Towards Health Safety: Conflict Detection and Resolution

Learn more »

linklab
logo3

Preclude: Conflict Detection in Textual Health Advice


Preclude Example A sample conflict between a pair of advice coming from a general weight loss app and WebMD

Health Apps Result in Conflicts Semantic decomposition of textual health advice

With the rapid digitalization of the health sector, people often turn to mobile apps and online health websites for health advice. Health advice generated from different sources can be conflicting as they address different aspects of health (e.g., weight loss, diet, disease) or as they are unaware of the context of a user (e.g., age, gender, physiological condition). Conflicts can occur due to lexical features, (such as, negation, antonyms, or numerical mismatch) or can be conditioned upon time and/or physiological status. We formulate the problem of finding conflicting health advice and develop a comprehensive taxonomy of conflicts. While a similar research area in the natural language processing domain explores the problem of textual contradiction identification, finding conflicts in health advice poses its own unique lexical and semantic challenges. These include large structural variation between text and hypothesis pairs, finding conceptual overlap between pairs of advice, and inference of the semantics of an advice (i.e., what to do, why and how). Hence, we develop Preclude, a novel semantic rule-based solution to detect conflicting health advice derived from heterogeneous sources utilizing linguistic rules and external knowledge bases. As our solution is interpretable and comprehensive, it can guide users towards conflict resolution too. We evaluate Preclude using 1156 real advice statements covering 8 important health topics that are collected from smart phone health apps and popular health websites. Preclude results in 90% accuracy and outperforms the accuracy and F1 score of the baseline approach by about 1.5 times and 3 times, respectively

[pdf] [link]




Preclude2 : Personalized conflict detection in heterogeneous health applications


Dependency among multiple activities Drug usage guidelines often poses different dependencies among activities of daily life


Sources of conflicts Conflicts may occur among health advice suggested by health applications and drug usage guidelines

Conflicting health information is a primary barrier of self-management of chronic diseases. Increasing number of people now rely on mobile health apps and online health websites to meet their information needs and often receive conflicting health advice from these sources. This problem is more prevalent and severe in the setting of multi-morbidities. In addition, often medical information can be conflicting with regular activity patterns of an individual. In this work, we formulate the problem of finding conflicts in heterogeneous health applications including health websites, health apps, online drug usage guidelines, and daily activity logging applications. We develop a comprehensive taxonomy of conflicts based on the semantics of textual health advice and activities of daily living. Finding conflicts in health applications poses its own unique lexical and semantic challenges. These include large structural variation between pairs of textual advice, finding conceptual overlap between pairs of advice, inferring the semantics of an advice (i.e., what to do, why and how) and activities, and aligning activities suggested in advice with the activities of daily living based on their underlying dependencies and polarity. Hence, we develop Preclude2, a novel semantic rule-based solution to detect conflicts in activities and health advice derived from heterogeneous sources. Preclude2 utilizes linguistic rules and external knowledge bases to infer advice. In addition, Preclude2 considers personalization and context-awareness while detecting conflicts. We evaluate Preclude2 using 1156 real advice statements covering 8 important health topics, 90 online drug usage guidelines, 1124 online disease specific health advice covering 34 chronic diseases, and 2 activity datasets. The evaluation is personalized based on 34 real prescriptions. Preclude2 detects direct, conditional, sub-typical, quantitative, and temporal conflicts from 2129 advice statements with 0.91, 0.83, 0.98, 0.85 and 0.98 recall, respectively. Overall, it results in 0.88 recall for detecting inter advice conflicts and 0.89 recall for detecting activity–advice conflicts. We also demonstrate the effects of personalization and context-awareness in conflict detection from heterogeneous health applications.

[pdf] [link] [Dataset]



A Corpus of Drug Usage Guidelines Annotated with Type of Advice


Sample drug usage guidelines An excerpt from the drug usage guideline document of Warfarin. Text underlined in blue, green, and red indicate advice related to drug administration, food interaction, and pregnancy, respectively.

Adherence to drug usage guidelines for prescription and over-the-counter drugs is critical for drug safety and effectiveness of treatment. Drug usage guideline documents contain advice on potential drug-drug interaction, drug-food interaction, and drug administration process. Current research on drug safety and public health indicates patients are often either unaware of such critical advice or overlook them. Categorizing advice statements from these documents according to their topics can enable the patients to find safety critical information. However, automatically categorizing drug usage guidelines based on their topic is an open challenge and there is no annotated dataset on drug usage guidelines. To address the latter issue, this paper presents (i) an annotation scheme for annotating safety critical advice from drug usage guidelines, (ii) an annotation tool for such data, and (iii) an annotated dataset containing drug usage guidelines from 90 drugs. This work is expected to accelerate further release of annotated drug usage guideline datasets and research on automatically filtering safety critical information from these textual documents.

[pdf] [Dataset]



MedRem: An Interactive Medication Reminder and Tracking System on Wrist Devices


MedRem System Architecture of MedRem

Medication adherence is pivotal for effective health outcomes. One of the main reasons behind poor medication adherence is forgetfulness, and reminder systems are often used in addressing the problem. This paper presents MedRem, a novel medication reminder and tracking system on wearable wrist devices. The system is handy and interactive, and it is enriched with several useful features. To address the limitations of the tiny display size of the wrist devices, MedRem incorporates speech recognition and text-to-speech features along with clever interface design. Users interact with the system using voice commands as well as using the display available on the device. A dictionary based training approach is used on top of the state of the art speech recognition systems to reduce the errors in recognizing the commands from the users. The system is evaluated for both native and non-native English speakers. The error rates for recognizing voice commands are 6.43% and 20.9% for the native and the non-native speakers, respectively, when a off-the-shelf speech recognition system is used. MedRem reduces the error to nearly zero for both types of users through a dictionary based training approach. On average, only 1.25 and 15 training commands are required to achieve this performance for the native and the non-native speakers, respectively.

[pdf]



Detection of Chronic Kidney Disease and Selecting Important Predictive Attributes


CKD Cost-accuracy trade-off for CKD detection with different predictive attributes

Chronic kidney disease (CKD) is a major public health concern with rising prevalence. In this study we consider 24 predictive parameters and create a machine learning classifier to detect CKD. We evaluate our approach on a dataset of 400 individuals, where 250 of them have CKD. Using our approach we achieve a detection accuracy of 0.993 according to the F1-measure with 0.1084 root mean square error. This is a 56% reduction of mean square error compared to the state of the art (i.e., the CKD-EPI equation: a glomerular filtration rate estimator). We also perform feature selection to determine the most relevant attributes for detecting CKD and rank them according to their predictability. We identify new predictive attributes which have not been used by any previous GFR estimator equations. Finally, we perform a cost-accuracy tradeoff analysis to identify a new CKD detection approach with high accuracy and low cost.

[pdf]



Harmony: A Hand Wash Monitoring and Reminder System using Smart Watches


Hand washing protocol Handwashing guidelines recommended by World Health Organization

Hand hygiene compliance is extremely important in hospitals, clinics and food businesses. Caregivers’ compliance with hand hygiene is one of the most effective tools in preventing healthcare associated infections (HAIs) in hospitals and clinics. In food businesses, hand hygiene compliance is essential to prevent food contamination, and thus food borne illness. Washing hands properly is the cornerstone of hand hygiene. However, the hand wash compliance rate by the workers (care givers, waiters, chefs, food processors and so on) is not up to the mark. Monitoring hand wash compliance along with a reminder system increases the compliance rate significantly. Quality of a hand wash is also important which can be achieved by washing hands in accordance with standard guidelines. In this paper, we present Harmony, a hand wash monitoring and reminder system that monitors hand wash events and their quality, provides real time feedback, reminds the person of interest when he/she is required to wash hands, and stores related data in a server for further use. Worker worn smart watches are the key components of Harmony that can differentiate hand wash gestures from other gestures with an average accuracy of about 88%. Harmony is robust, scalable, and easy to install, and it overcomes most of the problems of existing related systems.

[pdf]



EyePhy: Detecting Dependencies in Cyber-Physical System Apps due to Human-in-the-Loop


EyePhy Runtime Dependency Detection by EyePhy

EyePhy Installation time conflict detection on high level parameters

As app based paradigms are becoming popular, millions of apps are developed from many domains including energy, health, security, and entertainment. The US FDA expects that there will be 500 million smart phone users downloading healthcare related apps by 2015. Many of these apps are Cyber-Physical System (CPS) apps. In addition to sensing, communication, and computation, they perform interventions to control human physiological parameters, which can cause dependency problems as multiple interventions of multiple apps can increase or decrease each others effects, some of which can be harmful to the user. Such dependency problems occur mainly because each app is unaware about how other apps work and when an app performs an intervention to control its target parameters, it may affect other physiological parameters without even knowing it. We present EyePhy, a system that detects dependencies across interventions by having a closer eye on the physiological parameters of the human in the loop. To do that, EyePhy uses a physiological simulator HumMod that can model the complex interactions of the human physiology using over 7800 variables. EyePhy reduces app developers’ efforts in specifying dependency metadata compared to state of the art solutions and offers personalized dependency analysis for the user. We demonstrate the magnitude of dependencies that arise during multiple interventions in a human body and the significant ability of detecting these dependencies using EyePhy.

[pdf] [link]




Publications

  • S. Preum, M. Mondol, M. Ma, H. Wang, and J. Stankovic, Preclude: Conflict Detection in Textual Health Advice, Percom, March 2017. [pdf] [link]

  • S. Preum, M. Mondol, M. Ma, H. Wang, and J. Stankovic, Preclude2: Personalized and Context-aware Conflict Detection in Heterogeneous Health Applications, Journal of Pervasive and Mobile Computing, Vol. 42, Dec. 2017, pp. 226-247. [pdf] [link] [Dataset]

  • S. Preum, M. Rizwan Parvez, K. Chang, and J. Stankovic, A Corpus of Drug Usage Guidelines Annotated with Type of Advice, The International Conference on Language Resources and Evaluation (LREC), May 2018. [pdf] [Dataset]

  • S. Preum, M. Mondol, M. Ma, H. Wang, and J. Stankovic, Demo Abstract: Conflict Detection in Online Textual Health Advice, Information Processing in Sensor Networks (IPSN), April 2017. [link]

  • M. Mondol, I. Emi and J. Stankovic, MedRem: An Interactive Medication Reminder and Tracking System on Wrist Devices, Wireless Health, October 2016 [pdf]

  • A. Salekin and J. Stankovic, Detection of Chronic Kidney Disease and Selecting Important Predictive Attributes, Int. Conf. on Healthcare Informatics, October 2016. [pdf]

  • M. Mondol and J. Stankovic, Harmony: A Hand Wash Monitoring and Reminder System using Smart Watches, Mobiquitous, July 2015. [pdf] [link]

  • S. Munir, M. Ahmed, and J. Stankovic, EyePhy: Detecting Dependencies in Cyber-Physical System Apps due to Human-in-the-Loop, Mobiquitous, July 2015. [pdf] [link]



People

John A. Stankovic (PI)

BP America Professor

Director, Link Lab

Department of Computer Science, University of Virginia

Email: jas9f@virginia.edu


Sarah Masud Preum

PhD Candidate

Department of Computer Science, University of Virginia

Email: sp9cx@virginia.edu


Md Abu Sayeed Mondol

PhD Candidate

Department of Computer Science, University of Virginia

Email: mm5gg@virginia.edu


Asif Salekin

PhD Candidate

Department of Computer Science, University of Virginia

Email: as3df@virginia.edu


Ifat Afrin Emi

PhD Candidate

Department of Computer Science, University of Virginia

Email: iae4qb@virginia.edu


Meiyi Ma

PhD Candidate

Department of Computer Science, University of Virginia

Email: mm5tk@virginia.edu


Mohsin Yousuf Ahmed

PhD Candidate

Department of Computer Science, University of Virginia

Email: mya5dm@virginia.edu