Mobile Settings Research Articles

Memory-Aware Active Learning in Mobile Sensing Systems.

We propose a novel active learning framework for activity recognition using wearable sensors. Our work is unique in that it takes limitations of the oracle into account when selecting sensor data for annotation by the oracle. Our approach is inspired by human-beings' limited capacity to respond to prompts on their mobile device. This capacity constraint is manifested not only in the number of queries that a person can respond to in a given time-frame but also in the time lag between the query issuance and the oracle response. We introduce the notion of mindful active learning and propose a computational framework, called EMMA, to maximize the active learning performance taking informativeness of sensor data, query budget, and human memory into account. We formulate this optimization problem, propose an approach to model memory retention, discuss the complexity of the problem, and propose a greedy heuristic to solve the optimization problem. Additionally, we design an approach to perform mindful active learning in batch where multiple sensor observations are selected simultaneously for querying the oracle. We demonstrate the effectiveness of our approach using three publicly available activity datasets and by simulating oracles with various memory strengths. We show that the activity recognition accuracy ranges from 21% to 97% depending on memory strength, query budget, and difficulty of the machine learning task. Our results also indicate that EMMA achieves an accuracy level that is, on average, 13.5% higher than the case when only informativeness of the sensor data is considered for active learning. Moreover, we show that the performance of our approach is at most 20% less than the experimental upper-bound and up to 80% higher than the experimental lower-bound. To evaluate the performance of EMMA for batch active learning, we design two instantiations of EMMA to perform active learning in batch mode. We show that these algorithms improve the algorithm training time at the cost of a reduced accuracy in performance. Another finding in our work is that integrating clustering into the process of selecting sensor observations for batch active learning improves the activity learning performance by 11.1% on average, mainly due to reducing the redundancy among the selected sensor observations. We observe that mindful active learning is most beneficial when the query budget is small and/or the oracle's memory is weak. This observation emphasizes advantages of utilizing mindful active learning strategies in mobile health settings that involve interaction with older adults and other populations with cognitive impairments.

One-Bit Feedback Exponential Learning for Beam Alignment in Mobile mmWave

Efficient beam alignment in wireless networks capable of supporting device mobility is currently one of the major challenges in mmWave communications. In this context, we formulate the beam-alignment problem via the adversarial multi-armed bandit (MAB) framework, which copes with arbitrary network dynamics including non-stationary or adversarial components. Building on the well known exponential weights algorithm (EXP3) and by exploiting the structure and sparsity of the mmWave channel, we propose a modified (MEXP3) policy that requires solely one-bit of feedback information (reducing the amount of exchanged data during the beam-alignment process). Our MEXP3 comes with optimal theoretical guarantees in terms of asymptotic regret. Moreover, for finite horizons, our regret upper-bound is tighter than that of the original EXP3 suggesting better performance in practice. We then introduce an additional modification that accounts for the temporal correlation between successive beams and propose another beam-alignment policy. Our numerical results demonstrate that our beam-alignment policies outperform existing ones with respect to the regret but also to the outage, throughput and delay in typical mobile mmWave settings.

Driving Access to Care: Use of Mobile Units for Urine Specimen Collection During the Coronavirus Disease-19 (COVID-19) Pandemic

Patients with substance use disorders (SUD) are at increased risk of both coronavirus disease-19 complications as well as exacerbations of their current conditions due to social distancing and isolation. Innovations that provide increased access to support substance use disorder patients may mitigate long-term sequelae associated with continued or renewed drug use. To improve patient access during the coronavirus disease-19 pandemic, we deployed a mobile unit to enable access to urine drug testing where needed for patients suffering from substance use disorder. Over a 3-week pilot program, 54 patients received urine drug testing across 5 providers and 8 zip codes. The mobile unit was cost-effective, demonstrating a volume-dependent 19% lower cost compared to pre-coronavirus disease-19 patient service centers in a similar geographic region. The mobile unit was well-received by patients and providers with an average of 9 out of 10 satisfaction scores and allowed for access to urine drug testing for 67% patients who would not have received testing during this time frame. No statistically significant differences were found in substance use positivity rates in comparison to pre-coronavirus disease findings; however, some shifts in use included higher rates of fentanyl and opioid positivity and reductions in tetrahydrocannabinol and cocaine use in the mobile collections setting. Deployment of mobile collection services during the coronavirus disease-19 pandemic has shown to be an effective mechanism for supporting patients suffering from substance use disorder, allowing for access to care of this often stigmatized, vulnerable population.

Diagnostic Accuracy of Rapid Assessment of Avoidable Blindness: A Population-based Assessment

To evaluate the diagnostic accuracy of rapid assessment of avoidable blindness (RAAB). Population-based diagnostic accuracy study. A total of 2145 (95.3%, 2145/2250) subjects aged 50 years and older who participated in the RAAB survey were included. All the recruited participants underwent ophthalmic examination according to the RAAB protocol and then were reexamined with instruments in a mobile eye clinic set up in a village center on the same day. Examination in the mobile clinic included standardized visual acuity (VA) tests using logMAR charts, refraction, slit-lamp biomicroscopy, and dilated fundal examination with a binocular indirect ophthalmoscope. Blindness and economic blindness were defined as VA in the better-seeing eye <3/60 and <6/60, respectively. Visual impairment (VI) was defined as VA <6/18 in the better eye. The primary cause of blindness and VI was defined according to the cause of VI in the participant's better eye. The sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and area under the curve (AUC) of receiver operating characteristics of RAAB for detection of blindness and the principal causes of VI. A total of 1816 subjects (84.7%), including 686 men (37.8%) and 1130 women (62.2%), underwent ophthalmic examination in the mobile eye clinic. The mean (±standard deviation) age was 64.4 ± 9.6 years. The sensitivities, specificities, AUC, PLR, and NLR of RAAB were 90.3%, 99.3%, 0.948, 124.0, and 0.10, respectively, for detection of blindness (presenting visual acuity, PVA <3/60); 89.5%, 98.7%, 0.940, 69.2, and 0.11, respectively, for detection of economic blindness (PVA <6/60); and 90.3%, 97.7%, 0.940, 38.7, and 0.10, respectively, for detection of VI (PVA <6/18). The sensitivities, specificities, AUC, PLR, and NLR were 90.5%, 98.1%, 0.943, 48.1, and 0.10; and 60.4%, 98.7%, 0.796, 46.4, and 0.40 for detection of VI (PVA <6/18) owing to cataract and refractive error, respectively. The diagnostic performances of RAAB were high for detecting the prevalence of blindness, VI, and VI owing to cataract.

Increasing knowledge of HIV status in a country with high HIV testing coverage: Results from the Botswana Combination Prevention Project.

IntroductionAchieving widespread knowledge of HIV-positive status is a crucial step to reaching universal ART coverage, population level viral suppression, and ultimately epidemic control. We implemented a multi-modality HIV testing approach to identify 90% or greater of HIV-positive persons in the Botswana Combination Prevention Project (BCPP) intervention communities.MethodsBCPP is a cluster-randomized trial designed to evaluate the impact of combination prevention interventions on HIV incidence in 30 communities in Botswana. Community case finding and HIV testing that included home and targeted mobile testing were implemented in the 15 intervention communities. We described processes for identifying HIV-positive persons, uptake of HIV testing by age, gender and venue, characteristics of persons newly diagnosed through BCPP, and coverage of knowledge of status reached at the end of study.ResultsOf the 61,655 eligible adults assessed in home or mobile settings, 13,328 HIV-positive individuals, or 93% of the estimated 14,270 positive people in the communities were identified through BCPP. Knowledge of status increased by 25% over the course of the study with the greatest increases seen among men (37%) as compared to women (19%) and among youth aged 16–24 (77%) as compared to older age groups (21%). Although more men were tested through mobile than through home-based testing, higher rates of newly diagnosed HIV-positive men were found through home than mobile testing.ConclusionsEven when HIV testing coverage is high, additional gains can be made using a multi-modality HIV testing strategy to reach different sub-populations who are being missed by non-targeted program activities. Men and youth can be reached and will engage in community testing when services are brought to places they access routinely.

OMS: Opportunistic mules for short latency data collection in smart cities

SummaryIn smart city scenarios, data collected by sensors are required to be conveyed to central controllers for processing. Thus, efficient data collection mechanisms considering the urban environment characteristics should take place. To that end, the data mules approach is proved to be efficient. However, being applied to urban environment leads to high data delivery latency due to the characteristics of this environment. Motivated by the need to reduce the collection latency of such an approach, we propose to provide each road segment with a data mule to create a process of distributed data collection. Our scheme leverages opportunistic inter‐vehicular communications for assigning the mule role to adequate vehicles. The simulation in realistic mobility settings reveals that our scheme can provide more than 80% of the area of interest with data mules in less than 1 minute. In addition, the data collection latency can be reduced from 24 hours to few minutes.

Calculation of Stress-strain State of the Microtunneling Shield Housing

The work reviews the strength calculation of the microtunneling shield housing shell and bed plates of hydraulic cylinders of the mobile set. Based on the requirements for maximum loads on the shield, the analytical calculation of the housing shell thickness and basic design parameters of the tunneling shield bed plate was made. The bending moments and normal forces effecting on the housing were defined in analytical way. The calculation method was offered for the required solid shell of the shield housing considering the allowed values of its deformation due to the necessary construction clearance between the inner surface of the housing shell and mounted support. The numerical modelling of the stress-strain state of the tunneling shield housing was made using the finite-element method on the basis of solid model with due regard to the design peculiarities of the tunneling shield. The numerical modelling confirms the results of analytical calculation of design parameters of the tunneling shield.

On the Coloniality of “New” Mega‐Infrastructure Projects in East Africa

AbstractThis article responds to a preference for short‐term history in research on the infrastructure turn by engaging with the longue durée of East Africa’s latest infrastructure scramble. It traces the history of LAPSSET in Kenya and the Central Corridor in Tanzania, revealing the coloniality of new and improved transport infrastructure along both corridors. This exercise demonstrates how the spatial visions and territorial plans of colonial administrators get built in to new infrastructure and materialise in ways that serve the interests of global capital rather than peasant and indigenous peoples being promised more modern, prosperous futures. The article concludes by suggesting that a focus on the longue durée also reveals uneven patterns of mobility and immobility set in motion during the colonial scramble for Africa and reinforced after independence. These “colonial moorings” are significant as they shape political reactions to new mega‐infrastructure projects today and constrain the emancipatory potential of infrastructure‐led development.

LETSRP: A Secure Routing Protocol for Manets

Reliable link between the nodes playsvital role during the transmission of data in the routing protocols of Mobile Ad-hoc Networks (MANETs). In this research paper, a routing protocol Link named Expiration Time based Routing Protocol (LETSRP)is proposed which uses Winternitz One-time Signature Scheme to checkthe authentication of transmitted data in the network. The packet sending node calculates maximum LET (MaxLET), minimum LET (MinLET) and average LET (AvgLET) using greedy algorithm in our proposed solution. The number of sent packets dependson the available bandwidth. Various steps in implementation are repeated until all the packets reach the destination node. Simulations are done with variable number of nodes between10-20 nodes with mobility setting from 10 m/s to 20 m/s.

A Novel Mobile Tool (Somatomap) to Assess Body Image Perception Pilot Tested With Fashion Models and Nonmodels: Cross-Sectional Study.

BackgroundDistorted perception of one’s body and appearance, in general, is a core feature of several psychiatric disorders including anorexia nervosa and body dysmorphic disorder and is operative to varying degrees in nonclinical populations. Yet, body image perception is challenging to assess, given its subjective nature and variety of manifestations. The currently available methods have several limitations including restricted ability to assess perceptions of specific body areas. To address these limitations, we created Somatomap, a mobile tool that enables individuals to visually represent their perception of body-part sizes and shapes as well as areas of body concerns and record the emotional valence of concerns.ObjectiveThis study aimed to develop and pilot test the feasibility of a novel mobile tool for assessing 2D and 3D body image perception.MethodsWe developed a mobile 2D tool consisting of a manikin figure on which participants outline areas of body concern and indicate the nature, intensity, and emotional valence of the concern. We also developed a mobile 3D tool consisting of an avatar on which participants select individual body parts and use sliders to manipulate their size and shape. The tool was pilot tested on 103 women: 65 professional fashion models, a group disproportionately exposed to their own visual appearance, and 38 nonmodels from the general population. Acceptability was assessed via a usability rating scale. To identify areas of body concern in 2D, topographical body maps were created by combining assessments across individuals. Statistical body maps of group differences in body concern were subsequently calculated using the formula for proportional z-score. To identify areas of body concern in 3D, participants’ subjective estimates from the 3D avatar were compared to corresponding measurements of their actual body parts. Discrepancy scores were calculated based on the difference between the perceived and actual body parts and evaluated using multivariate analysis of covariance.ResultsStatistical body maps revealed different areas of body concern between models (more frequently about thighs and buttocks) and nonmodels (more frequently about abdomen/waist). Models were more accurate at estimating their overall body size, whereas nonmodels tended to underestimate the size of individual body parts, showing greater discrepancy scores for bust, biceps, waist, hips, and calves but not shoulders and thighs. Models and nonmodels reported high ease-of-use scores (8.4/10 and 8.5/10, respectively), and the resulting 3D avatar closely resembled their actual body (72.7% and 75.2%, respectively).ConclusionsThese pilot results suggest that Somatomap is feasible to use and offers new opportunities for assessment of body image perception in mobile settings. Although further testing is needed to determine the applicability of this approach to other populations, Somatomap provides unique insight into how humans perceive and represent the visual characteristics of their body.

Shoplifting in mobile checkout settings: cybercrime in retail stores

Purpose Retailers are implementing technology-enabled mobile checkout processes in their stores to improve service quality, decrease labor costs and gain operational efficiency. These new checkout processes have increased customer convenience primarily by providing them autonomy in sales transactions in that store employee interventions play a reduced role. However, this autonomy has the unintended consequence of altering the checks and balances inherent in a traditional employee-assisted checkout process. Retailers, already grappling with shoplifting, with an estimated annual cost of billions of dollars, fear that the problem may be exacerbated by mobile checkout and concomitant customer autonomy. The purpose of this paper is to understand the effect of mobile checkout processes in retail stores on cybercrime in the form of shoplifting enabled by a technology transformed the retail environment. Design/methodology/approach The authors conducted an online survey of a US sample recruited from a crowdsourced platform. The authors test a research model that aims to understand the factors that influence the intention to shoplift in three different mobile checkout settings − namely, smartphone checkout settings, store-provided mobile device checkout settings, and employee-assisted mobile checkout settings − and compare it with a traditional fixed location checkout setting. Findings The authors found that, in a smartphone checkout setting, intention to shoplift was driven by experiential beliefs and peer influence, and experiential beliefs and peer influence had a stronger effect for prospective shoplifters when compared to experienced shoplifters; in a store-provided mobile devices checkout setting, experiential beliefs had a negative effect on shoplifters’ intention to shoplift and the effect was weaker for prospective shoplifters when compared to experienced shoplifters. The results also indicated that in an employee-assisted mobile checkout setting, intention to shoplift was driven by experiential beliefs and peer influence, and experiential beliefs had a stronger effect for prospective shoplifters when compared to experienced shoplifters. Originality/value This study is the among the first, if not first, to examine shoplifters’ intention to shoplift in mobile checkout settings. We provide insights into how those who may not have considered shoplifting in less favorable criminogenic settings may change their behavior due to the autonomy provided by mobile checkout settings and also provide an understanding of the shoplifting intention for both prospective and experienced shoplifters in different mobile checkout settings.

P1460Seismocardiographic measurement of systolic time intervals in left ventricular dysfunction: validation with transthoracic echo

Abstract Background Systolic time intervals (STI) measured using Transthoracic Echo (TTE) have been shown to be sensitive indicators of contractile performance, but TTE is limited to a hospital setting. Recent advances in technology have enabled the simultaneous measurement of Electrocardiograms (ECG) and Seismocardiograms (SCG) using wearable devices. However, the use SCG to measure STIs has not been validated in a mobile setting, or on pathological subjects. Purpose To test the validity of an automated algorithm for measuring STIs using a wearable device recording ECG and SCG, in subjects with varying degrees of left ventricular dysfunction. Methods 179 Patients with suspected left ventricular dysfunction were assessed using TTE performed by a Cardiologist. Patients were simultaneously fitted with a novel wearable device worn on a chest strap, recording SCG and single-lead ECG data, while the cardiologist measured the pre-ejection period (PEP) and left ventricular ejection time (LVET) with the patient in a supine position. Of these subjects 29 (16.2%) were diagnosed with Dilated cardiomyopathy (DCM), and 109 (60.8%) had Ischemic Heart disease. The SCG and ECG data recorded on the wearable device was then analysed using a peak-detection algorithm, which detected the Q,R,S points on the ECG, and then determined the 4 most prominent peaks in the SCG signal corresponding to each R-peak. Heart rate, patient's age, gender and SCG time intervals were then used in a DecisionTree algorithm to determine the values of PEP and LVET, which were were then compared against those determined by the cardiologist using TTE. Results The correlation coefficient (r2) between PEP calculated using TTE, and the values obtained from the algorithm analyzing SCG data was 0.92 while the mean error was 7.47%. The r2 between the LVET calculated using the TTE and the algorithm was 0.75, while the mean error was 8.53% (p-value&lt;0.001 for all cases). Results All Subjects With IHD Without IHD With DCM Without DCM Number of Subjects 178 109 69 29 149 PEP (r2) 0.92 0.89 0.94 0.88 0.91 PEP (% age error) 7.47 7.50 7.42 6.99 7.56 LVET (r2) 0.75 0.81 0.66 0.55 0.83 LVET (% age error) 8.53 6.69 11.4 20.42 6.16 DCM, Dilate Cardiomyopathy; IHD, Ischemic Heart Disease. ECG + SCG signals for a DCM patient Conclusion The algorithm-derived STIs measured by SCG correlate well with those measured by TTE across most patient groups, including those with Dilated Cardiomyopathy and Ischemic Heart Disease, opening prospects for continuous remote monitoring of STIs in a mobile setting. Acknowledgement/Funding Fourth Frontier Technologies

ЭКСПЕРИМЕНТАЛЬНОЕ ИССЛЕДОВАНИЕ ЛИНЕЙНОГО ЭЛЕКТРОПРИВОДА ДВУХНОЖЕВОГО РЕЖУЩЕГО АППАРАТА ЖАТКИ ЗЕРНОУБОРОЧНОГО КОМБАЙНА

The article is devoted to the development and research in the work of an experimental installation of an electric drive of a two-blade cutting apparatus of a combine harvester harvester. The layout of the combine harvester elements is given, the cutting apparatus of which is supplied with electric drive based on two-sided flat linear asynchronous motors. The kinematic diagram of the linear electric drive of the cutting apparatus was compiled, its efficiency was evaluated in laboratory conditions. It was revealed that for efficient operation of the cutting apparatus, the operation of drive induction motors in the mode of forced oscillations is preferable. This eliminates the position sensors and it is possible to adjust the parameters of the oscillations of the knife strips according to the technological process of cutting. A study was also made of the linear electric drive of the cutting apparatus of the combine harvester in field conditions. For this purpose, a thyristor switching and control unit for linear asynchronous electric motors, designed and tested under laboratory conditions, which allows optimally adjusting the parameters of the oscillating process of knife strips for specific harvesting conditions, was connected to a mobile diesel generator set. In this case, the time dependences of the currents of inductors of two-sided flat linear asynchronous motors and the amplitude of oscillations of the blade strips of the cutting apparatus were experimentally obtained. When comparing the current consumed by inductors of linear motors by the maximum amplitude and pulse duration, it was concluded that when the linear electric drive of the cutting unit operates in a steady state, the current difference is no more than 8,8 % compared to the previously obtained theoretical data. The combination of the experimental time dependence of the oscillation amplitude of the knife strips of the cutting apparatus with the theoretical one allows us to conclude that the difference in the amplitude of oscillations does not go beyond 8,4 %. All this testifies to the adequacy of the previously obtained mathematical model.

Application of Nanotechnology For Cleaning of Teeth And Shining It

This research includes the using of Powder Of Snail Shell’s Rostellaria species for cleaning, and remove Staining of Teeth as well as shining it.As an alternative of a Conventional pumice when dentists and technicians’ in dental laboratories. It was powder processing on two types : first, it provides as a small size particles, with added of some rate of florid.it used by dentists, to clean the natural teeth. Second, it provides as a large size particles, it used by owners of dental laboratories, to clean the fixed and mobile sets of teeth. And making it easier for the possibility of the application of this powder as anano powder. they contain of a three-layer cover, resistant to external conditions. That’s making constant hardness powder, it range between (4 – 3), Unlike other types of pumice Available in the market. In addition to the containment of this crust of the fixed elements and chemical compositions such as : ( Cao, Sio2, AL2o3, Fe2o3, Mgo, So3 ), In the proportions : (52.70, 20.40, 0.68, 0.44, 1.5, 0.28 ) respectively.

Integration of air pollution data collected by mobile sensors and ground-based stations to derive a spatiotemporal air pollution profile of a city

ABSTRACTAir pollution has become a serious environmental problem causing severe consequences in our ecology, climate, health, and urban development. Effective and efficient monitoring and mitigation of air pollution require a comprehensive understanding of the air pollution process through a reliable database carrying important information about the spatiotemporal variations of air pollutant concentrations at various spatial and temporal scales. Traditional analysis suffers from the severe insufficiency of data collected by only a few stations. In this study, we propose a rigorous framework for the integration of air pollutant concentration data coming from the ground-based stations, which are spatially sparse but temporally dense, and mobile sensors, which are spatially dense but temporally sparse. Based on the integrated database which is relatively dense in space and time, we then estimate air pollutant concentrations for given location and time by applying a two-step local regression model to the data. This study advances the frontier of basic research in air pollution monitoring via the integration of station and mobile sensors and sets up the stage for further research on other spatiotemporal problems involving multi-source and multi-scale information.

RETRACTED ARTICLE: Simulator considering modeling and performance evaluation for high-performance computing of collaborative-based mobile cloud infrastructure

In recent years, as IT technology has progressed, mobile devices have been created that enable various manual tasks to be automated and portable. A variety of mobile devices has computing, storage, and Internet capabilities and can handle many tasks. When miniaturized mobile devices perform tasks that require a large amount of computing resources due to limited computing and storage, there is a delay in operation and a non-operation state. Therefore, collaborative-based mobile cloud infrastructure (MCI) research is being conducted to provide computing services composed of mobile devices. Computation off-loading studies have been conducted for MCI’s high-performance computing, but it is difficult to build various mobile infrastructures and verify algorithm performance. In addition, performance verification is performed in a predetermined MCI environment or is carried out through small-scale test equipment. This causes waste of time, cost, and manpower for constructing the environment. Various studies have been conducted for this purpose, but there is a difficulty in performance verification and analysis since only the results are displayed or outputted in text form. In this paper, we propose a mobile cloud infrastructure simulator (MCIS) for computing off-loading, resource management, mobile deployment, and mobile information for MCI. MCIS enables user tasks, resource allocation methods, and various mobile device performance settings. In addition, visualization of the operating state makes it easy to analyze the performance of the user, and it is possible to grasp the problems that occur during operation.

Factors Affecting Switching Behavior of Mobile Set Users: The Case of Khulna City, Bangladesh

This study leads to explore the important factors affecting the behaviors of switching mobile sets and people's attitudes toward particular features. The questionnaire which is provided to gather data regarding the study contains 16 items measured on a five-point Likert scale. The data were analyzed by using frequency analysis and descriptive statistics. The results indicate that problem with network, design, price consideration and problem in power supply, etc. has greater impact on switching behavior of mobile set users. And it also shows the factors which are less considered by the mobile handset users that includes: using new one as well as making a profit by selling it and gift from others. It is assumed that this research will enable the future researchers to work in an expanded field.

Mobile searching versus online searching: differential effects of paid search keywords on direct and indirect sales

Paid search advertising is a critical source of traffic for sellers on e-commerce platforms. As the mobile Internet gains prominence as the main channel for reaching consumers, sellers need to know how to select and bid on keywords in online versus mobile search settings. This study compares the differential effects of paid search keywords in online and mobile settings on both direct and indirect sales, along with the moderating roles of keyword specificity and keyword cost. Two field studies and two lab experiments consistently show that a paid keyword in a mobile setting can prompt higher direct sales than it does in an online setting, but it generates lower indirect sales in general. Moreover, while keyword cost attenuates this positive relationship between mobile searching and direct sales, keyword specificity and keyword cost attenuate this negative relationship between mobile searching and indirect sales. Empirical evidence also reveals the sales and return on investment of paid search keywords in mobile and online settings. This study enriches previous literature by detailing new insights into the distinct settings of mobile and online search advertising.

Efficient Deep Network Architectures for Fast Chest X-Ray Tuberculosis Screening and Visualization

Automated diagnosis of tuberculosis (TB) from chest X-Rays (CXR) has been tackled with either hand-crafted algorithms or machine learning approaches such as support vector machines (SVMs) and convolutional neural networks (CNNs). Most deep neural network applied to the task of tuberculosis diagnosis have been adapted from natural image classification. These models have a large number of parameters as well as high hardware requirements, which makes them prone to overfitting and harder to deploy in mobile settings. We propose a simple convolutional neural network optimized for the problem which is faster and more efficient than previous models but preserves their accuracy. Moreover, the visualization capabilities of CNNs have not been fully investigated. We test saliency maps and grad-CAMs as tuberculosis visualization methods, and discuss them from a radiological perspective.

A Hybrid Structure Dual-Path Step-Down Converter With 96.2% Peak Efficiency Using 250-m$\Omega$ Large-DCR Inductor

A dual-path step-down converter (DPDC) is presented for achieving high power efficiency in the mobile power management ICs (PMICs). Adopting a hybrid structure using one inductor and one flying capacitor, the proposed DPDC supplies a load current via two parallel paths, relieved an intrinsic problem of the conventional buck converter (CBC) topology, which is a significant power loss from a large DCR of the inductor ( ${R}_{\mathrm {DCR}}$ ). Therefore, DPDC achieves a high power efficiency and thus also reduces the heating problem, which is another critical issue in the mobile set. Moreover, DPDC can shrink the volume of the PMIC set with a low manufacturing cost by alleviating an ${R}_{\mathrm {DCR}}$ specification of the inductor. In this paper, although a 250 $\text{m}\Omega $ of large ${R}_{\mathrm {DCR}}$ inductor is used for our measurements, a 96.2% of peak efficiency was achieved and the power loss of total parasitic resistances can be reduced to up to 30% of that of CBC. Moreover, according to our measurement plots, it is verified that DPDC achieves the efficiency notably higher not only in a wide load current ( $I_{\mathrm {LOAD}}$ ) range but also in a wide conversion ratio ( $V_{\mathrm {OUT}}/V_{\mathrm {IN}}$ ) range, compared to CBC.