Subject-specific atlas for automatic brain tissue segmentation of neonatal magnetic resonance images.

Developing advanced systems for 3D brain tissue segmentation from neonatal magnetic resonance (MR) images is vital for newborn structural analysis. However, automatic segmentation of neonatal brain tissues is challenging due to smaller head size and inverted T1/T2 tissue contrast compared to adults. In this work, a subject-specific atlas based technique is presented for segmentation of gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) from neonatal MR images. It involves atlas selection, subject-specific atlas creation using random forest (RF) classifier, and brain tissue segmentation using the expectation maximization-Markov random field (EM-MRF) method. To increase the segmentation accuracy, different tissue intensity- and gradient-based features were used. Evaluation on 40 neonatal MR images (gestational age of 37-44 weeks) demonstrated an overall accuracy of 94.3% and an average Dice similarity coefficient (DSC) of 0.945 (GM), 0.947 (WM), and 0.912 (CSF). Compared to multi-atlas segmentation methods like SEGMA and EM-MRF with multiple atlases, our method improved accuracy by up to 4%, particularly in complex tissue regions. Our proposed method allows accurate brain tissue segmentation, a crucial step in brain magnetic resonance imaging (MRI) applications including brain surface reconstruction and realistic head model creation in neonates.

Transcranial magnetic stimulation and magnetoencephalography are feasible alternatives to invasive methods in optimizing responsive neurostimulation device placement.

Responsive neurostimulation (RNS) is a treatment option for patients with refractory epilepsy when surgical resection is not possible due to overlap of the irritative zone and eloquent cortex. Presurgical evaluations for RNS placement typically rely on invasive methods. This study investigated the potential of transcranial magnetic stimulation (TMS) and magnetoencephalography (MEG) to provide key presurgical information non-invasively. We hypothesized that these non-invasive methods may assist in optimizing RNS placement by providing useful information for seizure localization by MEG and eloquent cortex mapping by TMS. A retrospective chart review identified nine patients who underwent RNS placement (mean age = 20.4 years [SD = 5.6], two-thirds were female). Characterization of the irritative zone using MEG was successful in eight of nine patients. Non-invasive mapping of relevant eloquent cortex was attempted in all patients. TMS was successful in eight of nine patients, and MEG was successful in two of six patients. Importantly, patients mapped with non-invasive modalities experienced an average seizure reduction of 77 % at their most recent clinic visit, compared to 75 % seizure reduction in those with invasive evaluations, indicating appropriate RNS placement. These data demonstrate that TMS and MEG can provide key information for RNS and may be feasible alternatives to invasive methods for assisting in decision making regarding RNS placement. Non-invasive methods for determining RNS placement have a high rate of success when data from multiple non-invasive modalities converge and can inform more accurate placement of intracranial electrodes prior to RNS placement or mitigate their need.

Concordance between Wada, Transcranial Magnetic Stimulation, and Magnetoencephalography for Determining Hemispheric Dominance for Language: A Retrospective Study.

Determination of language hemispheric dominance (HD) in patients undergoing evaluation for epilepsy surgery has traditionally relied on the sodium amobarbital (Wada) test. The emergence of non-invasive methods for determining language laterality has increasingly shown to be a viable alternative. In this study, we assessed the efficacy of transcranial magnetic stimulation (TMS) and magnetoencephalography (MEG), compared to the Wada test, in determining language HD in a sample of 12 patients. TMS-induced speech errors were classified as speech arrest, semantic, or performance errors, and the HD was based on the total number of errors in each hemisphere with equal weighting of all errors (classic) and with a higher weighting of speech arrests and semantic errors (weighted). Using MEG, HD for language was based on the spatial extent of long-latency activity sources localized to receptive language regions. Based on the classic and weighted language laterality index (LI) in 12 patients, TMS was concordant with the Wada in 58.33% and 66.67% of patients, respectively. In eight patients, MEG language mapping was deemed conclusive, with a concordance rate of 75% with the Wada test. Our results indicate that TMS and MEG have moderate and strong agreement, respectively, with the Wada test, suggesting they could be used as non-invasive substitutes.

Comparing the efficacy of awake and sedated MEG to TMS in mapping hand sensorimotor cortex in a clinical cohort.

Non-invasive methods such as Transcranial Magnetic Stimulation (TMS) and magnetoencephalography (MEG) aid in the pre-surgical evaluation of patients with epilepsy or brain tumor to identify sensorimotor cortices. MEG requires sedation in children or patients with developmental delay. However, TMS can be applied to awake patients of all ages with any cognitive abilities. In this study, we compared the efficacy of TMS with MEG (in awake and sedated states) in identifying the hand sensorimotor areas in patients with epilepsy or brain tumors. We identified 153 patients who underwent awake- (n = 98) or sedated-MEG (n = 55), along with awake TMS for hand sensorimotor mapping as part of their pre-surgical evaluation. TMS involved stimulating the precentral gyrus and recording electromyography responses, while MEG identified the somatosensory cortex during median nerve stimulation. Awake-MEG had a success rate of 92.35 % and TMS had 99.49 % (p-value = 0.5517). However, in the sedated-MEG cohort, TMS success rate of 95.61 % was significantly higher compared to MEG's 58.77 % (p-value = 0.0001). Factors affecting mapping success were analyzed. Logistic regression across the entire cohort identified patient sedation as the lone significant predictor, contrary to age, lesion, metal, and number of antiseizure medications (ASMs). A subsequent analysis replaced sedation with anesthetic drug dosage, revealing no significant predictors impacting somatosensory mapping success under sedation. This study yields insights into the utility of TMS and MEG in mapping hand sensorimotor cortices and underscores the importance of considering factors that influence eloquent cortex mapping limitations during sedation.