Brain Magnetic Resonance Imaging Classification Using Deep Learning Architectures with Gender and Age.

Usage of effective classification techniques on Magnetic Resonance Imaging (MRI) helps in the proper diagnosis of brain tumors. Previous studies have focused on the classification of normal (nontumorous) or abnormal (tumorous) brain MRIs using methods such as Support Vector Machine (SVM) and AlexNet. In this paper, deep learning architectures are used to classify brain MRI images into normal or abnormal. Gender and age are added as higher attributes for more accurate and meaningful classification. A deep learning Convolutional Neural Network (CNN)-based technique and a Deep Neural Network (DNN) are also proposed for effective classification. Other deep learning architectures such as LeNet, AlexNet, ResNet, and traditional approaches such as SVM are also implemented to analyze and compare the results. Age and gender biases are found to be more useful and play a key role in classification, and they can be considered essential factors in brain tumor analysis. It is also worth noting that, in most circumstances, the proposed technique outperforms both existing SVM and AlexNet. The overall accuracy obtained is 88% (LeNet Inspired Model) and 80% (CNN-DNN) compared to SVM (82%) and AlexNet (64%), with best accuracy of 100%, 92%, 92%, and 81%, respectively.

Hippocampus Segmentation Using U-Net Convolutional Network from Brain Magnetic Resonance Imaging (MRI).

Hippocampus is a part of the limbic system in human brain that plays an important role in forming memories and dealing with intellectual abilities. In most of the neurological disorders related to dementia, such as, Alzheimer's disease, hippocampus is one of the earliest affected regions. Because there are no effective dementia drugs, an ambient assisted living approach may help to prevent or slow the progression of dementia. By segmenting and analyzing the size/shape of hippocampus, it may be possible to classify the early dementia stages. Because of complex structure, traditional image segmentation techniques can't segment hippocampus accurately. Machine learning (ML) is a well known tool in medical image processing that can predict and deliver the outcomes accurately by learning from it's previous results. Convolutional Neural Networks (CNN) is one of the most popular ML algorithms. In this work, a U-Net Convolutional Network based approach is used for hippocampus segmentation from 2D brain images. It is observed that, the original U-Net architecture can segment hippocampus with an average performance rate of 93.6%, which outperforms all other discussed state-of-arts. By using a filter size of [Formula: see text], the original U-Net architecture performs a sequence of convolutional processes. We tweaked the architecture further to extract more relevant features by replacing all [Formula: see text] kernels with three alternative kernels of sizes [Formula: see text], [Formula: see text], and [Formula: see text]. It is observed that, the modified architecture achieved an average performance rate of 96.5%, which outperforms the original U-Net model convincingly.

Diagnostic Yield of Closed Pleural Biopsy in Undiagnosed Exudative Pleural Effusions.

Introduction: In our practice, etiological diagnosis of pleural effusion is sometimes difficult despite cytological, biochemical and microbiological tests. The aim of the present study was to make an etiological diagnosis by means of closed pleural biopsy in undiagnosed pleural effusions. Methods:The study group consisted of patients with exudative pleural effusion where etiology was not determined by means of conventional cytological, biochemical and microbiological investigations. Pleural tissue was obtained by Abrams pleural biopsy needle. Pleural biopsy was subjected to histopathology, Ziehl-Neelsen (Z-N) staining and mycobacterial culture. Results:Four hundred fifty eight patients with pleural effusion were screened over three years and 82 of them were found to have undiagnosed exudative effusion after investigations, as mentioned in the Method section. The age of the 82 subjects [56 (68.29%) men and 26 (31.71%) women] ranged from 15 to 65 years (mean 32.6). Histopathology showed epitheloid granuloma with caseation necrosis in 50 (60.9%) patients and non-specific chronic inflammation in 14 (17.1%) subjects. Ziehl-Neelsen stain was positive for acid fast bacilli (AFB) in 10 (12.2%) patients and culture of pleural tissue showed the presence of Mycobacterium tuberculosis in 18 (21.9%) patients. Conclusion:In the era of thoracoscopy, percutaneous closed needle pleural biopsy still holds a significant role in cases of undiagnosed exudative pleural effusions where thoracoscopy is not available, being also a cost effective approach for developing countries. It can achieve specific diagnosis among 86.6% of cases with undiagnosed exudative pleural effusions.

Flexible Array of Microsupercapacitor for Additive Energy Storage Performance Over a Large Area.

An on-chip microsupercapacitor (MSC) pattern is obtained by layer-by-layer spray deposition of both manganese dioxide (MnO2) nanoparticle-coated carbon nanotubes (MnO2-CNTs) and MnO2 nanosheet-decorated reduced graphene oxide (MnO2-rGO) on mechanically robust, flexible polyethylene terephthalate. Layer-by-layer patterning of MSC electrodes offers rapid in-plane diffusion of electrolyte ions in electrodes the layered electrode and hence ultrahigh capacitance and energy density of 7.43 mF/cm2 (32300 mF/cm3) and 0.66 µW h/cm2 (2870 µW h/cm3), respectively, are obtained. A robust electrochemical response was measured under multiple bending of the solid-state flexible MSC as well as under repetitive cycles (∼5000).

Synergistic effect in the heterostructure of ZnCo2O4 and hydrogenated zinc oxide nanorods for high capacitive response.

Herein, a novel heterostructure was fabricated by combining electrochemically and optically active materials to achieve a high capacitive response of 896 F g-1 at 5 A g-1. A network of ZnCo2O4 nanorods (NRs) was directly grown on a three-dimensional matrix of H : ZnO NRs (ZnCo2O4/H : ZnO NRs) that offered synergistic advantages by providing an optimum ion/charge transportation path, large electrochemically active surface area, and stable capacitive response during the electrolytic process. Furthermore, the fabricated solid-state asymmetric supercapacitor ZnCo2O4/H : ZnO NRs//activated carbon induced a large potential window of 1.5 V that offered excellent energy and power densities. In addition, optically active ZnCo2O4/H : ZnO NRs were also used for the conversion of optical energy over a broad wavelength range; thus, the as-fabricated asymmetric solid-state supercapacitor could easily provide the required power for the operation of a photodetector. Therefore, the unique heterostructure of ZnCo2O4/H : ZnO NRs not only presents excellent capacitive response but also demonstrates great potential for energy conversion.

Unexplained dyspnea in a patient of chronic arsenicosis: A diagnostic challenge and learning curve for physicians.

Chronic arsenic exposure causes cutaneous effects like hyperkeratosis, peripheral vascular disease, hypertension, ischemic heart disease, non-cirrhotic portal hypertension, hepatomegaly, peripheral neuropathy, respiratory involvement, bad obstetrical outcome, hematological disturbances, and diabetes mellitus. Here we present a case of a 24-year-old lady, with chronic exposure to arsenic, presenting to us with progressive dyspnea. We found pulmonary arterial hypertension (PAH) as a cause of her dyspnea. PAH can occur in arsenicosis, secondary to arsenic-induced chronic obstructive pulmonary disease (COPD), lung fibrosis, and portal hypertension, which we excluded by appropriate investigations in our case. We also excluded a familial or heritable form of PAH. Thus, with the exclusion of all these secondary causes of PAH, as well as a hereditary cause, we came to a conclusion that this PAH might be due to chronic arsenic exposure. To the best of our knowledge, no case of PAH in chronic arsenicosis has been reported to date.

Solitary pulmonary nodule: A rare presentation of pulmonary mucormycosis in an immunocompetent adult.

Pulmonary mucormycosis is a rare opportunistic infection of immunocompromised individuals. Here, we report a case of 70-year-old male, smoker presenting with high-grade fever for 2 weeks and episodes of hemoptysis. Contrast-enhanced computed tomography (CT) thorax revealed a solitary pulmonary nodule measuring 2.3 × 1.6 cm in the right upper lobe. CT guided fine needle aspiration cytology and true cut biopsy showed plenty of typical fungal hyphae consistent with the diagnosis of mucormycosis. Fungal culture confirmed the organism as mucor. Positron emission tomography-CT scan showed a non- 18 fluorodeoxy glucose avid nodule ruling out possibility of malignancy. Investigation did not reveal any evidence of immunosuppression. Patient was treated with intravenous liposomal amphotericin B for 4 weeks. Follow-up chest X-ray and CT scan after 6 weeks were normal.

Non-resolving pneumonia: A rare presentation of progressive disseminated histoplasmosis.

Histoplasmosis, a fungal disease caused by Histoplasma capsulatum, is endemic in North and South America. Except few scattered cases, the disease is considered to be a non-entity in India. Furthermore, disseminated histoplasmosis is rare in the immunocompetent individuals. We report an adolescent boy presenting as middle lobe consolidation which did not respond to antibiotics. His condition deteriorated with the development of mediastinal lymphadenopathy, pleural effusion and hepatosplenomegaly. A diagnosis of progressive disseminated histoplasmosis was established by his clinical findings as well as bronchoscopic biopsy, transbronchial needle aspiration cytology and bronchoalveolar lavage culture demonstrating Histoplasma capsulatum. The case represents a unique example of progressive disseminated histoplasmosis in an immunocompetent individual in India.

Role of common investigations in aetiological evaluation of exudative pleural effusions.

BACKGROUND: Pleural effusion is a common problem encountered in daily practice. To Establish aetiology of exudative effusions is a diagnostic challenge to general practitioners and even to pulmonologists especially in resource poor government hospitals with lack of investigations like thoracoscopy. Some recent studies had shown that around 2% of patients remained undiagnosed even after these investigations. AIMS AND OBJECTIVE: To evaluate the role of the commonly available investigations such as pleural fluid study, blind pleural biopsy, sputum examination, CT scan thorax, bronchoscopy in the aetiological evaluation of exudative effusions and to ascertain the proportion of cases which remain undiagnosed after all the above investigations. MATERIAL AND METHODS: This was a prospective single-centred cross-sectional study carried out at the NRS Medical College, Kolkata, India from February 2008 to February 2013 which included 568 patients of exudative pleural effusions. We performed commonly available procedures like pleural fluid study, blind pleural biopsy, sputum examination, CT scan thorax, bronchoscopic procedures to the diagnosis. RESULTS: Total number of patients studied were 568. Tuberculosis was the most common cause (54.57%) followed by malignancy (28.17%), empyema (10.56%), parapneumonic effusion (5.28%) and others. Carcinoma of the lung was the commonest cause of malignant effusions and bronchoscopic biopsy was given the highest yield of histological diagnosis (84.6%) followed by CT guided FNAC (77.6%) and pleural fluid cytology (55%). Highest yield to diagnose tubercular effusion was found in lymph node FNAC (81.5%) followed by pleural biopsy (62%). Sputum smear for AFB was positive in only 27.4% cases. Bleeding followed by pneumothorax were the most common complications. Complications are very less (1.3% and 0.9% respectively). 2 patients (0.34%) remained undiagnosed even after these all above said investigations. CONCLUSION: Above mentioned commonly available investigations can ascertain diagnosis in most of the cases in the aetiological evaluation of exudative effusions and they are relatively safe procedures.

Recovery of fluconazole sensitive Candida ciferrii in a diabetic chronic obstructive pulmonary disease patient presenting with pneumonia.

Chronic obstructive pulmonary disease (COPD) exacerbations admitted in intensive care units (ICUs) occur rarely due to fungal respiratory tract infections, but may occur when associated co-morbidities like diabetes mellitus coexist. Candida ciferrii is a new agent, recently was isolated from lung infections but usually resistant to fluconazole. Here, we report a rare case of pneumonia due to fluconazole sensitive Candida ciferrii in a COPD patient with known diabetes, admitted in our ICU.

The Incidence of Hyponatraemia and Its Effect on the ECOG Performance Status among Lung Cancer Patients.

CONTEXT: Hyponatraemia is one of the common electrolytic disorders which are associated with lung cancer. Hyponatraemia may influence the ECOG performance status at presentation. Also, to the best of our knowledge, we found only limited Indian studies where the ECOG score was correlated with the serum sodium status in lung cancer patients on presentation. AIM: To assess the incidence of hyponatraemia among the patients of carcinoma of the lung before putting them into the specific treatment category for cancer and to check the effects on their ECOG performance status. SETTINGS AND DESIGN: A cross-sectional, observational study was conducted on 116 consecutive patients of lung cancer during the period from November 2011 to October 2012. MATERIAL AND METHODS: The patients with a histologically proven diagnosis of lung cancer were grouped initially according to their ECOG performance statuses. The serum sodium value of each patient was measured and the hyponatraemic patients were given treatment according to the protocol. The correlation of the ECOG performance status with the serum sodium of the lung cancer patients was measured. To check for any laboratory error in serum sodium, we selected (n = 58) age, sex and socioeconomic matched control patients. RESULTS: At presentation 44.8% of the lung cancer patients showed hyponatraemia [52/116]. The ECOG score was significantly poor in the advanced clinical stages (ECOG ≤2 Vs ECOG ≥ 3 in NSCLC cases, χ(2) =11.25, P=.0008). The ECOG performance status score at admission showed a negative correlation with the serum sodium status which was measured on admission among all the patients (Pearson correlation coefficient = - 0.186). The clinical stage of the lung cancer also showed a positive correlation with the ECOG score at admission in our study (Pearson correlation coefficient = 0.295). CONCLUSION: Hyponatraemia is not an uncommon condition and it should be suspected and screened in each patient, as it may influence the ECOG performance status score, which serves as an important factor in the prognosis of lung cancer.

Diagnosis of sclerosing hemangioma of lung: don't rely on fine-needle aspiration cytology diagnosis alone.

Sclerosing hemangioma is a rare variety of benign pulmonary neoplasm. It usually presents as asymptomatic, solitary, peripheral, circumscribed lesions in middle-aged women. Here, we describe a 46-year-old woman presenting to us for evaluation of right parahilar lung mass. Previous chest radiography done 10 years back showed a lung mass of almost similar size. Computed Tomography (CT)-guided fine-needle aspiration cytology (FNAC) was suggestive of adenocarcinoma of lung. A well-circumscribed, capsulated, ovoid mass measuring 5.6 cm × 4 cm × 3 cm, adjacent to the transverse fissure of the right lung was excised by lateral thoracotomy. Histopathological examination along with immunohistochemistry was suggestive of sclerosing hemangioma of lung. A pathologist must consider the clinicoradiological features before coming to a final diagnosis of lung malignancy from FNAC. Whenever there is any confusion regarding lung mass, thoracotomy must be done for arriving at an exact diagnosis from histopathology.