Automatic Detection of Scalp High-Frequency Oscillations Based on Deep Learning.

Scalp high-frequency oscillations (sHFOs) are a promising non-invasive biomarker of epilepsy. However, the visual marking of sHFOs is a time-consuming and subjective process, existing automatic detectors based on single-dimensional analysis have difficulty with accurately eliminating artifacts and thus do not provide sufficient reliability to meet clinical needs. Therefore, we propose a high-performance sHFOs detector based on a deep learning algorithm. An initial detection module was designed to extract candidate high-frequency oscillations. Then, one-dimensional (1D) and two-dimensional (2D) deep learning models were designed, respectively. Finally, the weighted voting method is used to combine the outputs of the two model. In experiments, the precision, recall, specificity and F1-score were 83.44%, 83.60%, 96.61% and 83.42%, respectively, on average and the kappa coefficient was 80.02%. In addition, the proposed detector showed a stable performance on multi-centre datasets. Our sHFOs detector demonstrated high robustness and generalisation ability, which indicates its potential applicability as a clinical assistance tool. The proposed sHFOs detector achieves an accurate and robust method via deep learning algorithm.

Efficacy and safety of perampanel as the first add-on therapy for children with epilepsy: A real-world multicenter prospective observational study.

OBJECTIVE: Perampanel (PER) is a new anti-seizure medication (ASM) with a novel mechanism of action. This study aimed to determine the efficacy and safety of PER when added to monotherapy in children and adolescents (age, 4-18 years) with epilepsy. METHOD: A multicenter prospective observational study was performed on children and adolescents (age, 4-18 years) with epilepsy who did not respond to ASM monotherapy between July 2021 and October 2022. PER was used as the first add-on therapy for the enrolled patients. Seizure-free rate, response rate, inefficacy rate, and drug retention rate were the main observation indicators during the 6 months of treatment. The patients were grouped based on treatment efficacy, and factors affecting efficacy were statistically analyzed. Adverse reactions were also recorded. RESULTS: In this study, 93 patients with epilepsy were enrolled; among them, 9 patients were lost to follow-up (attrition rate, 9.7 %), and 84 were included in the analysis. Five patients with unknown efficacy discontinued taking PER early due to intolerable adverse reactions, and 79 patients (48 males, 31 females; mean age, 11.0 ± 3.9 years) finally remained. Genetic epilepsy and structural epilepsy were found in 22 patients and 36 patients, respectively. The mean duration of epilepsy history at the time of PER initiation was 4.0 ± 3.8 years, and the mean maintenance dosage of add-on PER was 4.5 ± 1.8 mg/day (equivalent to 0.14 ± 0.07 mg/kg/day). Among the 79 patients, 28 patients were diagnosed with epilepsy syndrome, including 13 patients having self-limited epilepsy with centrotemporal spikes, among whom 9 patients were seizure-free after adding PER during the 6-month follow-up (seizure-free rate, 69.2 %). For these 79 patients, the seizure-free, response, and retention rates at the end of follow-up were 45.6 %, 74.7 %, and 82.1 %, respectively. Among the 84 patients included in the analyses, adverse reactions occurred in 20 patients, mainly dizziness (8 patients), somnolence (6 patients), and irritability (4 patients), and 4 patients developed two adverse reactions simultaneously. Univariate analyses revealed statistically significant differences in efficacy between groups with structural and non-structural epilepsy and between groups with different baseline concomitant ASMs, suggesting that these factors affected the efficacy of PER as the first add-on therapy. CONCLUSION: The overall response rate of PER as the first add-on therapy for children and adolescents with epilepsy who were followed up for 6 months was 74.7 %, indicating a relatively favorable safety and tolerability profile. The group of the baseline concomitant ASM administered and the etiological classification of epilepsy as either structural or non-structural were the factors influencing the efficacy of PER as the first add-on therapy.

Association of FAT1 with focal epilepsy and correlation between seizure relapse and gene expression stage.

PURPOSE: The FAT1 gene encodes FAT atypical cadherin 1, which is essential for foetal development, including brain development. This study aimed to investigate the relationship between FAT1 variants and epilepsy. METHODS: Trio-based whole-exome sequencing was performed on a cohort of 313 patients with epilepsy. Additional cases with FAT1 variants were collected from the China Epilepsy Gene V.1.0 Matching Platform. RESULTS: Four pairs of compound heterozygous missense FAT1 variants were identified in four unrelated patients with partial (focal) epilepsy and/or febrile seizures, but without intellectual disability/developmental abnormalities. These variants presented no/very low frequencies in the gnomAD database, and the aggregate frequencies in this cohort were significantly higher than those in controls. Two additional compound heterozygous missense variants were identified in two unrelated cases using the gene-matching platform. All patients experienced infrequent (yearly/monthly) complex partial seizures or secondary generalised tonic-clonic seizures. They responded well toantiseizure medication, but seizures relapsed in three cases when antiseizure medication were decreased or withdrawn after being seizure-free for three to six years, which correlated with the expression stage of FAT1. Genotype-phenotype analysis showed that epilepsy-associated FAT1 variants were missense, whereas non-epilepsy-associated variants were mainly truncated. The relationship between FAT1 and epilepsy was evaluated to be "Strong" by the Clinical Validity Framework of ClinGen. CONCLUSIONS: FAT1 is a potential causative gene of partial epilepsy and febrile seizures. Gene expression stage was suggested to be one of the considerations in determining the duration ofantiseizure medication. Genotype-phenotype correlation helps to explain the mechanisms underlying phenotypic variation.

Predicting Antiseizure Medication Treatment in Children with Rare Tuberous Sclerosis Complex-Related Epilepsy Using Deep Learning.

BACKGROUND AND PURPOSE: Tuberous sclerosis complex disease is a rare, multisystem genetic disease, but appropriate drug treatment allows many pediatric patients to have positive outcomes. The purpose of this study was to predict the effectiveness of antiseizure medication treatment in children with tuberous sclerosis complex-related epilepsy. MATERIALS AND METHODS: We conducted a retrospective study involving 300 children with tuberous sclerosis complex-related epilepsy. The study included the analysis of clinical data and T2WI and FLAIR images. The clinical data consisted of sex, age of onset, age at imaging, infantile spasms, and antiseizure medication numbers. To forecast antiseizure medication treatment, we developed a multitechnique deep learning method called WAE-Net. This method used multicontrast MR imaging and clinical data. The T2WI and FLAIR images were combined as FLAIR3 to enhance the contrast between tuberous sclerosis complex lesions and normal brain tissues. We trained a clinical data-based model using a fully connected network with the above-mentioned variables. After that, a weighted-average ensemble network built from the ResNet3D architecture was created as the final model. RESULTS: The experiments had shown that age of onset, age at imaging, infantile spasms, and antiseizure medication numbers were significantly different between the 2 drug-treatment outcomes (P < .05). The hybrid technique of FLAIR3 could accurately localize tuberous sclerosis complex lesions, and the proposed method achieved the best performance (area under the curve = 0.908 and accuracy of 0.847) in the testing cohort among the compared methods. CONCLUSIONS: The proposed method could predict antiseizure medication treatment of children with rare tuberous sclerosis complex-related epilepsy and could be a strong baseline for future studies.

A juvenile mouse model of anti-N-methyl-D-aspartate receptor encephalitis by active immunization.

Introduction: Anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis is a common autoimmune encephalitis, and it is associated with psychosis, dyskinesia, and seizures. Anti-NMDAR encephalitis (NMDARE) in juveniles and adults presents different clinical charactreistics. However, the pathogenesis of juvenile anti-NMDAR encephalitis remains unclear, partly because of a lack of suitable animal models. Methods: We developed a model of juvenile anti-NMDAR encephalitis using active immunization with an amino terminal domain peptide from the GluN1 subunit (GluN1356 - 385) against NMDARs in 3-week-old female C57BL/6J mice. Results: Immunofluorescence staining suggested that autoantibody levels in the hippocampus increased, and HEK-293T cells staining identified the target of the autoantibodies as GluN1, suggesting that GluN1-specific immunoglobulin G was successfully induced. Behavior assessment showed that the mice suffered significant cognition impairment and sociability reduction, which is similar to what is observed in patients affected by anti-NMDAR encephalitis. The mice also exhibited impaired long-term potentiation in hippocampal CA1. Pilocarpine-induced epilepsy was more severe and had a longer duration, while no spontaneous seizures were observed. Conclusion: The juvenile mouse model for anti-NMDAR encephalitis is of great importance to investigate the pathological mechanism and therapeutic strategies for the disease, and could accelerate the study of autoimmune encephalitis.

Activation Map Reveals Language Impairment in Children with Benign Epilepsy with Centrotemporal Spikes (BECTS).

Objective: Neuropsychological evidence revealed language impairment in children with benign epilepsy with centrotemporal spikes (BECTS). This study investigates language function using task-activated fMRI. Methods: We conducted a language task fMRI study on three groups on a 3.0T MRI scanner, including a new onset drug naïve group (NODN-BECTS, n=11, age=9.6±1.6), an established epilepsy with medication-treated group (Med-BECTS, n=17, age=10.7±2.2) and a healthy control group (HC, n=18, age=10.8±1.7). We use MATLAB14 and SPM12 to pre-process and analyze the data. A one-sample t-test was used to identify task-related brain activation changes in each group, based on the general linear model (GLM). And, then two sample t-test was performed to compare different activated regions between groups. In addition, scores on the most recent Mandarin school exams were acquired to examine and contrast extra-scanner language performance. Results: Statistical results show that some language-related brain regions (such as the left superior frontal gyrus and cerebellar vermis) were additionally activated in the NODN-BECTS group compared with the HC group. Compared with NODN-BECTS and HC groups, decreased activations were found in language-related regions in the Med-BECTS group, including the left insula, superior and middle frontal gyri, and bilateral middle occipital gyri. On the Mandarin school exams, the average score for HC was 87.3±8.2, NODN was 84.8±7.8, and Med was 78.2±13.2. There was a trend toward statistical significance between the Med and the HC (p = 0.074) as well as NODN (p = 0.092) groups. No statistically significant differences were found between the HC and the NODN-BECTS groups. Significance: Language task fMRI reveals additional areas of activation in new onset BECTS compared to healthy controls which may be compensatory in nature. Antiseizure medications (ASMs) and/or longer duration of BECTS additionally appears to affect language-related regions and reduce their functional ability.

COVID-19 vaccination for patients with epilepsy: A Chinese expert consensus.

Coronavirus disease-2019 (COVID-19) first emerged in late 2019 and has since spread worldwide. More than 600 million people have been diagnosed with COVID-19, and over 6 million have died. Vaccination against COVID-19 is one of the best ways to protect humans. Epilepsy is a common disease, and there are approximately 10 million patients with epilepsy (PWE) in China. However, China has listed "uncontrolled epilepsy" as a contraindication for COVID-19 vaccination, which makes many PWE reluctant to get COVID-19 vaccination, greatly affecting the health of these patients in the COVID-19 epidemic. However, recent clinical practice has shown that although a small percentage of PWE may experience an increased frequency of seizures after COVID-19 vaccination, the benefits of COVID-19 vaccination for PWE far outweigh the risks, suggesting that COVID-19 vaccination is safe and recommended for PWE. Nonetheless, vaccination strategies vary for different PWE, and this consensus provides specific recommendations for PWE to be vaccinated against COVID-19.

Effects of ketogenic diet on the classification and functional composition of intestinal flora in children with mitochondrial epilepsy.

The ketogenic diet (KD) has shown excellent performance in the treatment of refractory epilepsy, but how it works is not yet fully understood. Gut microbiota is associated with various neurological disorders through the brain-gut axis. Different dietary patterns have different effects on the composition and function of gut microbiota. Here, by analyzing fecal samples from some patients with mitochondrial epilepsy before and after KD treatment through 16SrRNA sequencing, we found that KD intervention reduced the abundance of Firmicutes in the patient's gut, while the abundance of Bacteroidota increased in the KD group. LefSe analysis showed that Actinobacteriota, Phascolarctobacterium had significant advantages in the control group, while Bacteroides increased significantly after KD intervention, especially Bacteroides fragilis. Functional analysis showed that there were significant differences in 12 pathways in level 3. These changes suggest that KD can change the composition and diversity of the gut microbiota in patients and affect their function. Changes in specific bacterial groups in the gut may serve as biomarkers for the therapeutic effects of KD on epilepsy.

Both epilepsy and anti-seizure medications affect bone metabolism in children with self-limited epilepsy with centrotemporal spikes.

OBJECTIVE: Bone metabolism can be influenced by a range of factors. We selected children with self-limited epilepsy with centrotemporal spikes (SeLECTS) and lifestyles similar to those of healthy children to control for the confounding factors that may influence bone metabolism. We aimed to identify the specific effects of epilepsy and/or anti-seizure medications (ASMs) on bone metabolism. METHODS: Patients with SeLECTS were divided into an untreated group and a monotherapy group, and the third group was a healthy control group. We determined the levels of various biochemical markers of bone metabolism, including procollagen type I nitrogenous propeptide (PINP), alkaline phosphatase (ALP), osteocalcin (OC), collagen type I cross-linked C-telopeptide (CTX), calcium, magnesium, phosphorus, parathyroid hormone (PTH), and vitamin D3 (VD3 ). RESULTS: A total of 1487 patients (from 19 centers) were diagnosed with SeLECTS; 1032 were analyzed, including 117 patients who did not receive any ASMs (untreated group), 643 patients who received only one ASM (monotherapy group), and 272 children in the healthy control group. Except for VD3 , other bone metabolism of the three groups were different (p < .001). Bone metabolism was significantly lower in the untreated group than the healthy control group (p < .05). There were significant differences between the monotherapy and healthy control group in the level of many markers. However, when comparing the monotherapy and untreated groups, the results were different; oxcarbazepine, levetiracetam, and topiramate had no significant effect on bone metabolism. Phosphorus and magnesium were significantly lower in the valproic acid group than the untreated group (adjusted p < .05, Cliff's delta .282-.768). CTX was significantly higher in the lamotrigine group than in the untreated group (adjusted p = .012, Cliff's delta = .316). SIGNIFICANCE: Epilepsy can affect many aspects of bone metabolism. After controlling epilepsy and other confounders that affect bone metabolism, we found that the effects of ASMs on bone metabolism differed. Oxcarbazepine, levetiracetam, and topiramate did not affect bone metabolism, and lamotrigine corrected some of the abnormal markers of bone metabolism in patients with epilepsy.

Recognizing Pediatric Tuberous Sclerosis Complex Based on Multi-Contrast MRI and Deep Weighted Fusion Network.

Multi-contrast magnetic resonance imaging (MRI) is wildly applied to identify tuberous sclerosis complex (TSC) children in a clinic. In this work, a deep convolutional neural network with multi-contrast MRI is proposed to diagnose pediatric TSC. Firstly, by combining T2W and FLAIR images, a new synthesis modality named FLAIR3 was created to enhance the contrast between TSC lesions and normal brain tissues. After that, a deep weighted fusion network (DWF-net) using a late fusion strategy is proposed to diagnose TSC children. In experiments, a total of 680 children were enrolled, including 331 healthy children and 349 TSC children. The experimental results indicate that FLAIR3 successfully enhances the visibility of TSC lesions and improves the classification performance. Additionally, the proposed DWF-net delivers a superior classification performance compared to previous methods, achieving an AUC of 0.998 and an accuracy of 0.985. The proposed method has the potential to be a reliable computer-aided diagnostic tool for assisting radiologists in diagnosing TSC children.

Case report: splicing effect of a novel heterozygous variant of the NUS1 gene in a child with epilepsy.

NUS1 is responsible for encoding of the Nogo-B receptor (NgBR), which is a subunit of cis-prenyltransferase. Over 25 variants in NUS1 have been reported, and these variants have been found to be associated with various phenotypes, such as congenital disorders of glycosylation (CDG) and developmental and epileptic encephalopathy (DEE). We report on the case of a patient who presented with language and motor retardation, epilepsy, and electroencephalogram abnormalities. Upon conducting whole-exome sequencing, we discovered a novel pathogenic variant (chr6:118024873, NM_138459.5: c.791 + 6T>G) in NUS1, which was shown to cause Exon 4 to be skipped, resulting in a loss of 56 amino acids. Our findings strongly suggest that this novel variant of NUS1 is responsible for the development of neurological disorders, including epilepsy. It is believed that the truncation of Nogo-B receptor results in the loss of cis-prenyltransferase activity, which may be the underlying cause of the disease.

Analysis of the efficacy and safety of inpatient and outpatient initiation of KD for the treatment of pediatric refractory epilepsy using generalized estimating equations.

Objective: To compare the efficacy and safety of inpatient and outpatient initiation ketogenic diet (KD) protocol of pediatric refractory epilepsy. Methods: Eligible children with refractory epilepsy were randomly assigned to receive KD with inpatient and outpatient initiation. The generalized estimation equation (GEE) model was used to analyze the longitudinal variables of seizure reduction, ketone body, weight, height, body mass index (BMI), and BMI Z-score at different follow-up times between the two groups. Results: Between January 2013 and December 2021, 78 and 112 patients were assigned to outpatient and inpatient KD initiation groups, respectively. There were no statistical differences between the two groups based on baseline demographics and clinical characteristics (all Ps > 0.05). The GEE model indicated that the rate of reduction of seizures≥50% in the outpatient initiation group was higher than that of the inpatient initiation group (p = 0.049). A negative correlation was observed between the seizure reduction and blood ketone body at 1, 6, and 12 months (all Ps < 0.05). There were no significant differences in height, weight, BMI, and BMI Z-score between the two groups over the 12-month period by the GEE models (all Ps > 0.05). Adverse events were reported by 31 patients (43.05%) in the outpatient KD initiation group and 46 patients (42.20%) in the inpatient KD initiation group, but these differences were not statistically significant (p = 0.909). Conclusion: Our study shows that outpatient KD initiation is a safe and effective treatment for children with refractory epilepsy.

The ketogenic diet for Dravet syndrome: A multicenter retrospective study.

OBJECTIVE: The ketogenic diet (KD) is one of the main treatments for drug-resistant epilepsy. However, there have been few multicenter reports on the use of the KD for the treatment of Dravet syndrome (DS). The aim of this study was to analyze the efficacy and safety of this approach based on a large number of multicenter cases. METHODS: This was a retrospective, multicenter cohort study from 14 centers in China. All patients were treated with the KD. We compared the effects of KD intervention time, age, and other factors. RESULTS: From March 2014 to March 2020, we treated 114 patients with DS with the KD. The male-to-female ratio was 67:47. The KD median initiation age was 3 y and 4 mo, and the median number of antiseizure medications (ASMs) was 2.4. KD therapy was the first choice for three patients. Exactly 10.5% of the patients started KD therapy after failure of the first ASM therapy, with 35.1% after failure of the second, 44.7% after the third, and 7% after the fourth or more. After KD therapy for 1, 3, 6, and 12 mo, the seizure-free rates were 14%, 32.5%, 30.7%, and 19.3%, respectively; KD efficacy (≥50% reduction in seizure frequency) were 57.9%, 76.3%, 59.6%, and 43%, respectively; the retention rates were 97.4%, 93%, 71.9%, and 46.5%, respectively; and the rates of adverse events were 25.2%, 19.9%, 11%, and 5.7%, respectively. CONCLUSIONS: Real-world, multicenter data analysis showed that the KD is effective for patients with DS and has a low incidence of side effects.

White-Sutton syndrome and congenital heart disease: case report and literature review.

BACKGROUND: White-Sutton syndrome is an autosomal dominant neurodevelopmental disorder caused by heterozygous mutation in POGZ (Pogo Transposable Element Derived with ZNF Domain). This syndrome is characterized by delayed psychomotor development apparent in infancy and abnormal facial features. To date, 80 cases have been reported in the literature; however, the phenotypic characterizations remain incomplete. CASE PRESENTATION: We herein describe a 2-year-old girl harboring a novel frameshift de novo POGZ variant: c.2746del (p.Thr916ProfsTer12). This patient presented with multisystem abnormalities affecting the digestive tract and neurological functioning, as well as congenital heart disease, which involved an atrial septal defect (18 × 23 × 22 mm) with pulmonary arterial hypertension (42 mmHg). The relationship between congenital heart disease and White-Sutton syndrome as described in both the GeneReview and OMIM databases (#616,364) remains unclear. A review of the current literature revealed 18 cases of White-Sutton syndrome with POGZ variants and congenital heart disease, and we summarize their clinical features in this study. CONCLUSIONS: Our findings based on the present case and those in the literature indicate a relationship between POGZ mutation and congenital heart disease.

Recent advances in highly accelerated 3D MRI.

Three-dimensional MRI has gained increasing popularity in various clinical applications due to its improved through-plane spatial resolution, which enhances the detection of subtle abnormalities and provides valuable clinical information. However, the long data acquisition time and high computational cost pose significant challenges for 3D MRI. In this comprehensive review article, we aim to summarize the latest advancements in accelerated 3D MR techniques. Covering over 200 remarkable research studies conducted over the past 20 years, we explore the development of MR signal excitation and encoding, advancements in reconstruction algorithms, and potential clinical applications. We hope that this survey serves as a valuable resource, providing insights into the current state of the field and serving as a guide for future research in accelerated 3D MRI.

Machine learning and statistic analysis to predict drug treatment outcome in pediatric epilepsy patients with tuberous sclerosis complex.

OBJECTIVES: We aimed to investigate the association between multi-modality features and epilepsy drug treatment outcomes and propose a machine learning model to predict epilepsy drug treatment outcomes with multi-modality features. METHODS: This retrospective study consecutively enrolled 103 epilepsy children with rare TSC. Multi-modality data were used to characterize risk factors for epilepsy drug treatment outcome of TSC, including clinical data, TSC1, and TSC2 genes test results, magnetic resonance imaging (MRI), computerized tomography (CT), and electroencephalogram (EEG). Three common feature selection methods and six common machine learning models were used to find the best combination of feature selection and machine learning model for epilepsy drug treatment outcomes prediction with multi-modality features for TSC clinical application. RESULTS: The analysis of variance based on selected 35 features combined with multilayer perceptron (MLP) model achieved the best area-under-curve score (AUC) of 0.812 (±0.005). Infantile spasms, EEG discharge type, epileptiform discharge in the right frontal area of EEG, drug-resistant epilepsy, gene mutation type, and type II lesions were positively correlated with drug treatment outcome. Age of onset and age of visiting doctors were negatively correlated with drug treatment outcome (p < 0.05). Our machine learning results found that among MRI features, lesion type is the most important in the outcome prediction, followed by location and quantity. CONCLUSION: We developed and validated an effective prediction model for epilepsy drug treatment outcomes of TSC. Our results suggested that multi-modality features analysis and MLP-based machine learning can predict epilepsy drug treatment outcomes of TSC.

Resective epilepsy surgery for West syndrome: The Hypsarrhythmic Asymmetric Scoring Scheme is a determining predictor of seizure outcome.

OBJECTIVE: It has been suggested that asymmetric hypsarrhythmia is associated with structural etiology. We devised the Hypsarrhythmic Asymmetric Scoring Scheme (HASS) to quantify the degree of hypsarrhythmic asymmetry in a retrospective series of patients who underwent surgical treatment at our center. The present study aimed to investigate the role of HASS in predicting the postsurgical seizure outcomes. METHODS: We retrospectively analyzed the records of 46 children with hypsarrhythmia who underwent resective epilepsy surgery between 2018 and 2020 and were followed up for at least 1 year after surgery. Hypsarrhythmia severity in each hemisphere was quantified and scored. The HASS score was calculated as the difference between the two hemispheres. Univariate results were submitted to logistic regression models to identify independent predictors for favorable surgical outcomes. RESULTS: Of the 46 patients who underwent resective surgery, Engel's class I-Ⅱ outcomes were achieved in 34 (73.9%). The Engel I-Ⅱ group had a significantly higher HASS score than the Engel Ⅲ-Ⅳ group (p<0.001). Multivariate analysis showed that the HASS score was the only significant predictor of good outcomes (p = 0.011). Further receiver operating characteristic analysis showed that a threshold of 7 yielded a better seizure outcome with a sensitivity of 97.06% and specificity of 83.33%. SIGNIFICANCE: As the first hypsarrhythmia scoring system specially designed for presurgical evaluation, the HASS score may contribute to predicting the postsurgical seizure outcome from the electroencephalography perspective.

Identification of Children's Tuberous Sclerosis Complex with Multiple-contrast MRI and 3D Convolutional Network.

Identifying rare tuberous sclerosis complex (TSC) children is valuable and crucial. Magnetic resonance imaging (MRI) is used for rare TSC diagnoses. In this work, T2w and FLAIR were combined as a new modality named FLAIR3 to maximize the contrast between TSC lesions and normal-appearing brain tissues. After that, for the first time, we propose to use two different 3D CNN combined with late fusion strategies to diagnose TSC. A total of 520 children were enrolled in the study, including 260 health and 260 TSC children. The experiments had shown that the FLAIR3 could effectively improve the conspicuity of TSC lesions and classification performance. And the results showed the proposed late fusion method can improve the classification performance and achieve the state-of-the-art performance of the AUC of 0.994 and the accuracy of 0.971, which could be treated as an effective computer-aided diagnostic tool to help clinical radiologists diagnose TSC children. Clinical Relevance- Our deep learning method can be a non-invasive, efficient, and reliable way to help clinical radiologists to identify TSC patients. FLAIR3 can provide clinicians with a new modality to accurately localize TSC lesions in TSC patients.