Detection and Classification of Diabetic Retinopathy Using DCNN and BSN Models

Diabetes is associated with many complications that could lead to death. Diabetic retinopathy, a complication of diabetes, is difficult to diagnose and may lead to vision loss. Visual identification of micro features in fundus images for the diagnosis of DR is a complex and challenging task for clinicians. Because clinical testing involves complex procedures and is time-consuming, an automated system would help ophthalmologists to detect DR and administer treatment in a timely manner so that blindness can be avoided. Previous research works have focused on image processing algorithms, or neural networks, or signal processing techniques alone to detect diabetic retinopathy. Therefore, we aimed to develop a novel integrated approach to increase the accuracy of detection. This approach utilized both convolutional neural networks and signal processing techniques. In this proposed method, the biological electro retinogram (ERG) sensor network (BSN) and deep convolution neural network (DCNN) were developed to detect and classify DR. In the BSN system, electrodes were used to record ERG signal, which was pre-processed to be noise-free. Processing was performed in the frequency domain by the application of fast Fourier transform (FFT) and mel frequency cepstral coefficients (MFCCs) were extracted. Artificial neural network (ANN) classifier was used to classify the signals of eyes with DR and normal eye. Additionally, fundus images were captured using a fundus camera, and these were used as the input for DCNN-based analysis. The DCNN consisted of many layers to facilitate the extraction of features and classification of fundus images into normal images, non-proliferative DR (NPDR) or early-stage DR images, and proliferative DR (PDR) or advanced-stage DR images. Furthermore, it classified NPDR according to microaneurysms, hemorrhages, cotton wool spots, and exudates, and the presence of new blood vessels indicated PDR. The accuracy, sensitivity, and specificity of the ANN classifier were found to be 94%, 95%, and 93%, respectively. Both the accuracy rate and sensitivity rate of the DCNN classifier was 96.5% for the images acquired from various hospitals as well as databases. A comparison between the accuracy rates of BSN and DCNN approaches showed that DCNN with fundus images decreased the error rate to 4%.     

Enhanced photocatalytic activity of sulfur doped TiO2 for the decomposition of phenol: A new insight into the bulk and surface modification

Sulfur ion (S⁶⁺) was doped into the anatase TiO₂ prepared by sol–gel method (SG-TiO₂) using sulfur powder as a sulfur source (S-TiO₂) and its photoreactivity was probed for the degradation of phenol under UV/solar light illumination. The S-TiO₂ and SG-TiO₂ were characterized by PXRD, UV–vis DRS, FTIR, SEM, XPS, BET and PL techniques. It was observed that S⁶⁺ ion was incorporated into the TiO₂ crystal lattice at Ti⁴⁺ lattice site and the sulfur on the surface gets modified to ₄SO²⁻${{\text{SO}}_4}^{2-}$ due to the heat treatment under atmospheric conditions. The high photocatalytic activity of S-TiO₂ compared to SG-TiO₂ is attributed to the surface modification of sulfur as sulfate which plays a crucial role in trapping electrons. S-TiO₂ shows significant increase in the surface area, reduced crystallite size, increased surface acidity, visible light absorption and prolonged lifetime of the photogenerated charge carriers. Hole scavengers like potassium iodide and tertiary butanol suggested the surface degradation mechanism rather than the bulk degradation pathway. Addition of oxidizing agents to the degradation reaction did not show any enhancement in the degradation rates since the presence of ₄SO²⁻${{\text{SO}}_4}^{2-}$ on the TiO₂ surface itself acts as the efficient electron trapping centers. Both trapping and detrapping of the electron takes place more efficiently at ₄SO²⁻${{\text{SO}}_4}^{2-}$ centers. The enhanced activity of S-TiO₂ is attributed to the synergistic effect between S⁶⁺ dopant with surface modified ₄SO²⁻${{\text{SO}}_4}^{2-}$.     

Correlation between structural properties and gas sensing characteristics of SnO2 based gas sensors

Undoped polycrystalline tin oxide sintered in the temperature range 500–1000 °C has been comprehensively characterized with respect to its response to CO, methane and H₂. Results obtained at an operating temperature of 300 °C show that increasing the sintering temperature leads to a gradual increase in CO sensitivity which reaches a maximum after sintering at 800 °C.     

The effect of grain size and composition on the fracture toughness of Ti–Al–Nb alloys undergoing stress-induced martensitic transformation

The effect of grain size and composition on the fracture toughness of Ti–Al–Nb alloys in β solution-treated condition was investigated. The fracture toughness of the alloys was found to increase with an increase in grain size initially, reach a maximum and subsequently decrease with further increase in grain size. This trend was attributed primarily to the effect of grain size on the enhancement of fracture toughness due to stress-induced martensitic transformation (SIMT) at the crack tip, which in turn can be related to the effect of grain size on trigger stress for SIMT. Alloys containing higher Al and Nb showed a higher toughness for the same grain size, which was also explained in terms of effect of composition on the trigger stress.     

An indigenously developed electronic control system for Langmuir-Blodgett film deposition set-up

An indigenous and simple electronic control system for Langmuir-Blodgett (LB) film deposition set-up has been developed. This set-up consists of a microstepping circuit to drive the stepper motors with precision and smooth motion, essential for controlled movement of the barriers and substrate in the LB set-up. Linear variable differential transformer (LVDT)-based displacement measuring device has been developed and used to measure the surface pressure of the monolayer material spread on the water surface. A control program is written which incorporates all operational modes required to drive the set-up and to acquire the datain situ using a set of user-friendly commands. This control set-up has been successfully used to plot the pressure-area isotherm of various amphiphilic compounds such as ferric stearate, zinc arachidate etc. and for deposition of ordered LB films of ferric stearate.     

Effect of manganese and cobalt doping on conductivity of ZnO based varistors: a study by complex plane modulus

Effect of manganese and cobalt doping (0.50 mole%) on electrical properties of ZnO based varistors has been studied using complex plane modulus analysis. It is found that total resistivity of Mn doped sample is more as compared to that of Co doped sample. This has been ascribed to existence of Mn in variable valence states viz. Mn²⁺, Mn³⁺ and Mn⁴⁺ which promotes hopping conduction leading to increase in the conductivity as compared to Co doped sample, in which Co exists predominantly in +3 state with traces of Co²⁺ or Co⁺⁴ states. This accounts for its less conductivity. Mechanism of conduction is the same for grains and grainboundaries.     

Development of nano cerium oxide incorporated aluminium alloy sacrificial anode for marine applications

Aluminium-zinc alloy sacrificial anodes are extensively used for cathodic protection. The performance of the sacrificial anodes can be significantly improved by incorporation of microalloying elements in the aluminium matrix. In the present work nano cerium oxide particles of different concentrations, ranging from 0 to 1 wt% were incorporated for activating and improving the performance of the anode. The electrochemical test results revealed the increased efficiency of the anode. The electrochemical impedance spectroscopy revealed the information that the presence of nano cerium oxide in the anode matrix caused effective destruction of the passive alumina film, which facilitated enhancement of galvanic performance of the anode. Moreover, the biocidal activity of cerium oxide prevented the bio accumulation considerably which enables the anodes to be used in aggressive marine conditions.     

Development of a PCR assay for the detection of nifH and nifD genes in indigenous photosynthetic bacteria

Molybdenum (Mo) nitrogenases consist of two components: dinitrogenase reductase (encoded by nifH) and the dinitrogenase or MoFe protein (encoded by nifDK). Nitrogenase enzyme of photosynthetic bacteria is responsible for hydrogen production. Therefore, primers were designed for the nitrogenase gene only. In this study, two primers (ND and NH) were designed after comparative genomic analysis of nifH and nifD gene sequences from public databases. The designed primers were used for the amplification of nifH and nifD genes to detect nitrogenase genes in photosynthetic bacteria. Initial detection was done using a monoplex Polymerase Chain Reactions (PCRs) followed by optimization of the PCR protocols. Subsequently, a duplex PCR was designed for amplification and detection of nifH and nifD genes in indigenous photosynthetic bacteria. Evaluation of the duplex PCR on six samples isolated from Palm Oil Mill Effluent (POME) showed that only four isolates contained both the nifH and nifD genes, indicating that these isolates were potential hydrogen-producing bacteria. PCR detection provides a rapid and efficient pre-identification of potential photosynthetic bacterial hydrogen producers.