​

​

 

Projects

​

(1) Interactive Predictive Analytics for Enhancing  Patient Adherence in Remote Health Monitoring

Although Remote Health Monitoring (RHM) systems have shown potential for improving quality of care and reducing healthcare costs, low adherence of human subjects can dramatically degrade the system efficacy. The purpose of this project is to design and

develop an interactive and human-centered framework with new data-driven techniques

and predictive analytics algorithms to enhance patients’ engagement and compliance with

RHM systems. In this project, we propose a novel interactive data-driven system for on demand

data acquisition to enhance human subjects’ adherence in a RHM system. In this

approach, we develop a predictive analytics model that attempts to predict medical

conditions with the least amount of collected data in order to reduce patients’ burden and

improve the adherence. The proposed framework includes a data-driven unit named

Interactive Learning for Data Acquisition (ILDA) for on-demand data collection in order to

enhance the prediction confidence and accuracy as needed. The ILDA automatically

decides whether it still needs to acquire additional information from some specific subjects

or not. It is responsible for evaluating the confidence of the predictions, detecting the need

to interact with specific subjects to collect new data, and identifying what additional

information should be requested from which subjects. The proposed method has been

tested and validated on two different datasets from diabetic and heart disease patients.

​

(2) Remote Health Monitoring Outcome Success Prediction Using Baseline and First Month

Intervention Data

Remote health monitoring (RHM) systems are becoming more widely adopted by clinicians

and hospitals to remotely monitor and communicate with patients while optimizing

clinician time, decreasing hospital costs, and improving quality of care. In the Women’s

heart health study (WHHS), we developed Wanda-cardiovascular disease (CVD), where

participants received healthy lifestyle education followed by six months of technology

support and reinforcement. Wanda-CVD is a smartphone-based RHM system designed to

assist participants in reducing identified CVD risk factors through wireless coaching using

feedback and prompts as social support. Many participants benefitted from this RHM

system. In response to the variance in participants’ success, we developed a framework to

identify classification schemes that predicted successful and unsuccessful participants. We

analyzed both contextual baseline features and data from the first month of intervention

such as activity, blood pressure, and questionnaire responses transmitted through the

smartphone. A prediction tool can aid clinicians and scientists in identifying participants

who may optimally benefit from the RHM system. Targeting therapies could potentially

save healthcare costs, clinician, and participant time and resources. Our classification

scheme yields RHM outcome success predictions with an F-measure of 91.9%, and

identifies behaviors during the first month of intervention that help determine outcome

success. We also show an improvement in prediction by using intervention-based

smartphone data. Results from the WHHS study demonstrates that factors such as the

variation in first month intervention response to the consumption of nuts, beans, and

seeds in the diet help predict patient RHM protocol outcome success in a group of young

Black women ages 25-45.

​

(3) A Robust Remote Health Monitoring and Data Processing System for Rural Area with Limited

Internet Access

​

Emerging technologies such as Body Sensor Networks (BSN) and Remote Health

Monitoring Systems (RHMS) allow for collecting continuous data from patients, providing

clinical interventions to improve patients’ physical and mental health, and to prevent

medically adverse events. Although RHMS have shown promises in reducing healthcare

costs and improving quality of care, designing a robust, flexible and portable RHMS that

can adequately handle technological and environmental restrictions such as data

communication disruptions caused by low-quality infrastructure is by large an open

problem.

In this study we present an end-to-end robust RHMS including body area networks for

monitoring and management of patients with HIV/AIDS. The proposed system includes an

Android application to be installed on mobile devices to collect and preprocess data from

patients, a wireless body area network to handle data communication and transmission,

and a backend cloud-based server coupled with a NoSQL database and web-portal for data

visualization and analysis.

The proposed system addresses a number of serious challenges in RHMS design including

data transmission and synchronization in a body area network with poor network

connection (e.g. in a rural area with poor or unavailable internet connection). The

proposed system is fully implemented and deployed in a large research study involving

600 women living with HIV/AIDS in rural parts of India. The results demonstrate the

effectiveness, reliability, robustness, and accuracy of the proposed RHMS.

(4) Multiple model analytics for adverse event

prediction in remote health monitoring systems

Remote health monitoring systems (RHMS) are gaining an important role in healthcare by

collecting and transmitting patient vital information and providing data analysis and

medical adverse event prediction (e.g. hospital readmission prediction). Reduction in the

readmission rate is typically achieved by early prediction of the readmission based on the

data collected from RHMS, and then applying early intervention to prevent the

readmission. Given the diversity of patient populations and the continuous nature of

patient monitoring, a single static predictive model is insu􀃞cient for accurately predicting

adverse events. To address this issue, we propose a multiple prediction modeling

technique that includes a set of accurate prediction models rather than one single

universal predictor. In this study, we propose a novel analytics framework based on the

physiological data collected from RHMS, advanced clustering algorithms and multiplemodel-

classification. We tested our proposed method on a subset of data collected

through a remote health monitoring system from 600 Heart Failure patients. Our proposed

method provides significant improvements in prediction accuracy and performance over

single predictive models.

​

(5) Context-Aware Data Analytics for Activity Recognition 

 

Remote Health Monitoring Systems (RHMS) are gaining an important role in healthcare by

collecting and transmitting patient information and providing data analytics techniques to

analyze the collected data and extract knowledge. Physical activity recognition and indoor

localization are two of the most important concepts in assistive healthcare, where tracking

the positions, motions and reactions of a patient or elderly is required for medical

observation or accident prevention. In this study, we propose a novel context-aware data

analytics framework to classify and recognize the physical activity based on the signals

received from a worn Smart Watch, the location information of the human subject, and

advanced machine learning algorithms. In this approach, we consider the physical location

of the human subject as contextual information to improve the accuracy of the activity

classification. The hypothesis is that the location information can get involved in classifier

decision making as a prior probability distribution to help improve the accuracy of activity

recognition. The results demonstrate improvements in accuracy and performance of the

activity classification when applying the proposed method compared to conventional

classifications.

​

(6) Remote health monitoring: Predicting outcome success based on contextual features

for cardiovascular disease

Current studies have produced a plethora of remote health monitoring (RHM) systems

designed to enhance the care of patients with chronic diseases. Many RHM systems are

designed to improve patient risk factors for cardiovascular disease, including physiological

parameters such as body mass index (BMI) and waist circumference, and lipid pro_les such

as low-density lipoprotein (LDL) and high-density lipoprotein (HDL). There are several

patient characteristics that could be determining factors for a patient’s RHM outcome

success, but these characteristics have been largely unidenti_ed. In this study, we analyze

results from an RHM system deployed in a six-month Women’s Heart Health study of 90

patients and apply advanced feature selection and machine learning algorithms to identify

patients’ key baseline contextual features and build e􀃠ective prediction models that help

determine RHM outcome success. We introduce Wanda-CVD, a smartphone-based RHM

system designed to help participants with cardiovascular disease risk factors by motivating

participants through wireless coaching using feedback and prompts as social support. We

analyze key contextual features that secure positive patient outcomes in both physiological

parameters and lipid pro_les. Results from the Women’s Heart Health study show that

health threat of heart disease, quality of life, family history, stress factors, social support,

and anxiety at baseline all help predict patient RHM outcome success.

(7) A Framework for Predicting Adherence in

Remote Health Monitoring Systems

Remote health monitoring (RHM) systems have shown potential e􀃠ectiveness in disease

management and prevention. In several studies RHM systems have been shown to reduce

risk factors for cardiovascular disease (CVD) for a subset of the study participants.

However, many RHM study participants fail to adhere to the prescribed study protocol or

end up dropping from the study prior to its completion. In a recent Women’s Heart Health

study of 90 individuals in the community, we developed Wanda-CVD, an enhancement to

our previous RHM system. Wanda-CVD is a smartphone-based RHM system designed to

assist participants to reduce identi_ed CVD risk factors by motivating participants through

wireless coaching using feedback and prompts as social support. Many participants

adhered to the study protocol, however, many did not completely adhere, and some even

dropped prior to study completion. In this study, we present a framework for analyzing

baseline features to predict adherence to prescribed medical protocols that can be applied

to other RHM systems. Such a prediction tool can aid study coordinators and clinicians in

identifying participants who will need further study support, leading potentially to

participants deriving maximal bene_t from the RHM system, potentially saving healthcare

costs, clinician and participant time and resources. We analyze key contextual features that

predict with an accuracy of 85.2% which participants are more likely to adhere to the study

protocol. Results from the Women’s Heart Health study demonstrate that factors such as

perceived health threat of heart disease, and perceived social support are among the

factors that aid in predicting patient RHM protocol adherence in a group of African

American women ages 25-45.

(8) Probabilistic segmentation of time-series

audio signals using Support Vector Machines

To allow health tracking, patient monitoring, and provide timely user interventions, sensor

signals from body sensor networks need to be processed in real-time. Time subdivisions of

the sensor signals are extracted and fed into a supervised learning algorithm, such as

Support Vector Machines (SVM), to learn a model capable of distinguishing di􀃠erent class

labels. However, selecting a short-duration window from the continuous data stream is a

signi_cant challenge, and the window may not be properly centered around the activity of

interest. In this work, we address the issue of window selection from a continuous data

stream, using an optimized SVM-based probability model. To evaluate the e􀃠ectiveness of

our approach, we apply our algorithm to audio signals acquired from a wearable nutritionmonitoring

necklace. Our optimized algorithm is capable of correctly classifying 86.1% of

instances, compared to a baseline of 73% which segments the time-series data with _xedsize

non-overlapping windows, and an exhaustive-search approach with an accuracy of

92.6%.

(9) Improving compliance in remote healthcare

systems through smartphone battery

optimization

Remote health monitoring (RHM) has emerged as a solution to help reduce the cost

burden of unhealthy lifestyles and aging populations. Enhancing compliance to prescribed

medical regimens is an essential challenge to many systems, even those using smartphone

technology. In this study, we provide a technique to improve smartphone battery

consumption and examine the e􀃠ects of smartphone battery lifetime on compliance, in an

attempt to enhance users’ adherence to remote monitoring systems. We deploy WANDACVD,

an RHM system for patients at risk of cardiovascular disease (CVD), using a wearable

smartphone for detection of physical activity. We tested the battery optimization technique

in an in-lab pilot study and validated its e􀃠ects on compliance in the Women’s Heart Health

Study. The battery optimization technique enhanced the battery lifetime by 192% on

average, resulting in a 53% increase in compliance in the study. A system like WANDA-CVD

can help increase smartphone battery lifetime for RHM systems monitoring physical

activity.

(10) Anti-Cheating: Detecting Self-In􀃖icted and

Impersonator Cheaters for Remote Health

Monitoring Systems with Wearable Sensors

In remote health monitoring of patient’s physical activity, ensuring correctness and

authenticity of the received data is essential. Although many activity monitoring systems,

devices and techniques have been developed, preventing patient cheating of an activity

monitor has been a primarily unaddressed challenge across the board. Patients can

manually shake an activity monitor device (sensor) with their hand and watch their physical

activity points or rewards increase, we de􀃕ne this as “self-in􀃖icted” cheating. A second type

of cheating, “impersonator” cheating, is when subjects hand the activity sensor over to a

friend or second party to wear and perform physical activity on their behalf. In this study,

we propose two novel methods based on classi􀃕cation algorithms to address the cheating

problems. The 􀃕rst classi􀃕cation framework improves the correctness of our data by

detecting self-in􀃖icted cheatings. The second technique is an advanced classi􀃕cation

scheme that extracts and learns unique patient-speci􀃕c activity patterns from prior data

collected on a patient to distinguish the true subject from an impersonator. We tested our

proposed techniques on Wanda, a remote health monitoring system used in our Women’s

Heart Health study of 90 African American women at risk of cardiovascular disease. We

were able to distinguish cheating from other physical activities such as walking and

running, as well as other common activities of daily living such as driving and playing video

games. The self-in􀃖icted cheating classi􀃕er achieved an accuracy of above 90% and an AUC

of 99%. The impersonator cheater framework results in an average accuracy of above 90%

and an average AUC of 94%. Our results provide insight into the randomness of cheating

activities, successfully detects cheaters, and attempts to build more context-aware remote

activity monitors that more accurately capture patient activity.

(11) Classi􀃕cation for Big Dataset of Bioacoustic

Signals Based on Human Scoring System and

Arti􀃕cial Neural Network

In this project, we propose a method to improve sound classi􀃕cation performance by

combining signal features, derived from the time-frequency spectrogram, with human

perception. The method presented herein exploits an arti􀃕cial neural network (ANN) and

learns the signal features based on the human perception knowledge. The proposed

method is applied to a large acoustic dataset containing 24 months of nearly continuous

recordings. The results show a signi􀃕cant improvement in performance of the detectionclassi

􀃕cation system; yielding as much as 20% improvement in true positive rate for a

given false positive rate.

(12) Applying machine learning techniques to

recognize arc in vehicle 48 electrical systems

Implementing higher voltages in vehicles like 48V mild hybrid or full-hybrid enables CO2

reduction and weight savings. However, the increase in the voltage demands an accurate

and robust protection system again potential fault conditions. Series arc is one of the fault

conditions which needs to be detected and addressed before the bene_ts of using higher

voltages in vehicle can be fully realized. In this project, an e􀃞cient arc detection algorithm

is presented based on advanced machine learning algorithms to recognize series arcs in

the electrical systems of 48V vehicles. The results demonstrate that the proposed

algorithm recognizes the arcs in the signal with accuracy of more than 99% detection. The

algorithm only uses measured current and therefore, no additional wiring will be added to

the system.

(13) Framework for Domain-Agnostic Gait Cycle

Detection

The utility of wearable sensors for continuous gait monitoring has grown substantially,

enabling novel applications on mobility assessment in healthcare. Existing approaches for

gait cycle detection rely on prede_ned or experimentally tuned platform parameters and

are often platform-speci_c, parameter sensitive, and unreliable in noisy environments with

constrained generalizability. To address these challenges, we introduce Framework1, a

novel framework for reliable and platform-independent gait cycle detection. Framework

o􀃠ers unique features: (1) It leverages physical properties of human gait to learn model

parameters; (2) captured signals are transformed into signal magnitude and processed

through a normalized cross-correlation module to compensate for noise and search for

repetitive patterns without prede_ned parameters; (3) an optimal peak detection algorithm

is developed to accurately _nd strides within the motion sensor data. To demonstrate the

e􀃞ciency of Framework, three experiments are conducted: a clinical study including a

visually impaired group of patients with glaucoma and a control group of healthy

participants; a clinical study involving children with Rett syndrome; and an experiment

involving healthy participants. The performance of Framework is assessed under varying

platform settings and demonstrates to maintain over 93% accuracy under noisy signal,

varying bit resolutions, and changes in sampling frequency. This translates into a recall of

95.3% and a precision of 93.4%, on average. Moreover, Framework can detect strides and

estimate cadence using data from di􀃠erent sensors, with accuracy higher than 95% and it

is robust to random sensor orientations with a recall of 91.5% and a precision of 99.2%, on

average.

(14) Predictive Analytics to Determine the

Potential Occurrence of Genetic Disease and

their Correlation: Osteoporosis and

Cardiovascular Disease

In this project, a Predictive Analytics Model is designed, developed, and validated to

determine the risk of manifesting osteoporosis in later life using big data processing. The

proposed model leverages the novel genetic pleiotropic information in the 1,000 Genome

Project of over 2,500 individuals world-wide. Also, the mutations associated with

osteoporosis and cardiovascular disease are speci_cally analyzed. The study proposes the

automatic histogram clustering as an e􀃠ective and intuitive visualization method for high

dimensional dataset. The results demonstrate a signi_cant correlation between a person’s

regional background and the frequency of occurrence of the 35 Single Nucleotide

Polymorphisms (SNPs) associated with osteoporosis and/or cardiovascular disease (CVD).

Machine learning algorithms, such as Logistic Regression, Adaboost, and KNN are then

applied to predict the occurrence of 7 osteoporosis-related SNPs based on the existing

CVD-related-SNPs input. Finally, the developed model is evaluated using a separate dataset

obtained through A􀃠ymetrix microarray mRNA expression signal values for the speci_c

SNP(s) in individuals with and without osteoporosis.

(15) Multi-label Classi_cation of Single and

Clustered Cervical Cells Using Deep

Convolutional Networks

Cytology-based screening through the Papanicolaou test has dramatically decreased the

incidence of cervical cancer. Convolutional neural networks (CNN) have been utilized to

classify cancerous cervical cytology cells but primarily focused on pre-processed nuclear

details for a single cell and binary classi_cation of normal versus abnormal cells. In this

study, we developed a novel system for multiple label classi_cation with a focus on both

nucleus and cytoplasm of single cells and cell clusters. In this retrospective study, we

digitalized cervical cytology slides from 104 patients. Based upon the Bethesda system, the

established criteria for diagnosing cervical cytology, cells of interest were categorized. With

10-fold cross validation, our CNN algorithm demonstrated 84.5% overall accuracy, 79.1%

sensitivity, 89.5% speci_city for normal versus abnormal. For 3 level classi_cation of

normal, low-grade, and high-grade, CNN demonstrated 76.1% overall accuracy. Results

show promise on the utility of CNNs to learn cervical cytology.

(16) Predicting Glucose Levels in Patients with

Type1 Diabetes Based on Physiological and

Activity Data

Managing blood glucose levels for type 1 diabetes patients is an absolute necessity to

better glycemic control. In this study, we present a predictive model that uses physiological

measurements and physical activity to predict continuous glucose levels and help patients

reduce and prevent hyperglycemia and hypoglycemia exposure, conditions that are

harmful to patient health.

The data of this study includes 4 months of physiological measurements, physical activity,

and nutrition information collected from 93 patients with diabetes using the Medtronic

MiniMed™ 530G insulin delivery system with Enlite™ sensor. After data preprocessing,

missing value imputation, feature extraction, and feature selection, a set of 180 features

were derived to represent the raw data. Then, an appropriate predictive model was

developed based on machine-learning algorithms to predict continuous glucose levels. The

prediction accuracy and error have been calculated to evaluate the performance of the

system. The results demonstrated that the predicted glucose levels closely followed the

actual sensor glucose (SG) values measured by subcutaneous glucose sensor.

(17) A novel fast solving method for targeted

drug-delivery capsules in the gastrointestinal

tract

As an innovative technique without cable connection, targeted drug-delivery capsules

improve diagnostic and therapeutic capabilities in the gastrointestinal (GI) tract. To fast

track targeted drug-delivery capsules in the GI tract, a tracking method based on the

multiple alternating magnetic sources with adaptive adjustment of the excitation intensity

has been investigated. The functional prototype of the tracking system has been

developed. The tracking model between the magnetic 􀃕eld strength and the capsule’s

location has been established, which shows a nonlinear equation group with multiple local

extremum. Particularly, an improved back-propagation (BP) neural network by particle

swarm optimization (PSO) is investigated to solve the tracking problem in real time. The

PSO is introduced at an early stage to optimize the weights and thresholds of the BP neural

network to improve the generalizability and global search ability. Consequently, the

Levenberg-Marquardt (LM) algorithm is used as the learning rule to obtain a higher

accuracy and convergence rate. The performance on the PSO-BP neural network is

experimentally analyzed by comparing it with the standard BP network and the LM-BP

network. The tracking experiments show that the PSO-BP neural network can solve the

tracking problem successfully. The PSO-BP network can get the solution faster than

iterative search algorithms.

(18) E􀃞ciency Optimization of Deep Workout

Recognition with Accelerometer Sensor for a

 

Mobile Environment

Recent advancements in deep learning have created numerous possibilities for real-world

application, among which the recognition of human motions with sensors. We employed a

convolutional neural network (CNN) to process workout motion data to solve the

segmentation and recognition problem. We focused on deploying the network architecture

in a mobile environment characterized by limited resources. Our experimental results

were promising in terms of both segmentation and recognition. Furthermore, we analyzed

the performance correlation between sampling the rate and recognition rate. This result

indicated 55Hz to provide an appropriate amount of information for workout motion. Then

we investigated the computational cost and memory space usage of the hyper-parameter

selection. The experimental results implied that appropriate hyper-parameter selection

could reduce the computational burden on the mobile environment. Subsequently, we

suggest an e􀃞ciency score regarding the computational cost, memory usage, and the

recognition rate. Our suggested e􀃞ciency score shows that our method could 􀃕nd hyperparameters

that o􀃠er minimum loss of accuracy and are computationally inexpensive with

optimal utilization of memory space.

(19) Analyzing the Mutation Frequencies and

Correlation of Genetic Diseases in Worldwide

Populations Using Big Data Processing,

Clustering, and Predictive Analytics

In this study, we utilize Big Data Processing and develop Predictive Analytics Models to

examine and analyze mutations associated with osteoporosis and cardiovascular disease.

The dataset consists of the genomic information of over 2,500 individuals. The genomic

data was collected from all around the world. The data visualization allowed us to see

geographical/regional clustering patterns in the above-mentioned speci􀃕c mutations. The

visualized data clearly shows a high correlation between a person’s regional background

and the occurrence of the 35 single nucleotide polymorphisms (SNPs). The 35 SNPs are

speci􀃕cally associated with osteoporosis and/or cardiovascular disease (CVD). A predictive

analytics model was developed based on machine learning algorithms to predict the risk of

an individual manifesting osteoporosis in later life. The results of this predictive model

con􀃕rmed the links between osteoporosis and Cardiovascular related parameters such as

High-Density Lipoprotein (HDL) and Systolic Blood Pressure (SBP), as determined by the

preceding studies.

(20) A 􀃖exible data-driven comorbidity feature

extraction framework

Disease and symptom diagnostic codes are a valuable resource for classifying and

predicting patient outcomes. In this study, we propose a novel methodology for utilizing

disease diagnostic information in a predictive machine learning framework. Our

methodology relies on a novel, clustering-based feature extraction framework using

disease diagnostic information. To reduce the data dimensionality, we identify disease

clusters using co-occurrence statistics. We optimize the number of generated clusters in

the training set and then utilize these clusters as features to predict patient severity of

condition and patient readmission risk. We build our clustering and feature extraction

algorithm using the 2012 National Inpatient Sample (NIS), Healthcare Cost and Utilization

Project (HCUP) which contains 7 million hospital discharge records and ICD-9-CM codes.

The proposed framework is tested on Ronald Reagan UCLA Medical Center Electronic

Health Records (EHR) from 3041 Congestive Heart Failure (CHF) patients and the UCI 130-

US diabetes dataset that includes admissions from 69,980 diabetic patients. We compare

our cluster-based feature set with the commonly used comorbidity frameworks including

Charlson’s index, Elixhauser’s comorbidities and their variations. The proposed approach

was shown to have signi_cant gains between 10.7-22.1% in predictive accuracy for CHF

severity of condition prediction and 4.65-5.75% in diabetes readmission prediction.

(21) A data-driven feature extraction framework

for predicting the severity of condition of

congestive heart failure patients

In this study, we propose a novel methodology for utilizing disease diagnostic information

to predict severity of condition for Congestive Heart Failure (CHF) patients. Our

methodology relies on a novel, clustering-based, feature extraction framework using

disease diagnostic information. To reduce the dimensionality, we identify disease clusters

using cooccurrence frequencies. We then utilize these clusters as features to predict

patient severity of condition. We build our clustering and feature extraction algorithm

using the 2012 National Inpatient Sample (NIS), Healthcare Cost and Utilization Project

(HCUP) which contains 7 million discharge records and ICD-9-CM codes. The proposed

framework is tested on Ronald Reagan UCLA Medical Center Electronic Health Records

(EHR) from 3041 patients. We compare our cluster-based feature set with another that

incorporates the Charlson comorbidity score as a feature and demonstrate an accuracy

improvement of up to 14% in the predictability of the severity of condition.

(22) Accurate tumor localization and tracking in

radiation therapy using wireless body sensor

networks

Radiation therapy is an e􀃠ective method to combat cancerous tumors by killing the

malignant cells or controlling their growth. Knowing the exact position of the tumor is a

very critical prerequisite in radiation therapy. Since the position of the tumor changes

during the process of radiation therapy due to the patients’ movements and respiration, a

real-time tumor tracking method is highly desirable in order to deliver a su􀃞cient dose of

radiation to the tumor region without damaging the surrounding healthy tissues. In this

study, we develop a novel tumor positioning method based on spatial sparsity. We

estimate the position by processing the received signals from only one implantable RF

transmitter. The proposed method uses a smaller number of sensors compared to

common magnetic transponder-based approaches. The performance of the proposed

method is evaluated in two di􀃠erent cases: (1) when the tissue con_guration is perfectly

determined (acquired beforehand by MRI or CT) and (2) when there are some uncertainties

about the tissue boundaries. The results demonstrate the high accuracy and performance

of the proposed method, even when the tissue boundaries are imperfectly known.

(23) Accurate Localization of In-Body Medical

Implants Based on Spatial Sparsity

Wearable and implantable wireless communication devices have in recent years gained

increasing attention for medical diagnostics and therapeutics. In particular, wireless

capsule endoscopy has become a popular method to visualize and diagnose the human

gastrointestinal tract. Estimating the exact position of the capsule when each image is

taken is a very critical issue in capsule endoscopy. Several approaches have been

developed by researchers to estimate the capsule location. However, some unique

challenges exist for in-body localization, such as the severe multipath issue caused by the

boundaries of di􀃠erent organs, inconsistency of signal propagation velocity and path loss

parameters inside the human body, and the regulatory restrictions on using highbandwidth

or high-power signals. In this study, we propose a novel localization method

based on spatial sparsity. We directly estimate the location of the capsule without going

through the usual intermediate stage of _rst estimating time-of-arrival or received-signal

strength, and then a second stage of estimating the location. We demonstrate the accuracy

of the proposed method through extensive Monte Carlo simulations for radio frequency

emission signals within the required power and bandwidth range. The results show that

the proposed method is e􀃠ective and accurate, even in massive multipath conditions.

(24) Indoor Localization, Tracking and Fall

Detection for Assistive Healthcare Based on

Spatial Sparsity and Wireless Sensor Network

Indoor localization and fall detection are two of the most paramount topics in assistive

healthcare, where tracking the positions and actions of the patient or elderly is required

for medical observation or accident prevention. Most of existing indoor localization

methods are based on estimating one or more location-dependent signal parameters.

However, some challenges caused by the complex scenarios within a closed space

signi_cantly limit the applicability of those existing approaches in an indoor environment,

such as the severe multipath e􀃠ect. In this study, the authors propose a new one-stage,

three-dimensional localization method based on the spatial sparsity in the x-y-z space. The

proposed method is not only able to estimate and track the accurate positions of the

patient, but also capable to detect the falls of the patient. In this method, the authors

directly estimate the location of the emitter without going through the intermediate stage

of TOA or signal strength estimation. The authors evaluate the performance of the

proposed method using various Monte Carlo simulation settings. The results show that the

proposed method is a very accurate even with a small number of sensors and ii very

e􀃠ective in addressing the multi-path issues.

​

(25) Robustness analysis of sparsity-based tumor

localization under tissue con_guration

uncertainty

Knowing the exact position of the tumor is a very critical prerequisite in radiation therapy.

Since the position of the tumor changes because of respiration or patient movements, a

real-time tumor tracking method must be applied during the process of radiation therapy

in order to deliver a su􀃞cient dose of radiation to the tumor region without damaging the

surrounding healthy tissues. In this study, we develop a novel tumor positioning method

based on spatial sparsity and then we investigate the sensitivity of this method to the

uncertainty of tissue con_guration. The proposed method is easier to implement, noniterative,

faster and more accurate compared to common magnetic transponder-based

approaches. The performance of the proposed method is evaluated in two di􀃠erent cases:

(1) when the tissue con_guration is perfectly determined (acquired beforehand by MRI or

CT) and (2) when there are some uncertainties about the tissue boundaries. The results

demonstrate the satisfactory accuracy and high performance of the proposed method,

even when the tissue boundaries are imperfectly known.

​

(26) A novel method for tumor localization and

tracking in radiation therapy

Because the position of a tumor changes during radiation therapy (because of respiration

or patient movements), real-time tumor tracking is necessary during radiation therapy in

order to deliver a su􀃞cient dose of radiation to the tumor without damaging the

​

​

surrounding healthy tissues. In this study, we propose a novel tumor positioning method

based on spatial sparsity. We estimate the position by processing the received signals from

only one implantable RF transmitter. The method is easier to implement, non-iterative,

faster and more accurate compared to common magnetic transponder-based methods.

We evaluate the performance of the proposed method using Monte-Carlo simulation.

​

(27) A novel method for medical implant in-body

localization

Wireless communication medical implants are gaining an important role in healthcare

systems by controlling and transmitting the vital information of the patients. Recently,

Wireless Capsule Endoscopy (WCE) has become a popular method to visualize and

diagnose the human gastrointestinal (GI) tract. Estimating the exact location of the capsule

when each image is taken is a very critical issue in capsule endoscopy. Most of the

common capsule localization methods are based on estimating one or more locationdependent

signal parameters like TOA or RSS. However, some unique challenges exist for

in-body localization due to the complex nature within the human body. In this study, we

propose a novel one-stage localization method based on spatial sparsity in 3D space. In

this method, we directly estimate the location of the capsule (as the emitter) without going

through the intermediate stage of TOA or signal strength estimation. We evaluate the

performance of the proposed method using Monte Carlo simulation with an RF signal

following the allowable power and bandwidth ranges according to the standards. The

results show that the proposed method is very e􀃠ective and accurate even in massive

multipath and shadowing conditions.

​

(28) Spatial sparsity based indoor localization in

wireless sensor network for assistive healthcare

Indoor localization is one of the key topics in the area of wireless networks with increasing

applications in assistive healthcare, where tracking the position and actions of the patient

or elderly are required for medical observation or accident prevention. Most of the

common indoor localization methods are based on estimating one or more locationdependent

signal parameters like TOA, AOA or RSS. However, some di􀃞culties and

challenges caused by the complex scenarios within a closed space signi_cantly limit the

applicability of those existing approaches in an indoor assistive environment, such as the

well-known multipath e􀃠ect. In this study, we develop a new one-stage localization method

based on spatial sparsity of the x-y plane. In this method, we directly estimate the location

of the emitter without going through the intermediate stage of TOA or signal strength

estimation. We evaluate the performance of the proposed method using Monte Carlo

simulation. The results show that the proposed method is (i) very accurate even with a

small number of sensors and (ii) very e_ective in addressing the multi-path issues.

 

​