postoperative care was performed.⁵ DBSI: A pair of 8-cm diameter volume and 15 × 8 mm² surface active-decoupled coils was used to cover T8 – T11 vertebrae. DBSI was performed at day 3 post contusion on a 4.7-T Agilent small-animal MR scanner utilizing a multiple-echo spin-echo diffusion-weighted sequence.⁶ A 25-direction diffusion scheme was performed.⁷ All images were obtained with following acquisition parameters: TR = 0.3 (with respiratory gating) s, TE = 43 ms, inter-echo delay = 16 ms, Δ = 25 ms, δ = 6 ms, maximal b-value = 2,200 s/mm², slice thickness = 1.8 mm, FOV (field of view) = 10 × 10 mm², in-plane resolution = 104 × 104 µm² (before zero-filled). Data analysis: A lab-developed DBSI code was performed on diffusion weighted MR data to estimate $$$\lambda$$$_{$$$\parallel$$$}, $$$\lambda$$$_$$$\perp$$$, and FA derived by DBSI and DTI, and DBSI specific fiber, restricted (putative cellularity) and non-restricted isotropic (putative edema) diffusion tensor fractions. Histology: mice were perfusion fixed immediately after the in vivo diffusion weighted MR measurements for immunohistochemical (IHC) staining of SMI-312, MBP, DAPI, and SMI-31. DTI and DBSI derived $$$\lambda$$$_{$$$\parallel$$$}, $$$\lambda$$$_$$$\perp$$$, and FA (Fig. 1 and 2) revealed that DBSI specifically reflected axonal pathologies without confounding effects from inflammation and axonal loss. The extent of inflammation, i.e., increased cellularity and edema, and axonal loss was reflected by DBSI derived restricted and non-restricted isotropic, and anisotropic (injured and non-injured axonal fibers) tensor fractions, respectively (Fig. 3, and 4). White matter volume estimated using diffusion weighted anatomic images increased at 3 days after SCI (Fig. 4C). In contrast, axonal fiber fraction (Fig. 4D) derived by DBSI significantly decreased in SCI group. A new metric “axon volume” was derived, i.e., fiber fraction × white matter volume, to quantitatively estimate the extent of axonal loss. DBSI derived “axon volume” significantly decreased in SCI (Fig. 4E). IHC confirmed the DBSI detected/quantified co-exiting pathologies, including axon and myelin damage, inflammation, and axon loss in mouse SCI.DBSI detected, distinguished, and quantified the co-exiting axonal pathologies in SCI mice 3 days after contusion, confirmed by post-MRI quantitative IHC. Results demonstrated that DBSI can “see through” confounding co-existing pathologies, e.g., inflammation and tissue loss, for quantifying the extent of axonal injury, demyelination, inflammation, and the irreversible axonal loss.