The so-called computational neuroanatomy is an established methodology to characterize the neuroanatomical configuration of different brains, encompassing mainly the following techniques: (1) voxel-based morphometry (VBM) which compares neuroanatomical differences on a voxel by voxel basis; (2) deformation-based morphometry (DBM) which provides information about global differences in brain shape; (3) and tensor-based morphometry (TBM) which provides information about local shape differences (cf. Ashburner et al. 2003 for review). These 3-D MRI analysis techniques which make use of local registration algorithms, however, are only validated for analyses at group level: it could be demonstrated that VBM could not satisfactorily delineate the spatial extent of focal lesions in individuals, even in simulations that avoided preprocessing artifacts (Mehta et al. 2003). To be applicable in the clinical practice and, in particular, in the presurgical setting, the techniques have to be sensitive to quantitative or qualitative alterations of normal brain structures in a given individual patient. In this respect, it is not useful to apply this computerized approach in lesions which are clearly assessable by conventional high-resolution MRI such as in neoplasms. Nevertheless, there are applications of computerized analysis of structural 3-D MRI for presurgical diagnostics, e.g., in the field of epilepsy surgery and here, e.g., in malformations of cortical development (MCD). First, there have been voxel-based 3-D MRI analysis tools developed for an automated observer-independent MCD lesion detection (Huppertz et al. 2005). These algorithms serve mainly as image screening tools in comparison to a healthy control sample in order to facilitate MCD detection and localization. Second, as a consequent continuation of the multimodality approach in the sense of a combined voxel-based mapping (Kassubek et al. 2000), a combination of ligand imaging using PET (which is able to provide relevant information on MCD; Cross 2003) and SPM-based MRI analysis has been proposed for the investigation of MCD (Richardson et al. 1997). Although at the expense of the measurement of absolute quantitative values (the voxel-based comparison of MRI and PET enables an automated objective whole-brain analysis), issues concerning partial volume effects and mixed tissue sampling in ligand images have to be resolved, and disproportionate changes in function and structure to be identified. In a study using the approach of voxel-based MRI analysis and subtraction of ictal SPECT co-registered to MRI (SISCOM), it could be demonstrated that multimodal assessment of MCD may contribute to an advanced observer-independent preoperative assessment of the seizure origin and might thus improve presurgical diagnostics in symptomatic epilepsy (Wiest et al. 2005). Wehner and Lüders (2008) summarized the role of neuroimaging in presurgical mapping focusing on structural MRI, PET, and SPECT.

Diffusion is a physical process involving the translational movement of molecules via thermally driven random motions, the so-called Brownian motion. The underlying microstructure of the cellular tissue influences the overall mobility of the diffusing molecules by providing numerous barriers and by creating various individual compartments within the tissue (e.g., intra-/extracellular space, neurons, glial cells, axons). Signal intensity of diffusion measured via diffusion-weighted imaging shows a greater signal enhancement and more rapid diffusion of water molecules along the direction parallel to the axis of the fiber bundles (so-called anisotropy) than in the perpendicular direction where signal intensity and diffusion velocity is lowest (so-called isotropy). As diffusion is characterized by hindrances of equal molecular movement caused by oriented barriers, it requires a general tensor which fully describes the mobility of molecules along each direction and the correlation between these directions. A common way to summarize diffusion measurements with DTI is the calculation of a parameter that reflects the overall anisotropy, such as the fractional anisotropy (FA) as a measure of the proportion of the diffusion tensor due to anisotropy. The FA can either be presented as a gray-scale image or the principal diffusion direction of the brain structure to be examined can be encoded in color resulting in color-coded parameter maps (Basser and Jones 2002; Le Bihan and van Zijl 2002).

Concretely, DTI measures the mean longitudinal direction of axons in white matter tracts. For this purpose, the application of fiber-tracking algorithms allows a delineation of the major white matter pathways (Mori and van Zijl 2002). With respect to presurgical mapping, the DTI technique provides information about the deformation, the displacement, or even the interruption of the white matter tracts around a lesion (Sinha et al. (2002). For instance, DTI is used to differentiate normal white matter from edematous brain and contrast-enhancing tumor margins in patients with high-grade gliomas. Witwer et al. (2002) reported that the ¬ isualization of white matter tracts by DTI was beneficial in the surgical planning for patients with intrinsic brain tumors showing that anatomically intact fibers may be present in abnormal-appearing areas of the brain. The knowledge of the course of major white matter tracts in relationship to the exact location of the lesion is of utmost importance to the neurosurgeon for the planning of the most appropriate surgical approach and may thus help to prevent the risk of new postoperative neurological deficits (Clark et al. 2003). For topographical identification, DTI atlas systems have been developed (Oishi et al. 2011). Figure 2 shows an example of a patient with a glioma showing DTI-based structural connectivity changes adjacent to the left cortex.

The role of DTI for the assessment of important pathways in the vicinity of tumor borders has already been described in 2004 by Ulmer and coworkers (2004) who concluded that DTI can significantly enhance the estimation of the proximity of functional brain structures/areas to brain tumor borders suggesting DTI to be a valuable tool for presurgical planning. This work was followed by a pre- and postoperative DTI-based analysis (white matter tractography) of brain tumors (Lazar et al. 2006) showing the preservation of white matter tracts during surgery. Ferda and coworkers (2010) reported that DTI offered significant additional information in 24 patients with glial tumors which might help for the differentiation of tumors with infiltrative growth character from more circumscribed tumors, and it has been possible to advance the accuracy of estimation of tumor grading prior to final determination of the morphological diagnosis. The finding of a reduced FA in the white matter around a tumor increases the probability of peritumorous white matter infiltration by tumor invasion. Figure 3 shows an example of the application of a DTI analysis in a patient with a biopsy-proven anaplastic oligoastrocytoma WHO III° by comparing the FA maps acquired before combined radio-/chemotherapy with those obtained 2, 5, and 9 months after radio-/chemotherapy (for the methodological approach, please cf. Müller and Kassubek 2013). After surgery, the patient received combined radio-/chemotherapy and 6 cycles of adjuvant chemotherapy.

Consecutively, DTI measurements are increasingly used in advanced neuronavigation protocols (Rohde et al. 2005). That way, the course of the pyramidal tract can be displayed with DTI for the intraoperative visualization in functional neuronavigation systems (Nimsky et al. 2005b). Moreover, DTI is helpful for the visualization of the optic radiation to preserve the patients’ visual field during the surgery of lesions adjacent to the course of the optic radiation. For instance, Coenen et al. (2005) could reproduce the course of the optic radiation by using neuronavigation with optic radiation DTI and preoperative and postoperative neuro-ophthalmological testing including perimetry. Also in epilepsy surgery, visual field defects are a well-recognized complication of anterior temporal lobe resection and may occur secondary to disruption of the Meyer loop. The use of DTI tractography for the visualization of the optic radiation before and after surgery has been concluded to be useful for the prediction and perhaps prevention of visual field defects following temporal lobe resection (Powell et al. 2005).

However, there are also limitations for DTI which is not able to correctly identify fiber crossings due to its inability to resolve more than one single axon direction within each voxel. Advanced approaches, such as the so-called q-ball imaging technique, can resolve intra-voxel white matter fiber crossings as well as white matter insertions into the cortex and might hence be able to uncover the complex intra-voxel fiber architecture and thus eliminate a key obstacle for the noninvasively mapping of neural connectivity in the human brain (Tuch et al. 2003). An approach which addresses the limits of diffusion resolution of “crossing” or “kissing” fibers and in the vicinity of tumor or edema is tractography method based on high-angular-resolution diffusion imaging (HARDI). With compressed sensing techniques, HARDI acquisitions with a smaller number of directional measurements can be used, thus enabling the use of HARDI-based fiber tractography in the clinical practice (Kuhnt et al. 2013; Bucci et al. 2013). More recent investigations applied Q-ball imaging as high-definition fiber tractography (HDFT) to patients with intracerebral lesions (Fernandez-Miranda et al. 2012) in order to understand the lesional patterns of structural injury and facilitate neurosurgical interventions. A further application of HFTD was reported in a patient with traumatic brain injury (Shin et al. 2012). An additional technical challenge, especially in presurgical mapping, i.e., the intraoperative displacement of major white matter tracts during mass resection due to brain shift, has been addressed by a comparison of preoperative with intraoperative DTI-based fiber tracking (Nimsky et al. 2005c). The authors described a marked shift and deformation of major white matter tracts due to tumor removal, herewith emphasizing the need for an intraoperative update – in navigation systems – during the resection of deep-seated tumor portions near eloquent brain areas. Consequently fiber tracking is not only useful for preoperative neurosurgical visualization but also for intraoperative planning and updating. Another emerging challenge for the clinical routine use is the lack of standardization which affects DTI as well as fMRI. There are many different software packages available for tractography, including traditional DTI-based methods (such as the fiber assignment by a continuous tracking algorithm), as well as more advanced methods that take into account each eigenvector within an individual voxel (Pillai 2010).

The integration of DTI into presurgical mapping protocols is also useful with respect to MRI multimodality, since the bi-parametric combination of DTI with fMRI provides valuable information that cannot be obtained by the use of either technique alone. The use of fMRI and DTI in a multimodality approach enables the connection of the advantages of both techniques. While fMRI is able to highlight the localization of function within the cortex, DTI visualizes white matter structures in vivo (Dimou et al. 2013; Kleiser et al. 2010; Wu et al. 2007). The validity and sensitivity of noninvasive functional mapping for surgical guidance can be improved by considering the results obtained with both methods, and the use of 3-D visualization seems crucial and unique for the representation and understanding of the complicated spatial relationship among the lesion, gray matter activation, and white matter fiber bundles (Hendler et al. 2003). In neoplastic brain lesions, the combined use of fMRI with DTI can provide a better estimation of the proximity of the tumor borders to eloquent brain systems, such as language, speech, vision, motor, and premotor functions (Ulmer et al. 2004). DTI and fMRI can thus be used as complementary techniques for the delineation of eloquent subcortical and cortical areas (Fig. 4).

The use of fMRI together with or without DTI currently seems to be a promising application of the available neuroimaging techniques in the clinical context. These multiparametric MRI data can be implemented in computer systems for functional neuronavigation or radiation treatment (Stippich 2010). There has been significant interest in using this bi-parametric MRI approach for the localization of potential brain regions which may be at risk during neurosurgical procedures. There is an increasing number of groups worldwide that have integrated both modalities in stereotaxic guidance systems, providing clinical insight in the mapping of eloquent gray matter and white matter regions to the neurosurgeon (Dimou et al. 2013). It has been reported that as the distance between tumor and eloquent cortex (“lesion to eloquent area distance” or “LEAD”) decreases, so does the likelihood of a complete surgical resection. Consequently, DTI was proposed as a tool for the identification of those patients who would benefit most from total surgical resection and, thus, having longer survival time (Castellano et al. 2011). Used in conjunction, both techniques provide important information both prior to surgery for a thorough preoperative planning, as well as during surgery for the intraoperative update of the information via integration into frameless stereotactic neuronavigational systems.

The use of fMRI landmarks for ROI selection has been shown to enhance the fiber-tracking performance compared to a strictly anatomically based ROI selection and is usually considered reliable in controls. Nevertheless, this approach may be inaccurate in patients with brain lesions (Schonberg et al. 2006). Kleiser and coworkers (2010) found that DTI-related results were more accurate than results using anatomical landmarks alone. Four patterns of regional anisotropy have been observed in relation to WM tracts in the vicinity of tumors (Jellison et al. 2004), i.e., (1) normal signal with altered position or direction, observed in cases of tract displacement; (2) decreased but present signal with normal direction and location, supposedly corresponding to vasogenic edema; (3) decreased signal with disrupted maps, believed to be related to tumor infiltration; and (4) loss of anisotropic signal corresponding to fiber tract destruction. In two trials addressing a potential modification of the surgical approach to corticotomy, the presurgical use of combined DTI/fMRI led to a change in the surgical approach in 16–21 % of cases (Dimou et al. 2013). Together, these techniques show great promise for improved neurosurgical outcomes. Clinical reports have suggested that the additional information provided by the combination of the two imaging modalities, via a co-registered map delivered to the neurosurgeon on a heads-up display, is valuable and facilitates a safe resection (Berntsen et al. 2010). While further research is required for a more widespread clinical validity and acceptance, results from the literature (e.g., Schonberg et al. 2006; Bello et al. 2008; Kleiser et al. 2010; Kuhnt et al. 2012, 2013) are very promising to cement these techniques into the clinical setup in the very new future. For instance, more recent works have applied DTI-based fiber tracking in intraoperative glioma surgery (Kuhnt et al. 2012).

In vivo proton (¹H)-MR spectroscopy (MRS) visualizes signals from carbon-bound, non-exchangeable protons, showing the highest information density in the spectral region from 1 to 5 ppm (Trabesinger et al. 2003). Principal metabolites detected by MRS include the neuronal marker N-acetylaspartate as an indirect expression of the integrity of neurons, choline-containing compounds (such as metabolites involved in phospholipid membrane synthesis) as markers for glial activity, creatine (including phosphocreatine) as a marker for energy metabolism, and lactate as an indicator for anaerobic glycolysis detected under pathologic conditions. For data acquisition, there are single-voxel and multiple-voxel techniques, depending on the information required. If applied in a larger area of expected signal alteration, the latter would be preferable as a screening for abnormalities in different locations within a bigger region (Barker 2005).

In the presurgical diagnostics of epilepsy surgery, MRS has shown to have additional potential to be used in imaging protocols for the presurgical evaluation of non-lesional frontal lobe epilepsy (Guye et al. 2005) and for the assessment of the surgical outcome in patients with temporal lobe epilepsy (Antel et al. 2002). In this context, the combination of different approaches such as MRS and 3-D MRI-based measurements of regions of interests (ROIs) such as the amygdala-hippocampal volume has been found to provide complementary results in presurgical neuroimaging protocols (Cendes et al. 1997). MRS has been established as part of presurgical multimodality neuroimaging studies using various MRI-based techniques and PET in patients with non-lesional unilateral temporal lobe epilepsy (Knowlton et al. 1997) and also in the in vivo analysis of metabolic disturbances associated with partial epilepsy (Knowlton 2004).

MRS may be used in the diagnostics of brain tumors for the identification of regions of viable tumor, investigation of the tumor extent, and monitoring of the effectiveness of the applied therapy (Vigneron et al. 2001). Consecutively, MRS has now been integrated into presurgical mapping protocols for glioma surgery to indicate the degree of tumor invasion by the co-registration of the metabolic maps derived from MRS postprocessing onto the anatomical 3-D MRI data set for its use in neuronavigation/frameless stereotaxy systems (Ganslandt et al. 2005). In this study, the stereotactic biopsies were acquired by the use of the spectral data, and the tumor areas as defined by the metabolic maps and as histopathologically confirmed by biopsy exceeded the T2-weighted signal change on conventional MRI in all seven cases investigated. The authors concluded that MRS may be useful for a more accurate definition of the tumor infiltration zone in glioma surgery than conventional anatomical MRI alone. Furthermore, a combination of MRS with functional landmarks obtained from fMRI is possible to increase the sensitivity for the assessment of tumor borders and the localization of functionally eloquent areas. Furthermore, intraoperative high-field MRI with integrated microscope-based neuronavigation enables the acquisition of MRS also during the surgical procedure for immediate intraoperative update and control (Nimsky et al. 2005a).

Recently, the application of MRS in a multimodal setup with MRI volumetry, EEG, PET, and SPECT has been reported in epilepsy (Cendes 2013) as well as in glioma surgery (Shang et al. 2012).

Magnetoencephalography (MEG) is an analysis technique which uses electromagnetic signals to localize the source of biomagnetic fields derived from intracellular electric current flows in the brain, thus providing direct information about neural activity (Hamalainen et al. 1993). One of the most important advantages of MEG is that its temporal resolution is very high (i.e., in the order of 1 ms) and is mainly restricted only by the sampling frequency of the recording system (Lounasmaa et al. 1996). Since magnetic fields are not distorted by conduction inhomogeneities in the head, MEG is a potentially powerful tool for the localization of brain function, providing a spatial discrimination below 5 mm. As a major limitation, MEG has a fundamental source identification problem (inverse problem), by determining the active site within the brain from the magnetic field pattern recorded outside the skull. Thus, some a priori assumptions must be made about the source. Usually, signal generators in the brain are described as current dipoles which are physiologically reasonable models for relatively small active cortical areas. MEG results can be directly co-registered to 3-D MRI. The merging of the data obtained from MEG with anatomical MRI images, i.e., magnetic source imaging (MSI), allows the integration of the functional information as obtained from MEG into the structural information of MR (Orrison 1999).