Adverse childhood experiences have lasting detrimental effects on mental health and are strongly associated with impaired cognition and increased risk of developing psychopathologies. Preclinical and neuroimaging studies have suggested that traumatic events during brain development can affect cerebral myelination particularly in areas and tracts implicated in mood and emotion. Although current neuroimaging techniques are quite powerful, they lack the resolution to infer myelin integrity at the cellular level. Recently demonstrated coherent Raman microscopy has accomplished cellular level imaging of myelin sheaths in the nervous system. However, a quantitative morphometric analysis of nerve fibers still remains a challenge. In particular, in brain, where fibres exhibit small diameters and varying local orientation. In this work, we developed an automated myelin identification and analysis method that is capable of providing a complete picture of axonal myelination and morphology in brain samples. This method performs three main procedures 1) detects molecular anisotropy of membrane phospholipids based on polarization resolved coherent Raman microscopy, 2) identifies regions of different molecular organization, 3) calculates morphometric features of myelinated axons (e.g. myelin thickness, g-ratio). We applied this method to monitor white matter areas from suicides adults that suffered from early live adversity and depression compared to depressed suicides adults and psychiatrically healthy controls. We demonstrate that our method allows for the rapid acquisition and automated analysis of neuronal networks morphology and myelination. This is especially useful for clinical and comparative studies, and may greatly enhance the understanding of processes underlying the neurobiological and psychopathological consequences of child abuse.