Bioinformatics-Guided Discovery of miRNAs Involved in Apoptosis Modulated by Parthenolide Combined with Vincristine in The NALM6 Cell Line.

OBJECTIVE: Acute lymphoblastic leukemia (ALL) is a highly heterogeneous leukemia. Despite the current improvement in conventional chemotherapy and high survival rates, the outcomes remain challenging. Sesquiterpen extracted from the Tanacetum parthenium, parthenolide, is a potential anticancer agent that can modulate the expression of miRNAs and induce apoptosis. The objective of this study was to investigate the effect of parthenolide in combination with vincristine and alone on the apoptosis rate and expression of miR-125b-5p, miR-181b-5p, and miR-17-5p in the NALM6 cell line. MATERIALS AND METHODS: In this experimental study, cell viability and metabolic activity were determined through MTT assay and PI staining. Flow cytometry was applied to evaluate the rate of apoptosis. The expression of miRNAs was assessed using real-time polymerase chain reaction. Bioinformatic analyses, including Cytoscape, RNAhybrid, and signaling pathway analysis were employed to investigate the association of miR-17-5p, miR-181b-5p and miR-125b- 5p with apoptosis. Further, molecular docking served to validate the modulation of these miRNAs by parthenolide and vincristine treatment. RESULTS: The MTT assay indicated that 7.7 µM of parthenolide decreased the metabolic activity to 50% after 48 hours. PI staining analysis indicated that at concentrations below the half maximal inhibitory concentration, parthenolide caused 50% cell death. Flow cytometric analysis indicated that parthenolide (1.925 µM) in combination with vincristine (1.2 nM) induced apoptosis in 83.2% of the cells. Real-time quantitative reverse transcription polymerase chain reaction (qRTPCR) analysis showed significant changes in the expression levels of miR-17-5p, miR-125b-5p, and miR-181b-5p. Moreover, the combination therapy downregulated the expression of miRNAs significantly. This was consistent with our bioinformatic analysis demonstrating that the studied miRNAs are regulators of apoptosis. Finally, molecular docking validated the modulation of the miRNAs by parthenolide and vincristine. CONCLUSION: Parthenolide in combination with vincristine triggers apoptosis at a high rate in the NALM6 cell line. Moreover, this combination therapy can decrease the expression of miR-17-5p, miR-181b-5p, and miR-125b-5p.

Cavernous hemangioma of corpus imitating endometrial polyp in a young non-pregnant woman: A case report study.

Key Clinical Message: Cavernous hemangiomas are rare and have various non-specific clinical presentations, such as menorrhagia. It can mimic different diseases such as endometrial polyps. Pathologists and gynecologists should be aware of performing histopathological examinations of this neoplasm for accurate diagnosis and to avoid unwarranted therapeutic interventions. Abstract: Internal genital tract hemangiomas are rare and can be divided into capillary and cavernous. We present a rare case of cavernous hemangioma (CH) of the corpus in a young, non-pregnant woman. The patient was a 28-year-old woman who had complained of menorrhagia for 2.5 years. Sonography showed a hypoechoic intramural area measuring 35 × 23 mm. Histomorphology revealed neoplastic proliferation of dilated thin-walled arteries of various sizes within the myometrium. Due to the variable clinical presentations of CH, histopathological examination should be performed for an accurate diagnosis. It is a rare entity and we recommend training pathologists and gynecologists on this neoplasm for accurate diagnosis and to avoid unwarranted therapeutic interventions.

Plasmonic porous micro- and nano-materials based on Au/Ag nanostructures developed for photothermal cancer therapy: challenges in clinicalization.

Photothermal therapy (PTT) has developed in recent decades as a relatively safe method for the treatment of cancers. Recently, various species of gold and silver (Au and Ag) nanostructures have been developed and investigated to achieve PTT due to their highly localized surface plasmon resonance (LSPR) effect. Concisely, the collective oscillation of electrons on the surface of Au and Ag nanostructures upon exposure to a specific wavelength (depending on their size and shape) and further plasmonic resonance leads to the heating of the surface of these particles. Hence, porous species can be equipped with tiny plasmonic ingredients that add plasmonic properties to therapeutic cargoes. In this case, a precise review of the recent achievements is very important to figure out to what extent plasmonic photothermal therapy (PPTT) by Au/Ag-based plasmonic porous nanomedicines successfully treated cancers with satisfactory biosafety. Herein, we classify the various species of LSPR-active micro- and nano-materials. Moreover, the routes for the preparation of Ag/Au-plasmonic porous cargoes and related bench assessments are carefully reviewed. Finally, as the main aim of this study, principal requirements for the clinicalization of Ag/Au-plasmonic porous cargoes and their further challenges are discussed, which are critical for specialists in this field.

Safety and efficacy of endovascular thrombolysis in patients with acute cerebral venous sinus thrombosis: A systematic review.

INTRODUCTION: Cerebral venous sinus thrombosis (CVST) is an uncommon but fatal cause of stroke worldwide. Endovascular treatments could be life-saving in patients who don't treat with anticoagulants as a mainstay of treatment. Currently, there is no consensus considering the safety, efficacy, and also selected approaches of endovascular intervention for these patients. This systematic review evaluates the literature on endovascular thrombolysis (EVT) in CVST patients. MATERIALS AND METHODS: A comprehensive search was conducted through PubMed and Scopus databases between 2010 and 2021, with additional sources identified through cross-referencing. The primary outcomes were the safety and efficacy of EVT in CVST, including catheter-related and non-catheter-related complications, clinical outcomes, and radiological outcomes. RESULTS: A total of 10 studies comprising 339 patients were included. Most of the patients presented with headaches (86.72%) and/or focal neurologic deficits (45.43%) (modified Rankin Scale of 5 in 55.88%). Acquired coagulopathy and/or consuming estrogen/progesterone medication were the most frequent predisposing factors (45.59%). At presentation, 68.84% had multi-sinus involvement, and 28.90% had venous infarcts and/or intracranial hemorrhage (ICH). The overall complication rate was 10.3%, with a 2.94%, 1.47%, and 1.17% rate of ICH, herniation, and intracranial edema, respectively. The complete and partial postoperative radiographic resolution was reported in 89.97% of patients, increasing to 95.21% during the follow-up. Additionally, 72.22% of patients had no or mild neurologic deficit at discharge, rising to 91.18% at the last follow-up. The overall mortality rate was 7.07%. CONCLUSIONS: EVT can be an effective and safe treatment option for patients with refractory CVST or contraindications to systemic anticoagulation.

Metabolic Pathways in Breast Cancer Reprograming: An Insight to Non-Coding RNAs.

Cancer cells reprogram their metabolisms to achieve high energetic requirements and produce precursors that facilitate uncontrolled cell proliferation. Metabolic reprograming involves not only the dysregulation in glucose-metabolizing regulatory enzymes, but also the enzymes engaging in the lipid and amino acid metabolisms. Nevertheless, the underlying regulatory mechanisms of reprograming are not fully understood. Non-coding RNAs (ncRNAs) as functional RNA molecules cannot translate into proteins, but they do play a regulatory role in gene expression. Moreover, ncRNAs have been demonstrated to be implicated in the metabolic modulations in breast cancer (BC) by regulating the metabolic-related enzymes. Here, we will focus on the regulatory involvement of ncRNAs (microRNA, circular RNA and long ncRNA) in BC metabolism, including glucose, lipid and glutamine metabolism. Investigation of this aspect may not only alter the approaches of BC diagnosis and prognosis, but may also open a new avenue in using ncRNA-based therapeutics for BC treatment by targeting different metabolic pathways.

Bcl11a and the Correlated Key Genes Ascribable to Globin Switching: An In-silico Study.

BACKGROUND: Reactivation of HbF is a potential strategy to ameliorate symptoms of hemoglobinopathies such as sickle cell disease and b-thalassemia. After birth, there is a switch from fetal to adult hemoglobin, for which the molecular mechanisms and key regulators await further understanding in order to develop effective methods for HbF reactivation. Bcl11a, one of the major HbF reactivation regulators, demonstrates no significant changes at transcriptional levels in F erythroblasts compared to the non-HbF expressing cells. Therefore, it is possible that posttranscriptional regulation and epigenetic effects, for which the miRNAs play an important role, are the primary causes of the decreased Bcl11a protein level in adult erythroblasts. OBJECTIVE: This paper aims to determine the differentially expressed mRNAs and miRNAs of erythroblasts in HSCs from the fetal liver and bone marrow. METHODS: Raw high-throughput sequencing data (GSE110936, GSE90878) was downloaded from Gene Expression Omnibus (GEO) database. After RNAseq analysis, several data sets and tools were used to select key genes and examine selection validation. RESULTS: We selected 42 DEmRNAs and nine DEmiRs, including hsa-let-7f-5p, hsa-miR-21-5p, hsamiR- 22-3p, hsa-miR-126-5p, hsa-miR-146b-5p, hsa-miR-181a-5p, hsa-miR-92a-3p, hsa-miR-25-3p and hsa-miR-191-5p. Furthermore, hub genes including hist1h2bl, al133243.2, trim58, abcc13, bpgm, and fam210b were identified in the coexpression network, as well as RPS27A in the PPI network. Functional analysis revealed that these DEmRNAs and DEmiRs might play a role in gene expression regulation at multiple levels. Gene set enrichment analysis, in particular, revealed a possible role for genes in the globin switching process. CONCLUSION: According to our findings, a number of the DEmRNAs and DEmiRs may play significant roles in globin switching regulation and thus have the potential to be applied for HbF reactivation.

Watch out for neuromyelitis optica spectrum disorder onset or clinical relapse after COVID-19 vaccination: What neurologists need to know?

INTRODUCTION: The ongoing global COVID-19 pandemic has dramatically impacted our lives. We conducted this systematic review to investigate the safety of the COVID-19 vaccines in NMOSD patients. METHODS: We systematically searched PubMed, Scopus, Web of Science, and Embase from the beginning of the COVID-19 vaccination to March 1, 2022. Except for the letters, posters, and reviews, we included all related articles to answer two main questions. Our first question examined the occurrence of NMOSD onset as an adverse effect of the COVID-19 vaccine. Our second question investigated the safety of the COVID-19 vaccines in NMOSD patients. RESULTS: Out of 262 records, nine studies, including five studies for the first question and four studies for the second question, met the inclusion criteria. Out of the six patients with NMOSD onset after COVID-19 vaccination, five (83.3%) were female. The median time to NMOSD onset was 6.5 days, and the frequency of the COVID-19 vaccine type was identical in all patients. The most common presentation was longitudinally extensive transverse myelitis, significantly improved by pulse methylprednisolone with or without plasma exchange. The maintenance therapy was described only in three patients: rituximab (n=2) and azathioprine (n=1). Regarding the second question, out of 67 patients, 77.61% were female, with a mean age of 54.75 years old, a mean EDSS of 2.83, and a mean disease duration of 9.5 years. 77% reported at least one preexisting comorbidity. 88.05% were under treatment, most of which were rituximab and azathioprine. 98.50% received two doses of the COVID-19 vaccine. mRNA vaccines were the most commonly used vaccine(86.56%), which were well tolerated. No significant adverse event was reported, and local pain was the most frequently reported. 4.67% of the patients experienced a clinical relapse after a mean interval of 49.75 days, which was mainly mild to moderate in severity. Unfortunately, the data on the COVID-19 vaccines were missing. CONCLUSION: The analysis suggests the safety profile of the COVID-19 vaccines. All NMOSD patients are strongly recommended to vaccinate for COVID-19. To maximize the effectiveness of the COVID-19 vaccines, further studies are needed to draw the best practice for vaccination.

Synchronous Occurrence of Papillary Thyroid Carcinoma and Medullary Carcinoma in the Setting of Hashimoto's Thyroiditis and Multi Nodular Goiter.

Coexistence of follicular epithelial and bilateral parafollicular cells derivative of carcinomas in the setting of Hashimoto's thyroiditis and multinodular goiter is a very rare event. Of course, all benign and malignant thyroid lesions are more prevalent in iodine deficient areas. It seems that the context for identifying the pathways influencing thyroid carcinogenesis especially coincidence form has not yet been fully understood and needs further investigation. Here, we present a case with the synchronous occurrence of papillary thyroid carcinoma and medullary thyroid carcinoma in the setting of Hashimoto's thyroiditis and multinodular goiter. A 54-year-old woman complained of a painless mass in the anterior region of the neck. The physical examination of the patient revealed multiple nodules in her thyroid gland. In ultrasound findings, she presented with thyroid enlargement associated with multiple isoechoic and hypoechoic nodules in both lobes. Thyroid fine needle aspiration results suggested a diagnosis of medullary thyroid carcinoma in the setting of Hashimoto's thyroiditis and multinodular goiter . The frozen sections, permanent sampling, and IHC examination showed the coexistence of papillary thyroid carcinoma with bilateral medullary thyroid carcinoma in the setting of Hashimoto's disease and multinodular goiter . Studies debated about the risk factors of these pathologies including the same environmental issues or mutations in genomes and they emphasized surgeons should be aware of these lesions for diagnosis and interventional treatments. Following up the Hashimoto's thyroiditis and multinodular goiter is required for detection ofoccult malignancies, and hence the proper management and treatment should be performed.

Cellular uptake and apoptotic properties of gemini curcumin in gastric cancer cells.

INTRODUCTION: Curcumin is a polyphenolic natural compound, which has demonstrated to possess antioxidant, anti-inflammatory, and anticancer effects in vitro & in vivo. However, its applicability in cancer therapy has been limited due to its poor cellular uptake. Here, we aimed to evaluate the anticancer effect of novel gemini curcumin (Gemini-Cur) on the gastric cancer AGS cells. METHOD: The AGS cancerous and HFF-2 non-cancerous cells were treated with Gemini-Cur and curcumin (Cur) in a time- and dose-dependent manner. Cellular toxicity was studied using MTT, fluorescence microscopy, annexin V/FITC, and cell cycle assays. Additionally, real-time PCR and western blotting were employed to evaluate the expression of Bax, Bcl-2 and survivin genes. RESULTS: Our data indicated that Gemini-Cur is significantly taken into AGS cells compared to Cur. Moreover, the viability of Gemini-Cur treated cells was significantly reduced in a time- and dose-dependent manner (p < 0.001). Gemini-Cur compound induced G2/M cell cycle arrest that was followed by apoptosis in a time-dependent manner (p < 0.0001). DISCUSSION: Taken together, our findings support the idea that Gemini-Cur has the potential to be considered as an anticancer agent.

Detection of COVID-19 in X-ray images by classification of bag of visual words using neural networks.

Coronavirus disease 2019 (COVID-19) was classified as a pandemic by the World Health Organization in March 2020. Given that this novel virus most notably affects the human respiratory system, early detection may help prevent severe lung damage, save lives, and help prevent further disease spread. Given the constraints on the healthcare facilities and staff, the role of artificial intelligence for automatic diagnosis is critical. The automatic diagnosis of COVID-19 based on medical images is, however, not straightforward. Due to the novelty of the disease, available X-ray datasets are very limited. Furthermore, there is a significant similarity between COVID-19 X-rays and other lung infections. In this paper, these challenges are addressed by proposing an approach consisting of a bag of visual words and a neural network classifier. The proposed method can classify X-ray chest images into non-COVID-19 and COVID-19 with high performance. Three public datasets are used to evaluate the proposed approach. Our best accuracy on the first, second, and third datasets is 96.1, 99.84, and 98 percent. Since detection of COVID-19 is important, sensitivity is used as a criterion. The proposed method's best sensitivities are 90.32, 99.65, and 91 percent on these datasets, respectively. The experimental results show that extracting features with the bag of visual words results in better classification accuracy than the state-of-the-art techniques.

The relationship between thyroid function tests and sleep quality: cross-sectional study.

INTRODUCTION: The relationship between sleep and hormones have long been recognized. Studies indicated that sleep quality is one of the major modulatory effects on the endocrine system. In this study, we aimed to assess the serum concentration of thyroid hormones in individuals who suffered from low quality sleep. MATERIAL AND METHODS: Based on the Pittsburgh Sleep Quality Index and ISMA Stress questionnaire, we divided 83 participants into two groups. Forty-one individuals with low quality sleep group and 42 with good quality sleep group, all from the male students of a medical school in Tehran, Iran, participated in this descriptive and cross-sectional study. Then, serum levels of thyroid hormones including free T3, free T4, and TSH were assessed and compared between two groups. RESULTS: There were a significant increase in serum levels of FT4 (p=0.01) and TSH (p=0.02). There were also meaningful correlations between sleep score and stress score (p=0.008) as well as stress score and FT4 (p=0.03) in the case group. CONCLUSIONS: The current study showed that thyroid function tests (T4 and TSH) significantly rose in the poor sleep condition. We also found correlations between sleep score, stress score, and FT4 in the poor sleep condition group that suggest low sleep quality can affect thyroid hormones.

EBST: An Evolutionary Multi-Objective Optimization Based Tool for Discovering Potential Biomarkers in Ovarian Cancer.

Ovarian cancer is the deadliest gynecologic malignancy, mainly due to limitations in early diagnosis. With advances in high-throughput technologies, research interest in identifying novel and customized tumor biomarkers for early detection and diagnosis is rapidly growing. Here we introduce a new tool called EBST to select microRNAs with biomarker potency in ovarian cancer. This tool has pre-processing options and Its core is the use of Modified Multi Objective Imperialist Competitive Algorithm and six objective functions based on the classifier performance/structure evaluation, clustering error and mRMR filter. In this paper, we used the FDR filter in the pre-processing stage and considered five objective functions, four of which relate to the l1-SVM classifier performance and one to the average mRMR ranking. The proposed method has identified 11 microRNAs including hsa-miR-6784-5p, hsa-miR-1228-5p, hsa-miR-8073, hsa-miR-6756-5p, hsa-miR-1307-3p, hsa-miR-4697-5p, hsa-miR-3663-3p, hsa-miR-328-5p, hsa-miR-1228-3p, hsa-miR-6821-5p, hsa-miR-1268a. Data classification by the proposed model showed 100 percent sensitivity, 99.38 percent specificity, 99.69 percent accuracy and 99.39 percent positive predictive value. In comparison with routine state-of-the-art methods, superiority of our method was confirmed. The biological evaluation of selected microRNAs using bioinformatics tools and published articles confirms their role in cancer signaling pathways. The tool and its MATLAB code are freely available at https://github.com/hanif-y.

Convolutional neural network with autoencoder-assisted multiclass labelling for seizure detection based on scalp electroencephalography.

OBJECTIVE: In long-term video-monitoring, automatic seizure detection holds great promise as a means to reduce the workload of the epileptologist. A convolutional neural network (CNN) designed to process images of EEG plots demonstrated high performance for seizure detection, but still has room for reducing the false-positive alarm rate. METHODS: We combined a CNN that processed images of EEG plots with patient-specific autoencoders (AE) of EEG signals to reduce the false alarms during seizure detection. The AE automatically logged abnormalities, i.e., both seizures and artifacts. Based on seizure logs compiled by expert epileptologists and errors made by AE, we constructed a CNN with 3 output classes: seizure, non-seizure-but-abnormal, and non-seizure. The accumulative measure of number of consecutive seizure labels was used to issue a seizure alarm. RESULTS: The second-by-second classification performance of AE-CNN was comparable to that of the original CNN. False-positive seizure labels in AE-CNN were more likely interleaved with "non-seizure-but-abnormal" labels than with true-positive seizure labels. Consequently, "non-seizure-but-abnormal" labels interrupted runs of false-positive seizure labels before triggering an alarm. The median false alarm rate with the AE-CNN was reduced to 0.034 h-1, which was one-fifth of that of the original CNN (0.17 h-1). CONCLUSIONS: A label of "non-seizure-but-abnormal" offers practical benefits for seizure detection. The modification of a CNN with an AE is worth considering because AEs can automatically assign "non-seizure-but-abnormal" labels in an unsupervised manner with no additional demands on the time of the epileptologist.

Anti-Cancer Effect of Melatonin via Downregulation of Delta-like Ligand 4 in Estrogen-Responsive Breast Cancer Cells.

BACKGROUND: The Notch signaling pathway has a key role in angiogenesis and Delta - Like Ligand 4 (DLL4) is one of the main ligands of Notch involved in cell proliferation in sprouting vessels. OBJECTIVE: In this study, we aimed to evaluate the expression of DLL4 in primary breast tumors and to examine the effect of melatonin on DLL4 expression in vitro. METHODS: Eighty-five breast tumor and paired adjacent non-tumor tissue samples were collected. Apoptosis assay was performed on breast cancer cells to evaluate melatonin effects. Western blot and quantitative RT-PCR were used to measure DLL4 expression. Then, we investigated the effect of melatonin on the expression of DLL4 in four breast cancer cell lines at RNA and protein levels. We also performed a probabilistic neural network analysis to study genes closely associated with DLL4 expression. RESULTS: Our results showed a significantly higher expression of DLL4 in tumor tissues compared to non-tumor tissues (P = 0.027). Melatonin treatment substantially attenuated DLL4 expression in BT474 and MCF-7 cells, but not in SK-BR-3 and MDA-MB-231 cells. Also, melatonin induced apoptosis in all four cell lines. Network analysis revealed a set of 15 genes that had close association and interaction with DLL4. DLL4 was overexpressed in breast cancer tissues as compared to the non-tumor tissues. CONCLUSION: It can be concluded that melatonin treatment attenuated DLL4 expression only in estrogen- responsive breast cancer cells and is able to induce apoptosis in breast cancer cells.

An Iranian man with increased thigh mass due to a hydatid cyst.

Background: Hydatid cyst is a zoonotic infection caused by Echinococcosis granulosus. Primary single-intramuscular hydatid disease is rare, even in endemic regions of the world. Here we report the case of exceptional thigh mass due to a hydatid cyst in an Iranian man. Case presentation: An 86-year-old man, initially presented to Velayat teaching hospital surgery clinic in May 2017 with a single right-thigh mass, but physical examinations of other organs were unremarkable. Based on sonographic findings, the differential diagnosis was hydatid cyst. He underwent surgical resection of the cyst. Histopathological results confirmed the diagnosis. There was no evidence of recurrence of the lesion during the 23-month follow-up. Conclusion: Increase thigh mass due to a hydatid cyst is a rare event. In endemic regions with the presence of hydatid cysts, especially physicians of surgical clinics have to consider differential diagnosis of hydatid cysts in unusual locations in case of such a lesion.

Novel Coronavirus Disease (COVID-19) and Central Nervous System Complications: What Neurologist Need to Know.

The novel coronavirus (Covid-19) is a family of large enveloped non-segmented positive-sense RNA viruses which has been considered as a global health concern as it has a very high transmissibility potential. Regarding to the similarity of the virus to SARS-CoV, it is postulated that the Covid-19 accumulates mainly in the nasal epithelia and lower respiratory airways. However, there is evidence suggesting the Covid-19 neurotropism which might contribute to respiratory failure. Here in we aim to review the central nervous system complications of the Covid-19 CoV since the emergence of the virus. Keywords: Novel Coronavirus, Covid19-Cov, CNS Complication, Nervous System.

The Impact of Pre-Cooling and CoQ10 Supplementation on Mediators of Inflammatory Cytokines in Elite Swimmers.

Chronic intensive exercise and hyperthermia may cause immune system function disturbance. We aimed to investigate the effect of 14-day coenzyme Q10 (CoQ10) supplementation and pre-cooling strategy on serum changes of inflammatory cytokines [interleukin-10 (IL-10), interleukin-8 (IL-8), interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-α)], and high-sensitivity C-reactive protein (hs-CRP), myeloperoxidase (MPO) and xanthine oxidase (XO) enzymes, leukocyte counts (WBC), and stress hormones (catecholamine and cortisol) responses in elite swimmers during competition phase. Thirty-six healthy males were randomly selected and divided into four groups of CoQ10, precooling, supplementation with precooling, and control. Blood sampling was done pre and post (before and after acute recoding bout) administration of CoQ10 and pre-cooling. There was no significant statistical difference among groups for the indices levels of IL-10, IL-8, IL-6, TNF-α, hs-CRP, catecholamine, cortisol, MPO, XO, and WBC counts at the pre sampling (P > 0.05). While, pre-cooling and control groups show a significant increase indices levels compared to the supplementation and supplementation with precooling groups in the post-sampling (two stages), (P ˂ 0.05). Short-term oral CoQ10 supplementation prevents adverse changes mediators of inflammatory cytokines following heavy swimming trainings and acute recording bout. In addition, pre-cooling strategy individually has no desired effect on the mediators of inflammatory cytokines.

Autoencoding of long-term scalp electroencephalogram to detect epileptic seizure for diagnosis support system.

INTRODUCTION: Epileptologists could benefit from a diagnosis support system that automatically detects seizures because visual inspection of long-term electroencephalograms (EEGs) is extremely time-consuming. However, the diversity of seizures among patients makes it difficult to develop universal features that are applicable for automatic seizure detection in all cases, and the rarity of seizures results in a lack of sufficient training data for classifiers. METHODS: To overcome these problems, we utilized an autoencoder (AE), which is often used for anomaly detection in the field of machine learning, to perform seizure detection. We hypothesized that multichannel EEG signals are compressible by AE owing to their spatio-temporal coupling and that the AE should be able to detect seizures as anomalous events from an interictal EEG. RESULTS: Through experiments, we found that the AE error was able to classify seizure and nonseizure states with a sensitivity of 100% in 22 out of 24 available test subjects and that the AE was better than the commercially available software BESA and Persyst for half of the test subjects. CONCLUSIONS: These results suggest that the AE error is a feasible candidate for a universal seizure detection feature.

Seizure detection by convolutional neural network-based analysis of scalp electroencephalography plot images.

We hypothesized that expert epileptologists can detect seizures directly by visually analyzing EEG plot images, unlike automated methods that analyze spectro-temporal features or complex, non-stationary features of EEG signals. If so, seizure detection could benefit from convolutional neural networks because their visual recognition ability is comparable to that of humans. We explored image-based seizure detection by applying convolutional neural networks to long-term EEG that included epileptic seizures. After filtering, EEG data were divided into short segments based on a given time window and converted into plot EEG images, each of which was classified by convolutional neural networks as 'seizure' or 'non-seizure'. These resultant labels were then used to design a clinically practical index for seizure detection. The best true positive rate was obtained using a 1-s time window. The median true positive rate of convolutional neural networks labelling by seconds was 74%, which was higher than that of commercially available seizure detection software (20% by BESA and 31% by Persyst). For practical use, the median of detected seizure rate by minutes was 100% by convolutional neural networks, which was higher than the 73.3% by BESA and 81.7% by Persyst. The false alarm of convolutional neural networks' seizure detection was issued at 0.2 per hour, which appears acceptable for clinical practice. Moreover, we demonstrated that seizure detection improved when training was performed using EEG patterns similar to those of testing data, suggesting that adding a variety of seizure patterns to the training dataset will improve our method. Thus, artificial visual recognition by convolutional neural networks allows for seizure detection, which otherwise currently relies on skillful visual inspection by expert epileptologists during clinical diagnosis.