Dinesh Dash

A Systematic Review of Wearable Systems for Cancer Detection: Current State and Challenges

Rapid growth of sensor and computing platforms have introduced the wearable systems. In recent years, wearable systems have led to new applications across all medical fields. The aim of this review is to present current state-of-the-art approach in the field of wearable system based cancer detection and identify key challenges that resist it from clinical adoption. A total of 472 records were screened and 11 were finally included in this study. Two types of records were studied in this context that includes 45% research articles and 55% manufactured products. The review was performed per PRISMA guidelines where considerations was given to records that were published or reported between 2009 and 2017. The identified records included 4 cancer detecting wearable systems such as breast cancer (36.3%), skin cancer (36.3%), prostate cancer (18.1%), and multi-type cancer (9%). Most works involved sensor based smart systems comprising of microcontroller, Bluetooth module, and smart phone. Few demonstrated Ultra-Wide Band (i.e. UWB) antenna based wearable systems. Skin cancer detecting wearable systems were most comprehensible ones. The current works are gradually progressing with seamless integration of sensory units along with smart networking. However, they lack in cloud computing and long-range communication paradigms. Artificial intelligence and machine learning are key ports that need to be attached with current wearable systems. Further, clinical inertia, lack of awareness, and high cost are altogether pulling back the actual growth of such system. It is well comprehended that upon sincere orientation of all identified challenges, wearable systems would emerge as vital alternative to futuristic cancer detection.

Restoring virtual backbone of Wireless Sensor Network on sensor failure

In Wireless Sensor Network (WSN) sensor nodes are energy constrained. The network is partitioned into segments due to power depletion or node failure. This requires the deployment of additional nodes / mobile actuators in order to restore network connectivity. In order to reduce the energy depletion of the network and to simplify the design of routing protocols and topology control algorithms Connected Dominating Set (CDS) is preferred as a virtual backbone of the network. The CDS simplifies routing by restricting the main routing path through the dominator nodes only. It is preferred the CDS is small and robust to node failure. In this paper, we consider CDS forms the backbone of the sensor networks. A small size CDS decrease the protocol overhead and energy consumption of the network. Sensors are prone to failure, after the initial construction of CDS. Therefore, restoring the CDS is an important issue in sensor network. In this paper, we study possibilities for CDS restoration after sensor node failure. Various cases are studied to check the possibilities of CDS restoration for the impaired network. Using the observations, we propose efficient algorithms for restoring a CDS for the Wireless Sensor Network after node failure.

Data aggregation in Wireless Sensor Network: An integer linear programming formulation for energy optimization

Wireless Sensor Network is huge number of sensor nodes disposed randomly in severe field to gather data. It monitors physical phenominan changes such as temperature, pressure, humidity, solar radiation, ambident light extra. Sensor nodes are small in size, having less energy, less communication power, less cost and multi functional nodes. This paper we firstly discuss about different types types of network, protocols extra i.e Network such as Direct network, hierarchical network etc and protocols like LEACH, PEGASIS. Next in this paper we deployed sensor nodes randomly in the grid network, We focus on data aggregation in different type of situation known as critical zone, in critical zone(critical grid) senor nodes are heavily loaded with information. Our objective is to maximize the data generation rate of sensor nodes. To got this objective we form different integer linear programming formulation. A optimization tool such as CPLEX IBM ILOG used for optimization and to show critical grids aggregation used MATLAB.

Energy efficient fault tolerant barrier coverage using limited mobility wireless sensors

Sensors are used for tracking objects, surveilling a region/territory, monitoring health and environmental parameters. Coverage problem in sensor network ensures monitoring a given region fully or partially. There are various definitions to measure the quality of coverage depending on its applications. Barrier coverage is a type of coverage, which ensures all paths that cross the boundary of a region intersect at least one sensor sensing region. The goal of the sensors is to detect intruders as they cross the boundary or as they penetrate a protected area. The sensors are dependent on their battery life. Repairing the barrier on sensor failure with minimum energy consumption is the major objective of this paper. A barrier coverage maintenance scheme is proposed to increase the robustness of the network. Barrier is maintained by finding existing alternate barrier or by recreating alternate barrier by moving existing sensors. Simulation results are discussed to evaluate the performance of our algorithm.

Data Gathering from Path-Constrained Mobile Sensors Using Data MULE

In Wireless Sensor Network (WSN), sensor nodes are deployed to sense useful data from environment. Sensors are energy-constrained devices. To prolong the sensor network lifetime, nowadays mobile robots (sometimes referred as data sink, data mules, or data collectors) are used for collecting the sensed data from the sensors. In this environment, sensor nodes may directly transfer their sensed data to the data mules. Sensed data are sometimes time sensitive; therefore, the data should be collected within a predefined period. Hence, depending on the speed of the data mules, the trajectory lengths of the data mules have upper limits. In this paper, an approximation algorithm is proposed for collecting data from the mobile sensors using data mules.

Approximation algorithm for data gathering from mobile sensors

Abstract In Wireless Sensor Network (WSN), sensors are deployed to sense useful data from environment. To prolong the sensor network lifetime in large-scale network, mobile sinks are employed for collecting data from the sensors directly. The major drawback of the system is slow speed of the mobile sinks, which causes long data gathering delay from the sensors. Since, sensors have limited memory and hence it causes buffer overflow in the sensors. Therefore, to avoid buffer overflow the data must be gathered by the mobile sinks within a predefined time interval. Data gathering from mobile sensors using mobile sinks is more challenging problem than data gathering from static sensors. A set of mobile sensors are moving arbitrarily on a set of predefined paths. Our objective is to collect data periodically from all mobile sensors using minimum number of mobile sinks and subsequently the mobile sinks visit a base station ( B S ) for final data delivery. We show that the problem is NP-hard and two approximation algorithms are proposed. We extend the proposed algorithms, where mobile sensors can deliver their sensed data to mobile sink within their circular communication regions and present a recovery algorithm from mobile sink’s failure. We analyze the performance and time complexity of the proposed algorithms.

Design of a low cost and novel Naive Bayes classifier based NodeMCU web server for fever type detection

This study presents a novel NodeMCU based web server to validate cost-effective and smart e-health diagnosis. Proposed work implements a novel Naive Bayes Classifier for automatic fever type detection as a proof of concept. The application is currently accessible on local network. Network users (patients) are first required to submit eight symptoms e.g. “fatigue”, “fever”, “chills”, “sore throat”, “cough”, “headache”, “muscle pain”, and “sneezing” on web portal in form of yes “y” or no “n”. Implied Naive Bayes Classifier engine predicts the probability of occurrence of fever either from flu or common cold as per symptoms provided earlier. The patients are simultaneously diagnosed by a medical practitioner from whom patient wise predictive percentage was received. The probabilistic values were then paired against flu and common cold typed fever patients independently. 22 patients were voluntarily gone through this experiment where p (0.089>0.05) and p (0.068>0.05) were not found to be statistically significant (i.e. no difference between proposed classifier and doctor’s diagnosis) for flu and common cold fever types, respectively.

An integer linear formulation scheme for data collection in wireless sensor network using mobile element (ME)

Nowadays wireless sensor network (WSN) plays an important role in every field of technology. Data collection is a major issue in WSN because the data has to be collected in a way that maximizes life time of WSN, minimizes the energy consumption and gives shortest path in case of mobile sinks. In this paper we divide the whole network into grids. Certain grids are critical (e.g., during a terror strike) in the sense that (a) it may not be possible for mobile elements to enter those grids, and (b) the sensor nodes in those grids are required to sense data at a higher rate due to the critically in the their surrounding environment. A subset of sensor nodes are designated as cache points such that each sensor node must deliver its sensed data to one of the cache points. Now the mobile movable sink has to make a tour schedule and visit each cache point exactly once such that it is able to collect data from all cache points in due time without causing buffer overflow at any of the cache points. We divide the problem into two parts. In the first part, we find cache points such that no cache point is also a critical sensor node and the energy consumption by critical nodes is minimized. In the second part, we find the shortest path among the caches points from the first step. Both problems are known to be NP-Complete and here we give ILP formulations for them.