The WIISEL project developed a wireless insole system to assess fall risk in elderly individuals in their home and community environments. The system collects gait data from insoles and analyzes parameters related to fall risk. It aims to allow for remote and quantitative assessment of fall risk, measuring activity and mobility under daily living conditions. The project was coordinated by CETEMMSA and funded by the European Commission over 41 months with a budget of 3.9 million euros and 8 partners from 6 countries. Validation studies showed the system can be useful as a research tool for studying fall risk and as a clinical tool for long-term monitoring of fall risk in home and community settings.
Telehealthcare for older people: barriers to large-scale roll-outsMaged N. Kamel Boulos
Kamel Boulos MN. Telehealthcare for older people: barriers to large-scale roll-outs (Round table: Use of technologies to promote healthy aging and improve disability). In: Proceedings of the 1st Barcelona Conference on Healthy Aging (University of Barcelona), Barcelona, Spain, 14-15 November 2013 (invited presentation). URL: http://www.healthyageingbarcelona.com/speakers8.html
This document discusses the use of mobile devices like PDAs and smartphones for public health work. It provides examples of software tools that have been used for data collection and medical protocols on mobile devices. These include EpiSurveyor for epidemiological surveys, Satellite Forms for application design, and Voxiva for large-scale projects. The document also discusses considerations for mobile health projects, like optimizing for small screens and taking advantage of connectivity. It announces a scholarship program to review literature on handheld computers in healthcare.
This document provides an overview of the Trauma Audit and Research Network (TARN). TARN collects data on severely injured patients from hospitals across the UK to support clinical audit and improve trauma care standards. It has a database of over 200,000 cases. TARN aims to analyze trauma management, provide comparative performance statistics to clinicians and hospitals, and identify areas for potential research. Participation has grown to over 50% of trauma-receiving hospitals. TARN also focuses on pediatric trauma cases through a separate initiative called TARNlet.
Computational Disease Management with Wearable DevicesPetteriTeikariPhD
Machine Learning modelling of disease trajectory with deep
learning and/or Gaussian Processes.
Alternative download link:
https://www.dropbox.com/s/jg73ymvkenx8rv4/computational_disease_management.pdf?dl=0
ARTIFICIAL NEURAL NETWORKING.
FIRST STEP TO KNOWLEDGE IS TO KNOW THAT we are ignorant
Knowledge in medical field is characterized by uncertanity and vagueness
Historically as well as currently this fact remains a motivation for the development of medical decision support system are based on fuzzy logics
Greek philosopher visualized a basic model of brain function as early as 300 bc
Till date nervous system is not completely understood to human kind.
Reducing Harm from Falls NZMJ final paper 2 DecemberCarmela Petagna
This document summarizes the Health Quality & Safety Commission's three-year program to reduce harm from falls, with an initial focus on reducing falls in hospitals. It discusses the serious consequences of falls, especially hip fractures, for older people. It describes the Commission's "adaptive approach" of promoting evidence-based interventions for providers to choose from, rather than imposing a single bundled approach. It outlines how targeted measurement of risk assessment and care planning practices led to significant nationwide reductions in hip fractures from falls and falls reported as serious adverse events.
1) The document discusses the use of artificial intelligence in orthodontics, including applications like automated cephalometric analysis, skeletal classification, predicting orthodontic treatment needs, and 3D tooth segmentation.
2) AI technologies like convolutional neural networks, artificial neural networks, and deep learning are being used in these orthodontic applications.
3) While AI is proving accurate and can help practitioners make decisions faster, limitations include cost, data protection concerns, and ensuring AI systems do not replace human clinicians for serious medical decisions.
Telehealthcare for older people: barriers to large-scale roll-outsMaged N. Kamel Boulos
Kamel Boulos MN. Telehealthcare for older people: barriers to large-scale roll-outs (Round table: Use of technologies to promote healthy aging and improve disability). In: Proceedings of the 1st Barcelona Conference on Healthy Aging (University of Barcelona), Barcelona, Spain, 14-15 November 2013 (invited presentation). URL: http://www.healthyageingbarcelona.com/speakers8.html
This document discusses the use of mobile devices like PDAs and smartphones for public health work. It provides examples of software tools that have been used for data collection and medical protocols on mobile devices. These include EpiSurveyor for epidemiological surveys, Satellite Forms for application design, and Voxiva for large-scale projects. The document also discusses considerations for mobile health projects, like optimizing for small screens and taking advantage of connectivity. It announces a scholarship program to review literature on handheld computers in healthcare.
This document provides an overview of the Trauma Audit and Research Network (TARN). TARN collects data on severely injured patients from hospitals across the UK to support clinical audit and improve trauma care standards. It has a database of over 200,000 cases. TARN aims to analyze trauma management, provide comparative performance statistics to clinicians and hospitals, and identify areas for potential research. Participation has grown to over 50% of trauma-receiving hospitals. TARN also focuses on pediatric trauma cases through a separate initiative called TARNlet.
Computational Disease Management with Wearable DevicesPetteriTeikariPhD
Machine Learning modelling of disease trajectory with deep
learning and/or Gaussian Processes.
Alternative download link:
https://www.dropbox.com/s/jg73ymvkenx8rv4/computational_disease_management.pdf?dl=0
ARTIFICIAL NEURAL NETWORKING.
FIRST STEP TO KNOWLEDGE IS TO KNOW THAT we are ignorant
Knowledge in medical field is characterized by uncertanity and vagueness
Historically as well as currently this fact remains a motivation for the development of medical decision support system are based on fuzzy logics
Greek philosopher visualized a basic model of brain function as early as 300 bc
Till date nervous system is not completely understood to human kind.
Reducing Harm from Falls NZMJ final paper 2 DecemberCarmela Petagna
This document summarizes the Health Quality & Safety Commission's three-year program to reduce harm from falls, with an initial focus on reducing falls in hospitals. It discusses the serious consequences of falls, especially hip fractures, for older people. It describes the Commission's "adaptive approach" of promoting evidence-based interventions for providers to choose from, rather than imposing a single bundled approach. It outlines how targeted measurement of risk assessment and care planning practices led to significant nationwide reductions in hip fractures from falls and falls reported as serious adverse events.
1) The document discusses the use of artificial intelligence in orthodontics, including applications like automated cephalometric analysis, skeletal classification, predicting orthodontic treatment needs, and 3D tooth segmentation.
2) AI technologies like convolutional neural networks, artificial neural networks, and deep learning are being used in these orthodontic applications.
3) While AI is proving accurate and can help practitioners make decisions faster, limitations include cost, data protection concerns, and ensuring AI systems do not replace human clinicians for serious medical decisions.
Shap Analysis Based Gastric Cancer DetectionIRJET Journal
This document proposes a novel deep learning framework to detect gastric cancer from endoscopic images of the stomach. The framework uses a patch-based analysis where features are extracted from image patches and evaluated for cancer risk. A bag-of-features technique is then applied to the extracted features from selected patches for analysis. Experimental results show the proposed framework can effectively and efficiently detect gastric cancer from images and identify minute lesions. It achieves higher accuracy than other models using the same dataset. The framework is also more accurate than existing methods for gastric cancer detection.
EdgeFall: a promising cloud-edge-end architecture for elderly fall careIJECEIAES
Elder citizens face sudden fall, which can lead to injuries of both destructive and non-virulent. These sudden falls are later more precarious than diseases like heart attack, blood sugar, blood pressure because these can be untreated for a lengthy time which can lead to death. Elder citizen who experiences a precipitous fall, carry out their communal life narrowed. Therefore, a shrewd and adequate anti-fallen system is required for aiding elderly health care, specifically to those who live individually. So, it can identify and anticipate a precipitous fall through appropriate human activity recognition. In this study, we have suggested an end-edge-cloud based wearable EdgeFall architecture for elderly care. We have performed simulation setups to clarify the query of why we need such a strategy, and its validity. We have achieved maximum 91.87% accuracy with 1.6% false alarm rate (FAR). These empirical results indicate the superiority of using tightly couple multiple information for recognizing human activity. We can accomplish a low FAR with an enhanced accuracy. We can observe that our proposed end-edge-cloud based architecture can reduce the execution time to millisecond range (ms) of 14.16 to 15.74. This work serves as the starting mark for future related research activities.
EYE DISEASE IDENTIFICATION USING DEEP LEARNINGIRJET Journal
This document presents a deep learning model for identifying eye diseases from images. The model was trained on datasets of five different eye conditions - conjunctivitis, cataracts, uveitis, bulging eyes, and crossed eyes. The model uses a convolutional neural network architecture with several convolutional and pooling layers. It achieves 96% accuracy on single-eye images and 92.31% accuracy on two-eye images. The authors conclude the model is effective and cost-efficient at classifying common eye diseases and recommending users seek treatment from ophthalmologists when needed.
The SENSACTION-AAL project aimed to assist older adults in maintaining independent mobility and reducing fall injuries through physical activity interventions using wearable sensors. It developed three main applications: a virtual trainer for home rehabilitation exercises providing audio feedback, a smart monitor for tracking daily mobility, and a remote assistant for detecting falls and alerting caregivers. An initial clinical trial with 18 older adults found the system feasible and that it increased adherence to exercises, awareness, and reduced falls. However, further validation is still needed including improved methods for simulating falls versus detecting real falls.
A FALL DETECTION SMART WATCH USING IOT AND DEEP LEARNINGIRJET Journal
The document describes a proposed fall detection smartwatch system using IoT and deep learning. It aims to enable smartwatches and algorithms to detect falls in smart homes. The proposed system, IMEFD-ODCNN, uses data collection, preprocessing, feature extraction using SqueezeNet, parameter tuning using SSO, and classification using SSOA-VAE. Video frames are preprocessed and features extracted before the SSOA-VAE classifier identifies falls. If a fall is detected, an alert is sent to the patient and caregiver for immediate assistance. The system aims to remotely monitor elderly people and help doctors treat patients by providing health data and history.
This document summarizes a study on preferred IoT technologies for tracking aged patients with non-communicable diseases (NCDs) in Malaysia. The study aims to analyze the current wearable health system scenario and preferences based on variables like age and gender. Literature reviews covered IoT and healthcare, wearable health devices including activity trackers and smartwatches. Research methodology involved a cross-sectional survey of 450 patients aged 50-70 across three Malaysian states. The study recommends strengthening technological capabilities and guidance to help prevent and treat NCDs in elderly patients.
This work aims to provide a practical guide to assist students of Computer Science
courses and related fields to conduct a systematic literature review. The steps proposed
in this paper to conduct a systematic review were extracted from a technical report
published by the researcher Bárbara Kitchenham [1] and arranged in a more objective
format, in order to make information more accessible and practical, especially for those
who are having their first contact with this technique.
PREDICTION OF COVID-19 USING MACHINE LEARNING APPROACHESIRJET Journal
This document summarizes a research paper that used machine learning models to predict the spread of COVID-19. The researchers used various machine learning algorithms like SVM, random forest, decision tree, and linear regression on COVID-19 case data. SVM had the highest error in predictions, while random forest and decision tree performed best with lowest error. The models were developed using Python and deployed on cloud platforms. The study aimed to accurately predict COVID-19 trends to help governments respond better to the pandemic.
The ALZCARE system aims to improve healthcare for elderly patients with dementia. It consists of three main components: 1) a mobile screening system that allows non-physicians to administer cognitive tests; 2) a clinical information system for physicians to track patient details, evaluations, and recommendations; and 3) an optional patient tracking system using GPS to monitor patients prone to wandering. The systems were designed based on user input and follow clinical guidelines. They are currently undergoing field testing and evaluation in Greece and Albania to assess their ability to screen for cognitive impairment and support patient management.
Social Distancing Detection, Monitoring and Management Using OpenCVIRJET Journal
This document proposes a system to detect social distancing violations using computer vision and deep learning algorithms. The system would identify individuals in video frames using a YOLOv3 model, calculate distances between detected individuals, and classify the risk level based on social distancing guidelines. It transforms frames into a bird's eye view to standardize distance measurements. The proposed system aims to help monitor social distancing and slow the spread of COVID-19 by identifying groups that are too close together. It achieved 92.8% precision in social distancing classification during testing.
This document summarizes the transition from clinical information systems to health grids and the future of health research infrastructure. It discusses trends like rising populations in Asia, increasing resource scarcity, and the need for multidisciplinary and open collaboration. Health grids are presented as enabling virtual collaborations across institutions. Key areas like medical imaging, computational models, and genomic medicine are highlighted. Adoption challenges and requirements like reliable, usable infrastructure are also summarized.
NEW CORONA VIRUS DISEASE 2022: SOCIAL DISTANCING IS AN EFFECTIVE MEASURE (COV...IRJET Journal
The document describes a proposed real-time system to monitor social distancing using computer vision and deep learning techniques. The system would use a camera to detect individuals and calculate distances between them in order to identify instances where social distancing guidelines are breached. When a breach is detected, an audio-visual cue would be emitted to alert individuals without identifying or saving personal data. The system aims to help reduce the spread of COVID-19 while respecting privacy and avoiding overreach. It outlines the technical approach including camera calibration, region of interest definition, object detection using YOLOv3, distance calculation techniques, and system architecture at a high level.
Report-Fog Based Emergency System For Smart Enhanced Living EnvironmentKEERTHANA M
Report-An ambient assisted-living emergency system exploits cloud and fog computing, an outdoor positioning mechanism, and emergency and communication protocols to locate activity-challenged individuals.
IRJET- A Survey on Vision based Fall Detection TechniquesIRJET Journal
This document reviews different vision-based fall detection systems that have been developed using computer vision and image processing techniques. It discusses how vision-based systems work by capturing images or videos using cameras and then analyzing the footage using algorithms to classify events as falls or non-falls. The document also examines some of the challenges of vision-based approaches, such as effects of lighting and background objects, and how newer techniques like convolutional neural networks have helped improve accuracy of fall detection.
This document provides a summary of a project report submitted by four students - Naina Sangole, Chetana Nimje, Prachi Dhawale, and Sejal Meshram - for their Bachelor of Computer Application degree. The report describes the development of a "Covid-19 Management System" project to allow users to register for Covid tests and vaccines online, track Covid cases, and access test reports digitally. The report includes sections acknowledging those who assisted with the project, certifying completion of the project, tables of contents, the project abstract, introduction, literature review on existing systems and their limitations, objectives of the proposed system, screenshots of the system, coding details, and conclusions.
The document is a project report submitted by 4 students (Naina Sangole, Chetana Nimje, Prachi Dhawale, Sejal Meshram) for their BCA degree. The report describes a "Covid-19 Management System" project aimed at developing a system to help track Covid testing, register users for vaccines, track Covid cases, and provide Covid reports online. The report includes an introduction describing the need for the system, literature review on existing systems, proposed system details including analysis, design, and screenshots, and coding details of the developed system.
A DEVICE FOR AUTOMATIC DETECTION OF ELDERLY FALLSIRJET Journal
The document presents a device for automatically detecting falls in elderly individuals. It uses an accelerometer to measure changes in acceleration along three axes and determine body position. When a fall is detected based on acceleration thresholds, the GPS receiver pinpoints the location and a GSM modem sends a text message notification. The system aims to promptly detect falls to reduce injuries and allow for timely medical assistance. It discusses related work on fall detection techniques using sensors like accelerometers and pose estimation. The proposed system design uses an ESP WiFi controller, GPS and GSM modules, accelerometer, and other components to detect falls, track location, and alert caregivers via SMS. It aims to help elderly individuals live independently safely.
The Power of Sensors in health & healthcareD3 Consutling
In a series of reports we explore key digital health trends and related opportunities for technology companies, healthcare providers and patients-consumers. We take both an international and Flemish perspective, the latter based on interviews with local stakeholders. In this report we focus on sensor-based applications.
STANLEY Healthcare received Frost & Sullivan's 2017 Global Company of the Year Award for its performance in the senior living industry. The company provides a full suite of solutions for senior living communities, including emergency call systems, wander management, staff security, and fall prevention. It also offers 24/7 customer support. STANLEY is working on innovative solutions like its Smart Resident Room concept using sensors to continuously monitor residents. Its visionary approach helps address challenges around technology adoption and respecting residents' dignity. STANLEY's strong performance and loyal customer base earned it the Company of the Year recognition.
This document describes a research project that developed a healthcare chatbot to provide medical information and advice to users from their homes. The chatbot uses machine learning algorithms like decision trees and logistic regression trained on medical datasets. It allows users to describe their symptoms and receives a predicted diagnosis along with the probability and potential future symptoms. For emergencies, it recommends expert doctors. The goal is to enable self-diagnosis and determine if hospital visits are needed, saving time. Evaluation showed the chatbot model achieved 95.52% accuracy in identifying illnesses based on symptom inputs. Future work aims to improve the chatbot and add translation features for wider accessibility.
More Related Content
Similar to WIISEL Final Report - 1- Publishable Report Final
Shap Analysis Based Gastric Cancer DetectionIRJET Journal
This document proposes a novel deep learning framework to detect gastric cancer from endoscopic images of the stomach. The framework uses a patch-based analysis where features are extracted from image patches and evaluated for cancer risk. A bag-of-features technique is then applied to the extracted features from selected patches for analysis. Experimental results show the proposed framework can effectively and efficiently detect gastric cancer from images and identify minute lesions. It achieves higher accuracy than other models using the same dataset. The framework is also more accurate than existing methods for gastric cancer detection.
EdgeFall: a promising cloud-edge-end architecture for elderly fall careIJECEIAES
Elder citizens face sudden fall, which can lead to injuries of both destructive and non-virulent. These sudden falls are later more precarious than diseases like heart attack, blood sugar, blood pressure because these can be untreated for a lengthy time which can lead to death. Elder citizen who experiences a precipitous fall, carry out their communal life narrowed. Therefore, a shrewd and adequate anti-fallen system is required for aiding elderly health care, specifically to those who live individually. So, it can identify and anticipate a precipitous fall through appropriate human activity recognition. In this study, we have suggested an end-edge-cloud based wearable EdgeFall architecture for elderly care. We have performed simulation setups to clarify the query of why we need such a strategy, and its validity. We have achieved maximum 91.87% accuracy with 1.6% false alarm rate (FAR). These empirical results indicate the superiority of using tightly couple multiple information for recognizing human activity. We can accomplish a low FAR with an enhanced accuracy. We can observe that our proposed end-edge-cloud based architecture can reduce the execution time to millisecond range (ms) of 14.16 to 15.74. This work serves as the starting mark for future related research activities.
EYE DISEASE IDENTIFICATION USING DEEP LEARNINGIRJET Journal
This document presents a deep learning model for identifying eye diseases from images. The model was trained on datasets of five different eye conditions - conjunctivitis, cataracts, uveitis, bulging eyes, and crossed eyes. The model uses a convolutional neural network architecture with several convolutional and pooling layers. It achieves 96% accuracy on single-eye images and 92.31% accuracy on two-eye images. The authors conclude the model is effective and cost-efficient at classifying common eye diseases and recommending users seek treatment from ophthalmologists when needed.
The SENSACTION-AAL project aimed to assist older adults in maintaining independent mobility and reducing fall injuries through physical activity interventions using wearable sensors. It developed three main applications: a virtual trainer for home rehabilitation exercises providing audio feedback, a smart monitor for tracking daily mobility, and a remote assistant for detecting falls and alerting caregivers. An initial clinical trial with 18 older adults found the system feasible and that it increased adherence to exercises, awareness, and reduced falls. However, further validation is still needed including improved methods for simulating falls versus detecting real falls.
A FALL DETECTION SMART WATCH USING IOT AND DEEP LEARNINGIRJET Journal
The document describes a proposed fall detection smartwatch system using IoT and deep learning. It aims to enable smartwatches and algorithms to detect falls in smart homes. The proposed system, IMEFD-ODCNN, uses data collection, preprocessing, feature extraction using SqueezeNet, parameter tuning using SSO, and classification using SSOA-VAE. Video frames are preprocessed and features extracted before the SSOA-VAE classifier identifies falls. If a fall is detected, an alert is sent to the patient and caregiver for immediate assistance. The system aims to remotely monitor elderly people and help doctors treat patients by providing health data and history.
This document summarizes a study on preferred IoT technologies for tracking aged patients with non-communicable diseases (NCDs) in Malaysia. The study aims to analyze the current wearable health system scenario and preferences based on variables like age and gender. Literature reviews covered IoT and healthcare, wearable health devices including activity trackers and smartwatches. Research methodology involved a cross-sectional survey of 450 patients aged 50-70 across three Malaysian states. The study recommends strengthening technological capabilities and guidance to help prevent and treat NCDs in elderly patients.
This work aims to provide a practical guide to assist students of Computer Science
courses and related fields to conduct a systematic literature review. The steps proposed
in this paper to conduct a systematic review were extracted from a technical report
published by the researcher Bárbara Kitchenham [1] and arranged in a more objective
format, in order to make information more accessible and practical, especially for those
who are having their first contact with this technique.
PREDICTION OF COVID-19 USING MACHINE LEARNING APPROACHESIRJET Journal
This document summarizes a research paper that used machine learning models to predict the spread of COVID-19. The researchers used various machine learning algorithms like SVM, random forest, decision tree, and linear regression on COVID-19 case data. SVM had the highest error in predictions, while random forest and decision tree performed best with lowest error. The models were developed using Python and deployed on cloud platforms. The study aimed to accurately predict COVID-19 trends to help governments respond better to the pandemic.
The ALZCARE system aims to improve healthcare for elderly patients with dementia. It consists of three main components: 1) a mobile screening system that allows non-physicians to administer cognitive tests; 2) a clinical information system for physicians to track patient details, evaluations, and recommendations; and 3) an optional patient tracking system using GPS to monitor patients prone to wandering. The systems were designed based on user input and follow clinical guidelines. They are currently undergoing field testing and evaluation in Greece and Albania to assess their ability to screen for cognitive impairment and support patient management.
Social Distancing Detection, Monitoring and Management Using OpenCVIRJET Journal
This document proposes a system to detect social distancing violations using computer vision and deep learning algorithms. The system would identify individuals in video frames using a YOLOv3 model, calculate distances between detected individuals, and classify the risk level based on social distancing guidelines. It transforms frames into a bird's eye view to standardize distance measurements. The proposed system aims to help monitor social distancing and slow the spread of COVID-19 by identifying groups that are too close together. It achieved 92.8% precision in social distancing classification during testing.
This document summarizes the transition from clinical information systems to health grids and the future of health research infrastructure. It discusses trends like rising populations in Asia, increasing resource scarcity, and the need for multidisciplinary and open collaboration. Health grids are presented as enabling virtual collaborations across institutions. Key areas like medical imaging, computational models, and genomic medicine are highlighted. Adoption challenges and requirements like reliable, usable infrastructure are also summarized.
NEW CORONA VIRUS DISEASE 2022: SOCIAL DISTANCING IS AN EFFECTIVE MEASURE (COV...IRJET Journal
The document describes a proposed real-time system to monitor social distancing using computer vision and deep learning techniques. The system would use a camera to detect individuals and calculate distances between them in order to identify instances where social distancing guidelines are breached. When a breach is detected, an audio-visual cue would be emitted to alert individuals without identifying or saving personal data. The system aims to help reduce the spread of COVID-19 while respecting privacy and avoiding overreach. It outlines the technical approach including camera calibration, region of interest definition, object detection using YOLOv3, distance calculation techniques, and system architecture at a high level.
Report-Fog Based Emergency System For Smart Enhanced Living EnvironmentKEERTHANA M
Report-An ambient assisted-living emergency system exploits cloud and fog computing, an outdoor positioning mechanism, and emergency and communication protocols to locate activity-challenged individuals.
IRJET- A Survey on Vision based Fall Detection TechniquesIRJET Journal
This document reviews different vision-based fall detection systems that have been developed using computer vision and image processing techniques. It discusses how vision-based systems work by capturing images or videos using cameras and then analyzing the footage using algorithms to classify events as falls or non-falls. The document also examines some of the challenges of vision-based approaches, such as effects of lighting and background objects, and how newer techniques like convolutional neural networks have helped improve accuracy of fall detection.
This document provides a summary of a project report submitted by four students - Naina Sangole, Chetana Nimje, Prachi Dhawale, and Sejal Meshram - for their Bachelor of Computer Application degree. The report describes the development of a "Covid-19 Management System" project to allow users to register for Covid tests and vaccines online, track Covid cases, and access test reports digitally. The report includes sections acknowledging those who assisted with the project, certifying completion of the project, tables of contents, the project abstract, introduction, literature review on existing systems and their limitations, objectives of the proposed system, screenshots of the system, coding details, and conclusions.
The document is a project report submitted by 4 students (Naina Sangole, Chetana Nimje, Prachi Dhawale, Sejal Meshram) for their BCA degree. The report describes a "Covid-19 Management System" project aimed at developing a system to help track Covid testing, register users for vaccines, track Covid cases, and provide Covid reports online. The report includes an introduction describing the need for the system, literature review on existing systems, proposed system details including analysis, design, and screenshots, and coding details of the developed system.
A DEVICE FOR AUTOMATIC DETECTION OF ELDERLY FALLSIRJET Journal
The document presents a device for automatically detecting falls in elderly individuals. It uses an accelerometer to measure changes in acceleration along three axes and determine body position. When a fall is detected based on acceleration thresholds, the GPS receiver pinpoints the location and a GSM modem sends a text message notification. The system aims to promptly detect falls to reduce injuries and allow for timely medical assistance. It discusses related work on fall detection techniques using sensors like accelerometers and pose estimation. The proposed system design uses an ESP WiFi controller, GPS and GSM modules, accelerometer, and other components to detect falls, track location, and alert caregivers via SMS. It aims to help elderly individuals live independently safely.
The Power of Sensors in health & healthcareD3 Consutling
In a series of reports we explore key digital health trends and related opportunities for technology companies, healthcare providers and patients-consumers. We take both an international and Flemish perspective, the latter based on interviews with local stakeholders. In this report we focus on sensor-based applications.
STANLEY Healthcare received Frost & Sullivan's 2017 Global Company of the Year Award for its performance in the senior living industry. The company provides a full suite of solutions for senior living communities, including emergency call systems, wander management, staff security, and fall prevention. It also offers 24/7 customer support. STANLEY is working on innovative solutions like its Smart Resident Room concept using sensors to continuously monitor residents. Its visionary approach helps address challenges around technology adoption and respecting residents' dignity. STANLEY's strong performance and loyal customer base earned it the Company of the Year recognition.
This document describes a research project that developed a healthcare chatbot to provide medical information and advice to users from their homes. The chatbot uses machine learning algorithms like decision trees and logistic regression trained on medical datasets. It allows users to describe their symptoms and receives a predicted diagnosis along with the probability and potential future symptoms. For emergencies, it recommends expert doctors. The goal is to enable self-diagnosis and determine if hospital visits are needed, saving time. Evaluation showed the chatbot model achieved 95.52% accuracy in identifying illnesses based on symptom inputs. Future work aims to improve the chatbot and add translation features for wider accessibility.
Similar to WIISEL Final Report - 1- Publishable Report Final (20)
1. 1
WIISEL FINAL REPORT
Grant Agreement number: 288878
Project acronym: WIISEL
Project title: Wireless Insole for Independent and Safe Elderly Living
Funding Scheme: Collaborative project (FP7 – ICT)
Period covered: From 1st
November 2013 to 31st
March 2015
Project Coordinator: CETEMMSA
Phone: +34.93.741.91.00
Contact: fbreuil@cetemmsa.com; ereixach@cetemmsa.com
Website: www.wiisel.eu
All rights reserved.
The document is proprietary of the WIISEL consortium members. No copying or distributing, in any form or by any
means, is allowed without the prior written agreement of the owner of the property rights.
This document reflects only the authors’ view. The European Community is not liable for any use that may be made of
the information contained herein.
WIISEL is co-funded by the European Commission under the Seventh (FP7 - 2007-2013)
Framework Programme for Research and Technological Development.
3. 3
1 - WIISEL FINAL PUBLISHABLE REPORT
Grant Agreement number: 288878
Project acronym: WIISEL
Project title: Wireless Insole for Independent and Safe Elderly Living
Funding Scheme: Collaborative project (FP7 – ICT)
Period covered: From 1st
November 2013 to 31st
March 2015
Project Coordinator: CETEMMSA
Phone: +34.93.741.91.00
Contact: fbreuil@cetemmsa.com; ereixach@cetemmsa.com
Website: www.wiisel.eu
All rights reserved.
The document is proprietary of the WIISEL consortium members. No copying or distributing, in any form or by any
means, is allowed without the prior written agreement of the owner of the property rights.
This document reflects only the authors’ view. The European Community is not liable for any use that may be made of
the information contained herein.
WIISEL is co-funded by the European Commission under the Seventh (FP7 - 2007-2013)
Framework Programme for Research and Technological Development.
4. 4
Contents
1- Executive summary .....................................................................................................5
2- Summary of project context and objectives..................................................................6
Defining the need (Identifying the gap).......................................................................................... 6
Project objectives ........................................................................................................................... 6
3- Description of the S&T results/foregrounds .................................................................8
Validation studies......................................................................................................................... 12
Conclusions ..................................................................................................................................18
4- Potential impact, main dissemination activities & exploitation of results.................... 20
Potential impact ........................................................................................................................... 20
Dissemination activities................................................................................................................22
Exploitation of our results ............................................................................................................28
Address of the project website and all relevant contact details...................................................30
5. 5
1- Executive summary
The main goal of WIISEL was to develop a flexible research tool to collect and analyze gait data
from real users and correlate parameters related with the risk of falls from the elderly
population.
The developed tool consists of a combination of a flexible software platform together with
wearable insole device collecting data related with gait. Risk of falls is assessed based on multiple
gait parameters and gait pattern recognition. WIISEL allows quantifying activity, assessing the
quality of gait under real life conditions and enable researchers to evaluate and monitor fall risk in
elderly patients, in the home and community environment, mostly reflecting everyday life
behavior.
Coordinated by CETEMMSA, WIISEL is co-funded by the European Commission (FP7-ICT) for 41
months, with 3.9 M€ budget and 8 partners from 6 different countries.
The potential utility and impact of using the WIISEL system on the research and clinical community
is the following:
- Allowing for remote and quantitative assessment of a user’s fall risk
- Measuring activity and mobility in daily living conditions
- As a clinical assessment tool, allowing its use as part of any research and assessment
of gait parameters.
- Enabling the early identification of functional mobility decline in performance (i.e.
assessment of motor fluctuations and disease progression)
- Enabling fall detection in the home setting
A system like WIISEL, that in an unobtrusive way will allow to analyze movement, posture and
activity of the elderly population by extracting a direct and continuous information from gait, is
beyond any doubt of utmost importance in a challenging ageing of Europe’s population context, as
by assessing the risk of falls among the elders indirectly enables the implementation of early and
adapted interventions to lower this risk of falls.
The WIISEL Project will impact the research and clinical communities by developing an innovative
tool for research and assessment on gait parameters.
Moreover, the WIISEL system as a flexible tool may lay the ground for a commercial pathway as a
continuous and remotely monitoring platform, contributing in the future to the improvement of
the competitiveness of the European industry and to its positioning as a global leader in the field of
Information and Communication technologies (ICT) related to the “Ageing well” concept.
The results obtained during the project support the idea that the WIISEL system can be useful as a
research tool for studying fall risk and as a clinical tool for long-term monitoring of fall risk in the
home and community setting.
6. 6
2- Summary of project context and objectives
Defining the need (Identifying the gap)
Identifying persons at risk for falls is considered a challenge as falls are regarded as rare events that
are unpredictable by nature. However due to the severe consequences of falls, addressing an
increased risk could prevent future falls and break the “vicious cycle” that can result from it.
Traditionally, the assessment of fall risk has been performed in the clinical setting, using
performance based measures that address static and dynamic balance as well as quality of gait. The
Timed Up & Go test (TUG), the Berg Balance Scale (BBS) and the Dynamic gait index (DGI) have
been developed to assess balance and gait in different populations. Although these methods can
provide some indication of a person’s risk of falling, these measures are typically carried out in a
clinical or laboratory setting, where (reverse) white coat effects may limit the utility of the test. In
addition, many of these tests require an experienced professional and are restricted to a limited
and controlled environment, not always reflecting the person's actual difficulties. The at-home
setting may be much more complex and challenging to control, however, traditional performance
based measures may not fully reflect capabilities in these types of situations. Ideally, a fall risk test
should be capable of stratifying and monitoring fall risk even among those who do not yet have
marked impairment and who have not yet fallen.
These identified limitations have been driving the research in this field to try and use more dynamic
wearable devices. Technological advances in wireless communications, miniaturized sensors,
sensorized garments, and low power electronics enabled the penetration of these devices into the
medical field as a tool for assessment in biomedical applications. In this context, wearable sensors,
typically consisting of accelerometers, pressure sensors and/or gyroscopes, represent a promising
technology for preventing and mitigating the effects of falls. The advances in the wearable sensors
systems allow monitoring outside the laboratory environment in more realistic and comfortable
scenarios. This has been already confirmed by a number of studies which have used
accelerometers to assess the risk of falls. The need for simple, unsupervised and quantitative
methods for fall risk assessment is well justified. Moreover, the number of wearable sensors for
gait analysis will keep increasing in the coming future.
Project objectives
Background
Falls are a major health problem for older adults, through both immediate effects such as fractures
and head injuries and in the longer term, problems such as disability, fear of falling that derive in
loss of independence. Fall’s primary risk factors are muscle weakness, history of falls, gait deficit
and balance deficits. The population aged 65 years and over in Europe is estimated to reach 30% of
the total population by the year 2060. Early detection and prediction of elderly falls is thus a major
challenge for the European health and social systems. Its control and reduction should lead to
longer independence, self-confidence and quality of life for the older population, their caregivers
and relatives, and, at the same time, it should help to reduce costs related to falls, thus
contributing to more sustainable healthcare systems.
The analysis of movement, posture and activity in the elderly population has been an important
focus of research for the scientific community, with special attention to the assessment of risk of
7. 7
falls. Many of the systems developed for these analyses are based on single or multiple inertial
sensors, located in different parts of the body, that provide limited information based in body
segment accelerations or rotations. The interpretation of the resulting data is rather problematical.
On the other hand, systems for gait analysis have been mainly focused on platforms or complex
insole systems that constrain the studies to the clinical environment.
The WIISEL solution
The global ambition of the project was to provide a tool to continuous and remotely monitor gait
and fall risk in the elderly and collect information on long term gait data for researchers in this
field.
This tool is made of two elements: a flexible software platform together with a fully-wireless
wearable insole device, collecting data related with gait. Risk of falls is calculated as a new Fall
Risk Index based on multiple gait parameters and gait pattern recognition.
WIISEL solution allows quantifying activity, while assessing the quality of gait under real life
conditions. Furthermore, it provides a tool to researchers in order to monitor and evaluate fall risk
in elderly patients, at home and community environment, mostly reflecting real settings and
regular life behaviour. The system can be used as a research or rehabilitation tool and it enables
the recording of fall events to better recognize and correlate fall-associated gait patterns and
increasing fall risk.
WIISEL has been able to produce a flexible research tool that may have an important impact in the
scientific community, and it is composed by the following elements:
1. A constant monitoring system for elderly people through a wearable and unobtrusive sensing
insole connected to a data analysis system. The WIISEL system continuously captures spatial–
temporal data (e.g. stride time, single support time, and swing time, double support time, cadence,
nº steps per day, step length, stride length, gait speed, heel acceleration) related to human gait and
balance.
2. Intelligent algorithms embedded in a Gait Analysis Software tool to analyse captured data with
high flexibility. Pattern recognition techniques allow quantifying the fall risk and provide useful
information on fall risk assessment. With these results, the project has developed a self learning
analysis framework as a basis for further research in optimizing fall risk prediction and identifying
fall risk factors.
3. A Fall Risk Index based on multiple gait parameters (e.g. stride time, gait speed, step length and
double support time and their variability) and gait pattern recognition to assess and quantify the
risk of fall of elderly population.
4. Real-life and long term human gait data, useful for the scientific community to enrich existing
databases.
5. A fall detection algorithm to feed gait pattern recognition.
Moreover, the WIISEL system, as a flexible and configurable platform, is paving the ground for a
future commercial exploitation as a continuous and remote monitoring platform.
8. 8
3- Description of the S&T results/foregrounds
Description of the system modules
The WIISEL system is a non invasive ambient device, designed to continuously monitor gait
parameters and assess fall risk of a user wearing the WIISEL sensor insoles .The WIISEL system
continuously captures spatial–temporal data (e.g. stride time, single support time, and swing time,
double support time, cadence, nº steps per day, stride length, gait speed, heel acceleration slope,
maximum pressure on heel and toe) related to human gait and balance from a user in his home as
well as in any other locations he may go by foot.
The WIISEL system includes three main modules.
• One Pair of Instrumented Insoles with embedded pressure and inertial sensors and an inductive
wireless charging station
• One Smartphone (off the shelf Bluetooth 4 or above)
• A data analysis system which consist of a back-end server, a web application and standalone
program
The current WIISEL constant monitoring system uses the following working procedure:
The sensing insoles transfers captured gait data from the sensors and transmit them to a
commercial smartphone by using the Bluetooth Low Energy. The smartphone collects data
wirelessly and forward them to a backend computer server via mobile internet connection. Under
the supervision of an expert, the Gait Analysis Software tool assesses the subjects’ main
parameters for the period of time being considered, after filtering the relevant gate data.
WIISEL Insoles
WIISEL insoles are designed to fit different models of footwear placed over the inner surface of the
shoe. A semi-rigid embedded wireless rechargeable battery powers each insole. The components in
the insoles consist of 14 pressure sensors, 1 accelerometer and 1 gyroscope, a microprocessor with
integrated Bluetooth Low Energy module and its Bluetooth antenna, a small storage memory chip,
a Qi chip and an induction coil to facilitate the charging of the battery. The insoles have autonomy
of 16 hours and the recharge time to the batteries full capacity is 90 minutes.
9. 9
Figure 1: Final insole in its charging station
WIISEL Smartphone
The Smartphone serves two primary functions;
Firstly the Smartphone acts as a communications hub, whereby data from the insole sensors are
sent wirelessly via Bluetooth Low Energy (BLE) to the Smartphone, which will be worn by the user.
The Smartphone then uploads this data to a server via a Wi-Fi or 3G/4G connections.
Secondly the Smartphone acts as a system interface for the user wearing the insole. A WIISEL
application on the phone provides a number of features such as:
• Battery power status and signal strength status
• Error messages if problems occur with the insole or communications system
• Summary of the user’s gait and activity, such as steps per day counts, distance
measurements and the current fall risk (customisable)
• A fall detection system interface
The gait pattern analysis and the calculation of fall risk takes place in the Gait Analysis Tool of the
related clinician/researcher whereas the real time fall detection takes place locally at the
Smartphone, using the events then logged with the sensors data. Alarm messages via SMS and
email can be configured individually, via the administration tool on the server.
10. 10
Figure 2: Front page of the application
WIISEL Server and administration
The server is used for data storage and administrative tasks, collecting data from WIISEL
Smartphones and then performing preliminary formatting on these data. After that process, data is
available to the clinicians/researchers through the Gait Analysis Tool, according to their
corresponding access rights. These rights are set through the administration tool.
The administration tool is accessible through a web interface. It allows setting the user roles, the
relations between clinicians/researchers and users of the insoles, and the access rights to the
Smartpone app, as well as to the Gait analysis Tool.
Gait Analysis Tool (GAT)
Once data has been uploaded to the server from the Smartphone, it is accessible through the Gait
Analysis Tool where gait analysis and fall risk is calculated
Several features can be accessed according to the role of the GAT user. The features are organized
in tabs, which are:
- Data input selection
The GAT supports either data input from the WIISEL hardware (insoles) or the import of
data collected with other systems (tools, researchers have already in use)
- Raw data filtering
Raw sensor data as well as extracted gait information and parameters can be visualised and
common filter methods for filtering gait from daily life data can be applied. Also, an
overview of falls of users – recorded by the fall detection app in the smart phone - can be
accessed.
11. 11
- Gait parameter definition
The definition of each gait parameter calculated in the GAT can be accessed. The GAT –
internal script engine allows the modification of definitions and allows adding new
definitions and gait parameters for researchers.
- Gait Parameter History
All calculated gait parameters can be viewed in charts showing the change over time in
long time monitoring
- Pattern extraction
Framework to automatically extract pattern from gait parameters to allow the classification
between fallers and non fallers. In a supervised process, pattern for the calculation of a Fall
Risk Index (FRI) can be selected.
- Fall Risk Index definition
Framework to build FRI(s) from patterns extracted from the Pattern extraction function.
Quantifiers can be set manually or generated automatically using algorithms. As result, the
FRI is represented as a ‘traffic light system’, indicating red as high fall risk, yellow as
medium fall risk and green as low fall risk.
- Fact sheet
Output of a user specific fact sheet that gives a one – page – overview of the users gait and
his current fall risk.
Figure 3 Raw data filtering
12. 12
Figure 4 Fact sheet
Validation studies
WIISEL system was tested during the Validation studies with elderly volunteers. The objective of
the validation studies was mainly to assess the feasibility, usability and functionality of the system
and the ability of the potential target user to use the system and receive valuable information from
it to help them address their risk of falls. More specifically the validation studies addressed the
following aims:
1- To assess the comfort of the WIISEL insole and its durability over long term use by older
adults.
2- To evaluate functionality of the system in collecting gait data from the insole
3- To measure the ability of the system to accurately recognize walking patterns, activity and
sedentary periods and to identify potential risk behaviours based on the data collected.
4- To assess the feasibility of the users to operate the system including independent use at
home, charging the insole, donning and doffing.
5- To evaluate the usability of the system and the ability of the users to understand the
information provided to them from the system.
13. 13
Figure 5 describes the configuration of the studies aimed in assessing the above goals and the
feedback loop intended to improve the system before the final validation in the user’s home.
Figure 5: Configuration of the validation studies
The technical validation was carried out to measure the accuracy of the WIISEL system in its ability
to measure the gait parameters compared to gold standards. In this task, participants walked along
the GaitRite walkway while wearing the WIISEL insoles. Data collected using the GaitRite and the
insoles were processed separately.
In total 86 crossings over the GaitRite were analysed, 928 heel strike and toe off events and 73
strides.
Heel strike and toe off
Mean and standard absolute error between the two systems in heel strike and toe off events are
presented:
Heel strike[sec] Toe off[sec]
Mean absolute error 0.0157 0.0646
Standard absolute error 0.0130 0.0278
As expected toe off has demonstrated larger difference In order to evaluate HS detection a
commonly used gait parameter of stride time was calculated from the insoles using two
consecutive heel strike events from the same leg and then compared to the analysis that was
performed by the GaitRite system.
14. 14
Stride Time
Result of the stride times that both systems calculated during the walking trials are given at the
following table:
GaitRite [sec] Insoles[sec]
Mean stride time 1.1205 1.1210
Standard error in stride
time
0.2160 0.2164
It is possible to see from the tables that both systems provide almost identical results. In order to
perform a more accurate comparison between the two systems a Blant-Altman plot was used
(Figure 6).
Figure 6: Blant-Altman plot to compare the two systems: Insoles and GaitRite.
The analysis demonstrated that the majority of the pressure points are within the confidence lines
(93.89% of the points). Most of the outliers are due to incorrect detection of the heuristic
algorithm. Nevertheless considering that sample rate of the insoles is 30Hz, meaning there is a
temporal resolution of 0.033 seconds between two insole samples, therefore, the error between
the two systems is reasonable and can be used for evaluation of gait parameters.
The comparison of gait parameters collected using the WIISEL system and the GaitRite system
revealed an error of detection of less than 5% which suggests the systems are similar and the
WIISEL system is sufficiently accurate in evaluating gait parameters. This low error rate includes the
analysis of both spatial measures as well as temporal measures taking into account that temporal
measures were highly consistent even more so than spatial measures. This is due to a slightly
higher error rate in detecting toe off. Several reasons exist that may explain this findings. These
include, a specific ‘foot’ flat’ gait pattern (when subjects compensate for ankle movement by
utilizing hip and knee strategies), a short insole as compared to the shoe, a longer shoe which
lengthens the step on the GaitRite as compared to the recordings from the insoles. Nevertheless,
considering the differences in temporal resolution between the systems, we believe this error rate
(~7 %) is acceptable.
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
Mean=-0.00049661
1.96STD
0.038735
-1.96STD
-0.039728
Average between GaitRite and Insoles stride time[sec]
DifferancebetweenGaitRiteandInsolesstridetime[sec]
Bland-Altman plot - stride time
15. 15
The clinical validation studies were divided into 2 phases:
1. The pilot phase: to enable a thorough assessment of the system and its applicability by older
adults.
2. The validation phase: to further establish the feasibility, usability and validity of the final WIISEL
system for continuous monitoring of fall risk in older adults.
The clinical validation included trials in a structured environment and in the home of the
participant and they run in different countries by INRCA (Italy), Nui Galway (Ireland) and TASMC
(Israel). This was a far more substantial test of the system as it was subject to real usage conditions.
Usability, human factors and user experience, as well as user safety, were also be closely observed.
For the development of the pilot and validation studies, a total of 45 pairs of insoles were
manufactured (2 sizes, 38/39 and 42/43) and distributed to the partners together with 15
Smartphones, battery extenders and 18 charging stations.
Pilot study
The protocol of the pilot phase consisted of a total of 3 days of assessment, including 1 day in a
structured controlled laboratory environment followed by 2 days of using the system in the
participant’s home. Pilot phase was carried out with a total of 15 participants aged over 65 years in
INRCA, TASMC and NUI Galway. All individuals recruited had a history of falls. Data collected in this
phase was used to improve and adjust the overall system before starting the validation phase.
Validation study
Validation phase was carried out with a total of 39 participants aged over 65 years in INRCA ,
TASMC and NUI Galway from November 2014 to February 2015. Subjects recruited were
volunteers with a history of falls and healthy age-matched older adults to enable a wide range of
gait patterns and data to be used and validated and a comparison between ‘high risk patterns’ and
normal patterns of older adults with minimal fall risk. These participants used the insole system
briefly in a structured environment while assessment of clinical measures was conducted (FES-I;
POMA, TUG, DGI, FSST, 2MWT, PASE and EQ-5D, see Figure 6 ) and then for about two weeks in
their usual environment during activities of daily living. At the end these two weeks, participants
provided their feedbacks on usability and acceptance of the WIISEL system (SUS and QUEST 2.0).
Data such as gait, postural control, functional status of the subjects, fall risk assessment, device
acceptance, perceived degree of safety, social and psychological aspects was collected and
analyzed to determine the success of the WIISEL system in achieving the expected outcomes.
16. 16
Figure 7: Procedure for the clinical validation studies.
Figure 8: Examples of validation trials in clinical partners’ laboratories
17. 17
Results from validation studies
Results obtained in terms of FRI are shown in Table 1. According to the system 20 volunteers have
a low fall risk; 2 of them have a medium fall risk, and 8 have a high risk of fall. For the remaining 9
subjects the system could not collect enough data to be used for the FRI extraction.
Table 1: Fall Risk Index (FRI) scores
FRI n %
Low risk of fall 20 51.28
Medium risk of fall 2 5.13
High risk of fall 8 20.51
Missing 9 23.08
Total 39 100
A first validation of the FRI was obtained by comparing the index scores with the main standardized
functional tests results included in the validation study. The Tinetti Performance Oriented Mobility
Assessment (POMA), the Timed-Up and Go test (TUG) and the Dynamic Gait Index (DGI) were
selected since their scoring is similar to the FRI and there are very few missing answers.
Table 2: Comparison between FRI and POMA, DGI, TUG
Fall Risk Index (FRI)
Low Medium High
POMA
(p=0.139)
High fall risk 0 0 1
Medium fall risk 2 0 3
Low fall risk 18 2 4
DGI
(p=0.001)
Predictive of falls 0 0 4
Uncertain 7 2 0
Safe ambulator 13 0 4
TUG
(p=0.304)
Freely Mobile 7 0 1
Mostly independent 12 2 6
Variable mobility 0 0 1
Total 20 2 8
Note: statistical significance evaluated through Pearson’s Chi-squared test.
18. 18
Table 2 shows satisfactory adherence between the FRI and the main standardized functional tests.
As regard those classified with Low risk of fall, results from the FRI and from the tests are very
similar, especially as regards POMA and DGI. The medium risk of fall is better predicted by DGI and
TUG, but in this case the number of subjects is very low. As regard the high risk of fall, results from
the system and results from the test are quite scattered and there is a small adherence. The clinical
test best approximated by the FRI in terms of statistical significance is the DGI.
In order to compare the FRI results with just one comprehensive clinical indicator, clustering
analysis techniques have been applied. The cluster analysis is a multivariate statistical technique
aimed at detecting within a population the presence of groups of cases similar to each other and,
at the same time, as different as possible from other groups (difference between).
The selected indicator are POMA, DGI and TUG as these are the three clinical test which estimate
the risk of fall and that are categorized into three main group.
Results from WIISEL project show that the FRI predicts quite well the clinical fall risk identified by
the cluster in a statistically significant way (p=0.006). For the Low fall risk, only 3 subjects do not
overlap between the two indicators: more specifically the system identifies these 3 subjects as Low
risk of fall while they have been clustered as medium. As for the medium risk of fall, the system
identifies only two subjects: one of them was intercepted also by the cluster, while the other one
was identified as low risk of fall. Finally, for the high risk of fall, 4 subjects were identified both by
the cluster and by the FRI, while the other 4, even if classified as low fall risk, are recognized by the
system as high risk.
Overall, among 29 subjects for whom it was possible to calculate both FRI and clusters, for 21 of
them the results matched, while for 8 they did not. The sample size of 29 is not large, in part
because of technical issues related to connection problems. Nonetheless, these initial findings
suggest that the fall risk level assignment based on the FRI is valid, as it performs similarly to
conventional fall risk tests.
Conclusions
The present results support the idea that the WIISEL system can, potentially, be useful as a
research tool for studying fall risk and as a clinical tool for long-term monitoring of fall risk in the
home and community setting, advancing research in this area and leading to relevant savings in
terms of time and money.
At the end of the project, the FRI apparently captures fall risk to a degree that is achieved using
conventional, widely used performance-based tests of fall risk like the Tinetti gait and balance test
and the Dynamic Gait Index. The software tool is flexible and can be adjusted and tailored as
additional parameters are identified. These promising findings need to be confirmed on a larger
number of subjects, first cross sectionally and then longitudinally. Large scale, prospective studies
are needed to further validate the FRI, to determine its added value compared to traditional
measures, to evaluate its ability to predict changes in fall risk over time, and to assess its sensitivity
to changes in response to therapeutic interventions. Nonetheless, the work to date is an important
step forward as it suggests that the FRI appears to be a valid tool for quantifying fall risk. These
initial findings are encouraging.
19. 19
Other key findings of the present work relate to the user’s acceptance of this new technology. Here
too, in general, the results were encouraging. Many aspects of the Smartphone interface and the
insole were already deemed to be acceptable by most users. Important improvements in usability
and user acceptance were achieved after the system was adjusted in response to the feedback
obtained from experts and users. Testing indicated the robustness of certain aspects of the
technology (e.g., connections, durability of the insoles) needs to be improved before the system
can be tested widely. Still, most features of the system functionality are “user friendly” and
acceptable. Larger scale trials designed to further evaluate the FRI and the robustness and usability
of WIISEL system will help to further establish the potential utility of this new system as a research
and clinical tool. The work to date shows that the system has potential and promise.
20. 20
4- Potential impact, main dissemination activities & exploitation of results
Potential impact
Falls are a major health problem for older adults. Immediate effects of falls include fractures and
head injuries. Furthermore, falls also result in long term problems such as disability, fear of falling
and loss of independence1
.
In response to this situation, the main goal of the project has been to develop a flexible research
tool to collect and analyze gait data from real users and correlate parameters related with the risk
of falls from the elderly population.
The project objective has been to create a tool to continuously and remotely monitor gait and fall
risk in the elderly and collect information on long term gait data for researchers in this field. This
tool consisted of a combination of a flexible software platform together with wearable insole
device collecting data related with gait. Risk of falls was calculated as a new Fall Risk Index based
on multiple gait parameters and gait pattern recognition.
Information provided by the WIISEL system allows quantifying activity, assessing the quality of gait
under real life conditions, enabling researchers to evaluate and monitor fall risk in elderly patients,
in the home and community environment, mostly reflecting everyday life behaviour. This
important feature will both help users to become conscious of their increased risk of falling so that
they will take practical steps to decrease this risk, improve their gait and function and possibly
prevent future falls, decrease fear of falling and hopefully enhance an independent lifestyle. For
clinicians, the WIISEL system allows monitoring remotely patients (eg, after discharge), in
order to verify potential improvements during rehabilitation. In addition to this, clinicians will be
able to identify those patients with highest risk of fall in order to screen those more in need of
urgent medical controls. The benefit of this aspect can bring to enormous savings both in term of
time (for clinicians themselves) and costs (for National Health Systems).
Moreover, considering WIISEL as a research or rehabilitation tool, our system enables the recording
of fall events to better recognize and correlate fall-associated gait patterns and increased fall risk or
facilitates finding of new useful gait parameters and new thresholds. During the WIISEL workshop
at the European Falls Festival (Stuttgart, 24-25 March 2015), the WIISEL consortium has discussed
in a round table the potential application of the system for the future imagining an impact in the
following areas:
1. Clinical
2. Sports Medicine
3. Rehabilitation
4. Research and Development
5. Orthotics, prosthetics, braces and shoes
1
Kannus P, Sievänen H, Palvanen M, Järvinen T, Parkkari J. Prevention of falls and consequent injuries in elderly people.
Lancet. 2005 Nov 26;366(9500):1885-93.
21. 21
6. Performing arts
Figure 9: Examples of WIISEL potential applications
Direct impact on partners, achieved during the project
- 11 persons working full time were recruited thanks to the project. In total, 78 persons were
involved in WIISEL with a large range of expertise.
- Several partners completed their facilities with new equipments and tools, which enable them
to be at the forefront of the technology in the field of WIISEL research.
- All partners have gained scientific knowledge and experience, each in its own expertise linked
to WIISEL.
- 48 dissemination presentations, 24 publications, 2 videos released and a Final Dissemination
Event have allowed partners gaining high visibility.
- 1 patentability report was launched and several letters of intention are being signed among
partners to carry on with further research and protection of the foreground.
22. 22
Dissemination activities
Our dissemination plan has been built around a set of questions which gives the right indication
about the direction in which the plan needs to go.
• WHAT information do we have to disseminate? e.g. Project findings: Technologies
(improvements, implementation, future technologies, on-going project work, advances made on
the state-of-the-art, …).
• WHEN dissemination took place?
Dissemination Plan according to project calendar
From month 1 to month 41 Dissemination to the scientific community to gain acceptance of
recognition of the idea about the use of WIISEL as a research tool
for fall risk assessment.
From month 18 Disseminate first results and boost synergies and soft clustering
with other projects and initiatives.
From month 36 Show final results and start looking for potential partners to
implement further activities after the project.
• TO WHOM?
Scientific &Clinical
community (experts in gait
analysis, fall risk)
Dissemination via academic partners, universities and clinical
partners
Technical community Dissemination via academic, research and technological centres
Large Audience Dissemination by all partners
Industrial community Dissemination by SME partners and technology centres
• HOW is the information disseminated?
In order to give the right visibility of the project, especially outside of the consortium, it has been
necessary to build a set of tools which helps partners in doing such communication activity:
- Leaflet which explains, as general idea, the target of the WIISEL project, the purpose and the
consortium
- Website for Scientifics, researchers and General Public (www.wiisel.eu)
- Scientific congresses / Conferences / Final Workshop
23. 23
WIISEL Survey to experts in the field of gait analysis and fall risk assessment.
A small survey was launched among the community of researchers and clinicians with expertise
related to gait analysis and fall risk assessment, in order to explore possible interest from external
partners towards the WIISEL project results. A total of 27 results were obtained with high level of
interest, as it is summarised in the charts below:
Figure 10: Interest in WIISEL features for research purposes
As per the survey results, in general WIISEL seems to be well accepted by the research community
and by clinicians too, not only from a theoretic point of view, but also according their willingness to
use the system.
Effectively, respondents expressed interest in the Fall risk index and Fall detection algorithm and
they recognize in WIISEL the possibility of a deep customization in order to monitor subjects during
their everyday life and to prevent any worsening of their living conditions.
Moreover answerers see in the project the possibility of further development in the future and
they seem to be interested to be kept up to date about this highlighting that research community is
interested in the scientific value added by the WIISEL system.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Constant monitoring system
Analysis framework for research in gait
Real life and long term human gait data
(only on-line survey)
Fall risk index
Fall detection algorithm
Not at all interested Interested Very interested
24. 24
WIISEL Final Dissemination Event
The WIISEL Consortium decided to hold its final dissemination event during the EU Falls Festival
organised by the ProFounD network.
The 1st
European Union Falls Festival (EUFF) was held at the Robert-Bosch-Krankenhaus Atrium,
Stuttgart, Germany, 24th-25th March 2015. The theme for the 2015 Festival is "Technology in the
prediction, detection and prevention of falls".
See the event details at: http://eufallsfestival.eu/ .
The event brought high visibility to WIISEL as one of the event affiliated project, through the
following actions:
1- WIISEL specific workshop
WIISEL Workshop was one of the 8 Parallel Sessions within the “EU Project Showcase” program.
The workshop aimed at disseminating the project results and exploring possible synergies and
actions after the project end based on the different exploitation lines. The workshop was planned
as to get the maximum interaction between speakers and attendees.
WIISEL Workshop at EU Falls Festival, 24-25 March 2015, Stuttgart
25. 25
WIISEL Workshop at EU Falls Festival, 24-25 March 2015, Stuttgart
2- WIISEL Booth
WIISEL had one of the 15 booths presented in the festival.
During the 2 days, a high percentage of the 220 event attendees went through the WIISEL booth,
were demonstrations were running during all the coffee and lunch breaks. Several WIISEL posters
were used in this space and in the general access to the central Auditorium. The WIISEL video has
been running during all these time slots.
WIISEL booth at EU Falls Festival, 24-25 March 2015, Stuttgart
26. 26
3- WIISEL presentation within the “Footwear and Sensors” Workshop.
One of the 7 parallel workshop sessions aimed at footwear and sensor technologies, chaired by
Christopher Moufawad El Achkar, EPFL, Lausanne, Switzerland. Within this workshop, one of the
three speakers was Juan V. Durà from IBV (WIISEL subcontractor). He gave an additional
presentation on WIISEL focusing on the advantages of the WIISEL for gait researchers.
WIISEL presentation at “Footwear and Sensing” workshop, EU Falls Festival, 24-25 March 2015, Stuttgart
27. 27
Summary of dissemination actions
Targeted
results
Means &
channels
Indicators Target
(full project)
Accumulated
(all project period)
1. Constant
monitoring
system
2. Gait
Analysis
Tool
3. Fall Risk
Index
4. Real-life
and long
term
human gait
data
5. Fall
detection
algorithm
Website -update frequency Every 3 months Periodic update done
Newsletter -No. newsletters
released
1 / 6 months -4 newsletters released
Official Press
releases
-No. articles in press -1 in kick off -2 press releases to
several EU-National
supports.
Media kit -No. posters released
-No. leaflets
distributed
-No. videos released
-TV and radio
interviews
2
1500 in project
1
- 5 posters
- 1.800 leaflets released
(2 versions)
- 2 videos released
- 3 radio and 3 TV
interviews done
Dissemination to
different actors
-No. presentations
done
5 in project -48 oral presentations:
‐27 to clinicians &
researchers (gait, fall risk,
rehabilitation etc)
‐ 12 to ICT, electronics,
eHealth community
‐ 9 to industry
Publications -No. publications in
national, EU and
specialized press,
peer-reviewed
journals
3 in the project -15 scientific publications
-9 other publications (2
year 1, 2 year 2)
Other actions -No. booths
-No. visitors /
contacts in fairs
1 in the project
>100
-4 booths
- Participation at 12 fairs.
About 45 contacts with
high potential for future
actions
28. 28
Exploitation of our results
The exploitation strategy followed a methodology based on the definition of the list of all results
valued as potentially exploitable by one (or more) consortium partner. A deep analysis on
technological and market state of the art was performed during the project (constantly compared
with results obtained by the consortium), in order to ensure the innovative degree and potential
demand on the market.
Within our Exploitation Plan, the following tables for each Exploitable Result are available:
- Fiche
- Risk and Action Plan
- Process/Product Evaluation
- Exploitation Strategy
All these tables are an integrant part of the description of each Exploitable Result and had been
elaborated and updated during the project duration, with the specific aim to offering to all partners
an exhaustive summary of all relevant data for the exploitation of each result.
The table below summarizes the exploitable results. Some of them are being analyzed for further
protection and research, with the goal of ensuring exploitation of the results.
Exploitable results
Constant monitoring system through a wearable and unobtrusive sensing insole
connected to a data analysis system
Intelligent algorithms
which utilize data analysis including pattern recognition to quantify fall risk
Fall Risk Index based on multiple gait parameters
Real-life and long term human gait data
Fall detection algorithm
Gait analysis software tool
Smartphone application
29. 29
Thus, the aim of the exploitation strategy has been to optimize the impact of the project trough
targeted exploitation activities and management, involving and informing potential users, by:
- Identifying potential markets, exploitation channels and third parties.
- Assuring project results exploitation by the partners.
- Managing Intellectual/industrial property rights.
30. 30
Address of the project website and all relevant contact details
Website: www.wiisel.com
Project Management:
Fundacio Privada CETEMMSA
Av. d'Ernest Lluch 36, E-08302 Mataro, Spain
www.cetemmsa.com
Project Manager: Fanny Breuil
Project Technical Coordinator : Dr. Elisenda Reixach
Contact: fbreuil@cetemmsa.com – ereixach@cetemmsa.com
Consortium Members:
Universitat Autònoma de Barcelona UAB
Istituto Nazionale di Riposo e Cura per Anziani INRCA INRCA
The Foundation for Medical Research Infrastructural Development and
Health Services next to the Medical Center TEL AVIV
TASMC
Spring Techno GMBH & CO. KG SPRING TECHNO
Tejidos Indesmallables GEISA, S.L. GEISA
National University of Ireland, GALWAY NUI Galway
ACREO AB ACREO
Click here to see the
WIISEL VIDEO