16.1 Introduction

Vision is a multistep process that involves several different anatomical regions whereby optical signals begin in the retina and are transmitted throughout various parts of the brain. ^{1 ,​ 2 ,​ 3} In order for visual perception to occur, the physiological signal must be identified by the retina, the impulses must be transmitted through the optic nerves to the optic pathways, and then these nerve signals must be processed by various areas of the brain including the occipital, parietal, and temporal lobes and eventually the frontal lobes. ^{1 ,​ 2 ,​ 3} This process is complex and contributes the human ability to perform higher level tasks in addition to just visualizing an object, such as object identification, motion acquisition, recognition, and naming, among others. ^{1 ,​ 2 ,​ 4} Visual functions, including visual fields and visual processing, are critical components of one’s daily functioning and intraoperative damage to these pathways can significantly affect one’s quality of life. ⁵ Lesions, namely gliomas, can reside adjacent to and/or interfere with the function of these visual pathway components. ^{1 ,​ 2 ,​ 4} An understanding of these pathways and how to map these pathways intraoperatively can help facilitate extensive resection and complication avoidance. ^{1 ,​ 2 ,​ 4 ,​ 5} In this chapter, we report how we perform awake brain mapping with direct electrical stimulation for lesions involving the visual pathways.

16.2 Visual Pathway and Cognitive Processing of Vision

Visual perception is the process whereby an object is visualized and cognitively recognized. ^{1 ,​ 2 ,​ 3} This process involves multiple steps, several different anatomical structures, and a connectome that spans the entire sagittal distance of the brain. ^{6 ,​ 7 ,​ 8} Lesions can occur along this entire pathway and lead to different deficits depending on the location and degree of pathway disruption. ^{6 ,​ 7 ,​ 8} Knowledge about the process of visual perception is in constant evolution as the brain connectome is being better understood with brain imaging and brain mapping techniques. ^{1 ,​ 2 ,​ 3}

Light strikes the retinal ganglion cells and the impulse is conducted through the retinal ganglion cell axons into the optic nerve. ^{6 ,​ 7 ,​ 8} The optic nerve then traverses from the retina through the intraorbital space, followed by the optic canal, and then to the optic chiasm where there is partial crossing of the temporal optic nerve axons. ^{6 ,​ 7 ,​ 8} The postchiasmatic optic nerve or optic tract, which carries axons from both optic nerves, synapses in the lateral geniculate nucleus of the thalamus. ^{6 ,​ 7 ,​ 8} The nerve fibers from the lateral geniculate nucleus form the optic radiations that travel through the white matter adjacent to the lateral ventricles to the primary and late visual cortices in the occipital, temporal, and parietal lobes. ^{6 ,​ 7 ,​ 8}

The occipital lobe is referred to as the primary visual cortex, while the temporal and parietal lobes are the late cortices. ^{6 ,​ 7 ,​ 8} In the primary occipital cortex, the lower bank of the calcarine sulcus represents the upper visual field, while the upper back represents the lower visual field. ^{6 ,​ 7 ,​ 8} Moreover, the right visual field is represented in the left occipital lobe and the left visual field is represented in the right occipital lobe. ^{6 ,​ 7 ,​ 8} The fovea is represented in a large cortical area within the occipital lobe where there are a large number of neurons that enable fine spatial vision. ^{6 ,​ 7 ,​ 8} The ability to visualize an object, therefore, requires the optic pathway from the retina to the primary visual cortex of the occipital lobe to be intact. ^{1 ,​ 2 ,​ 3}

Although visualization is a primary function, there are other hierarchical functions that vision is a component of. These hierarchical systems are involved in a number of tasks including object recognition, object naming, detection of motion, and memory, among others, and require input from other sensory systems and brain regions. ^{1 ,​ 2 ,​ 3} In this hierarchical system, several different networks combine their inputs from lower levels and the information is combined and processed in higher levels that allow for visual analysis. ^{1 ,​ 2 ,​ 3} As with language and hearing, there are ventral and dorsal streams that subserve vision. ⁹ In this dual-stream model, the ventral stream is involved with object identification and recognition, while the dorsal stream is involved with spatial location. ⁹ The ventral stream involves the processing of information within the parvocellular layer of the lateral geniculate nucleus of the thalamus and projects this information to the V1 cell layer of the occipital cortex, followed by the V2 and V4 cell layers, and then to the inferior temporal lobe primarily through inferior longitudinal fasciculus (ILF). ⁹ Damage to this region results in difficulty with object recognition, but mostly recognizing faces and facial expression. ⁹ The posterior portion of the ILF, especially the temporal-occipital area plays a crucial role in reading. ⁹ In comparison to the ventral stream, the dorsal stream involves the V1 layer of the primary visual cortex of the occipital lobe and transmits information into the parietal lobe, and plays a role in detecting and analyzing movements and spatial awareness. ⁹ This spatial awareness is most prominent at the right parieto–temporal junction and the information is transmitted through the second portion of the superior longitudinal fasciculus (SLF II). ⁹ Damage to this region results in incoordination and poor spatial resolution. In addition to these dual streams, the occipital-callosal fibers of the corpus callosum allow information to be communicated to the bilateral occipital lobes for visual processing. ^{1 ,​ 2 ,​ 3}

16.3 Preoperative Imaging

Preoperative imaging can help identify and delineate components of the optic pathway. ^{10 ,​ 11 ,​ 12 ,​ 13 ,​ 14} The optic nerves are typically best visualized with high-resolution T2-weighted magnetic resonance imaging (MRI). ^{10 ,​ 11 ,​ 12 ,​ 13 ,​ 14} The differential in intensity between the optic nerves and the cerebrospinal fluid space allows the delineation of the optic nerves from the back of the globe, through the optic canal and to the chiasm. ^{10 ,​ 11} In the setting of lesions involving or adjacent to the optic nerves, sequences such as inversion recovery (fast gray and white matter acquisition T1 inversion recovery), contrast-enhanced fast imaging employing steady-state acquisition (FIESTA), and heavily weighted T2 sequences can help better delineate the optic nerve and chiasm especially nearby bony structures. ^{12 ,​ 13 ,​ 14}

The retrochiasmatic optic pathway involving the optic radiations can be visualized with a number of different MRI modalities. ^{10 ,​ 11} The most commonly used method is diffusion tensor imaging (DTI; Fig. 16‑1). ^{10 ,​ 11} This method is based on the diffusion of water molecules in the extra and intracellular spaces. ^{10 ,​ 11} When this diffusion is not random and limited to directional structures such as axons, the diffusion of water follows these axon bundles rather than random path. ^{10 ,​ 11} This anisotropic diffusion allows the identification of these white matter tacts. ^{10 ,​ 11} In addition to DTI, functional MRI (fMRI) can also help identify components of the optic pathway, namely the primary visual cortex. ^{10 ,​ 11} FMRI is based on neurovascular coupling, where neural activity in a specific cortical region will trigger a change in the regional blood flow that is captured on the MRI. ^{10 ,​ 11} A component of this fMRI is called retinotopic mapping. ^{10 ,​ 11} In retinotopic mapping, visual stimuli are presented to different visual fields that create a wave of specific neural activity within the cortex. ^{10 ,​ 11} This allows one to identify the correspondence between the position of the stimuli within the visual field and the cortex to be identified. ^{10 ,​ 11} More complex functional mapping involves hierarchical functions such as object perception and recognition of object motion, where cortical areas that are activated during specific tasks are identified. ^{10 ,​ 11} For object perception, an object is shown in its entire form or scrambled, while, for motion, an object is shown in motion or stationary. ^{10 ,​ 11} The corresponding cortical areas that are activated are then identified for their respective function. ^{10 ,​ 11}

These preoperative imaging modalities have significant limitations for intraoperative utilization. ^{10 ,​ 11 ,​ 12 ,​ 13 ,​ 14} This is especially true for the optic radiations. ^{10 ,​ 11} For DTI, the normal anisotropic diffusion of water can be disrupted by various pathologies. ^{10 ,​ 11} Lesions that induce perilesional edema interfere with the normal anisotropic diffusion of water thereby making this imaging modality less precise and subject to false positive and false negative identification intraoperatively. ^{1 ,​ 2 ,​ 4} Moreover, pathologies that destroy and/or infiltrate these white matter tracts can also interfere with the anisotropic diffusion of water within these tracts. ^{1 ,​ 2 ,​ 4} In addition to the limitations associated with DTI, fMRI can be associated with error as this imaging modality requires large-scale neural activation with specific tasks. ^{10 ,​ 11} These foci of neural activation can be associated with false positive and false negative localization with intraoperative direct electrical stimulation. ^{1 ,​ 2 ,​ 4} Despite the widespread availability of these imaging modalities and others, the problem remains that they are associated with false positive and false negative identification of eloquent and noneloquent areas. ^{1 ,​ 2 ,​ 4} Most importantly, they do not accurately identify functional from nonfunctional areas and therefore should not be used solely for functional brain area identification and avoidance during surgery. ^{1 ,​ 2 ,​ 4} This is why many advocate a 5 to 10 mm avoidance of these critical cortical and subcortical regions identified on DTI and fMRI during surgery, which can significantly limit extent of resection. ^{1 ,​ 2 ,​ 4}