In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

/pdf/review-of-deep-learning-concepts-cnn-architectures-hpbk62qzc2.pdf

Review of deep learning: concepts, CNN architectures, challenges, applications, future directions

Improving the performance of CNN to predict the likelihood of COVID-19 using chest X-ray images with preprocessing algorithms.

In this paper, an effective approach for the feature extraction of raw Electroencephalogram (EEG) signals by means of one-dimensional local binary pattern (1D-LBP) was presented. For the importance of making the right decision, the proposed method was performed to be able to get better features of the EEG signals. The proposed method was consisted of two stages: feature extraction by 1D-LBP and classification by classifier algorithms with features extracted. On the classification stage, the several machine learning methods were employed to uniform and non-uniform 1D-LBP features. The proposed method was also compared with other existing techniques in the literature to find out benchmark for an epileptic data set. The implementation results showed that the proposed technique could acquire high accuracy in classification of epileptic EEG signals. Also, the present paper is an attempt to develop a general-purpose feature extraction scheme, which can be utilized to extract features from different categories of EEG signals.

1D-Local Binary Pattern Based Feature Extraction for Classification of Epileptic EEG Signals, Applied Mathematics and Computation243 (2014): 209-219.

Acute kidney injury (AKI) consists of a rapid renal function decline which usually increases serum urea and creatinine levels. Since kidney injury begins by inducing biological and molecular changes which evolve to cellular damage, biomarkers could be used as tools for monitoring early AKI appearance, and predicting its recovery. Among the main AKI biomarkers the neutrophil gelatinase-associated lipocalin, cystatin C, kidney injury molecule-1, monocyte chemotactic peptide-1, N-acetyl-β-D-glucosaminidase, interleukin-18, liver-type fatty acid-binding protein, netrin-1, cycle arrest markers, endogenous ouabain, selenium-binding protein 1, and BPIFA2 marker, have been described. Even though novel biomarkers seem to be more helpful to early detect AKI and/or predict the need for renal replacement, and mortality compared to serum creatinine, more comprehensive studies are still required to determine their clinical utility.

Novel acute kidney injury biomarkers: their characteristics, utility and concerns

Diabetic nephropathy (DN) is the most serious microvascular complication of diabetes and the largest single cause of end-stage renal disease (ESRD) in many developed countries. DN is also associated with an increased cardiovascular mortality. It occurs as a result of interaction between both genetic and environmental factors. Hyperglycemia, hypertension, and genetic predisposition are the major risk factors. However, the exact mechanisms of DN are unclear. Despite the benefits derived from strict control of glucose and blood pressure, as well as inhibition of renin-angiotensin-aldosterone system, many patients continue to enter into ESRD. Thus, there is urgent need for improving mechanistic understanding of DN and then developing new and effective therapeutic approaches to delay the progression of DN. This review focuses on recent progress and future perspective about mechanistic insight and management of DN. Some preclinical relevant studies are highlighted and new perspectives of traditional Chinese medicine (TCM) for delaying DN progression are discussed in detail. These findings strengthen the therapeutic rationale for TCM in the treatment of DN and also provide new insights into the development of novel drugs for the prevention of DN. However, feasibility and safety of these therapeutic approaches and the clinical applicability of TCM in human DN need to be further investigated.

/pdf/mechanistic-insight-and-management-of-diabetic-nephropathy-3qn3lgzgsb.pdf

Mechanistic Insight and Management of Diabetic Nephropathy: Recent Progress and Future Perspective.

Melanoma is a type of skin cancer with a high mortality rate The different types of skin lesions result in an inaccurate diagnosis due to their high similarity Accurate classification of the skin lesions in their early stages enables dermatologists to treat the patients and save their lives This paper proposes a model for a highly accurate classification of skin lesions The proposed model utilized the transfer learning and pre-trained model with GoogleNet The model parameters are used as initial values, and then these parameters will be modified through training The latest well-known public challenge dataset, ISIC 2019, is used to test the ability of the proposed model to classify different kinds of skin lesions The proposed model successfully classified the eight different classes of skin lesions, namely, melanoma, melanocytic nevus, basal cell carcinoma, actinic keratosis, benign keratosis, dermatofibroma, vascular lesion, and Squamous cell carcinoma The achieved classification accuracy, sensitivity, specificity, and precision percentages are 9492%, 798%, 97%, and 8036%, respectively The proposed model can detect images that do not belong to any one of the eight classes where these images are classified as unknown images

/pdf/skin-lesions-classification-into-eight-classes-for-isic-2019-2awpgfqpu5.pdf

Skin Lesions Classification Into Eight Classes for ISIC 2019 Using Deep Convolutional Neural Network and Transfer Learning

Melanoma is deadly skin cancer. There is a high similarity between different kinds of skin lesions, which lead to incorrect classification. Accurate classification of a skin lesion in its early stages saves human life. In this paper, a highly accurate method proposed for the skin lesion classification process. The proposed method utilized transfer learning with pre-trained AlexNet. The parameters of the original model used as initial values, where we randomly initialize the weights of the last three replaced layers. The proposed method was tested using the most recent public dataset, ISIC 2018. Based on the obtained results, we could say that the proposed method achieved a great success where it accurately classifies the skin lesions into seven classes. These classes are melanoma, melanocytic nevus, basal cell carcinoma, actinic keratosis, benign keratosis, dermatofibroma, and vascular lesion. The achieved percentages are 98.70%, 95.60%, 99.27%, and 95.06% for accuracy, sensitivity, specificity, and precision, respectively.

Classification of Skin Lesions into Seven Classes Using Transfer Learning with AlexNet

Introduction Early detection of Acute Kidney Injury (AKI) is important. We conducted a prospective observational study to test the ability of cystatin C (CysC) to predict AKI and mortality among in intensive care unit (ICU) patients after acute myocardial infarction (MI). Methods One hundred patients were enrolled in this study, CysC and serum creatinine (SCr) were measured at admission in the ICU and then daily for three days.We assessed the ability of CysC and SCr at admission to predict sustained AKI and short-term mortality at one month using the area under the receiver operator characteristic curve (AUC). Results The mean relative increase in CysC after 24 hours was significantly higher than the mean relative increase in SCr (83±66% versus 14±48%; P>0.001).Of 100 patients, 35 patients had >50% increase in either CysC or SCr, seven patients had >50% increase in both markers and 58 patients had neither. CysC increased prior to elevation of SCr more frequently than the reverse (P > 0.0001). Baseline CysC was a good predictor of sustained AKI with an AUC of 0.86 (95% CI; 0.79-0.95), while the AUC of SCr was only 0.54 (95% CI; 0.40-0.69).Baseline CysC was a fair predictor of death with an AUC of 0.76 (95% CI; 0.65-0.87), while the AUC of SCr was only 0.65 (95% CI; 0.50-0.79). Conclusion CysC was an effective and earlier surrogate marker of decreased renal function than SCr in ICU population after MI. High CysC concentrations predict substantially increased risks of short-term mortality in the ICU after MI.

Cystatin C as an early marker of acute kidney injury and predictor of mortality in the intensive care unit after acute myocardial infarction.

This paper introduces an unsupervised fabric defect detection algorithm that does not need any user-adaptation and reveals high detection accuracy rates at the same time. This algorithm is mainly applied to patterned fabrics and can be extended to plain fabric. To achieve this, firstly, non-extensive standard deviation filtering is used with a suitable window size. Secondly, the image is split into blocks with size equals the periodic pattern of the fabric image. Thirdly, the square difference between each block median and the mean of all block medians are calculated. Finally, all these differences are forwarded to K-means clustering to classify the blocks into defected and non-defected ones. The proposed algorithm showed promising performance with overall detection success rate that reaches 95%.

Unsupervised patterned fabric defect detection using texture filtering and K-means clustering

Introducao: A deteccao precoce da nefropatia diabetica (ND) e importante. O acido urico serico (AUS) tem um papel ainda desconhecido no desenvolvimento de ND. Objetivo: Estudar a relacao entre AUS e hipertensao, nefropatia precoce e progressao da doenca renal cronica (DRC) no diabetes mellitus tipo 2 (DM2). Metodos: O estudo contou com 986 participantes, de acordo com a presenca e a duracao do diabetes, os pacientes foram classificados em tres grupos. O Grupo I incluiu 250 participantes saudaveis. O Grupo II incluiu 352 pacientes com inicio de diabetes 5 anos. Todos os participantes foram submetidos a exame clinico completo, medidas antropometricas, exames laboratoriais - incluindo hemoglobina glicosilada (HbA1C), bem como a razao entre triglicerides e lipoproteina de alta densidade (TG/HDL-C), AUS, razao creatinina/albumina (RCA) urinaria, e taxa estimada de filtracao glomerular (eTFG). Resultados: A razao AUS, PA, HbA1c, TG/HDL-C e RCA foi significativamente maior no grupo III do que no grupo I, II e em II do que I. A eTFG foi significativamente menor no grupo III do que nos grupos I, II e no II do que no I (p 6,1 mg/dl, > 6,2 mg/dl e > 6,5 mg/dl apresentou maior sensibilidade e especificidade para identificar hipertensao, nefropatia precoce e declinio da eTFG, respectivamente. Conclusao: Mesmo elevados niveis de AUS, foi associado ao risco de hipertensao, nefropatia precoce e declinio da eTFG. Alem disso, o nivel de AUS pode identificar o inicio da hipertensao, nefropatia precoce e progressao da DRC em DM2.

https://www.bjnephrology.org/wp-content/uploads/articles_xml/2175-8239-jbn-S0101-28002016000400403/2175-8239-jbn-S0101-28002016000400403-pt.pdf

Mohamed M. Fouad

Papers

Skin Lesions Classification Into Eight Classes for ISIC 2019 Using Deep Convolutional Neural Network and Transfer Learning

Classification of Skin Lesions into Seven Classes Using Transfer Learning with AlexNet

Cystatin C as an early marker of acute kidney injury and predictor of mortality in the intensive care unit after acute myocardial infarction.

Unsupervised patterned fabric defect detection using texture filtering and K-means clustering

Serum uric acid and its association with hypertension, early nephropathy and chronic kidney disease in type 2 diabetic patients