Automated Recognition of Complex Agitation Behavior of Dementia Patients Using Video Camera

Fook, Foo Siang; Thang, Pham Viet; Htwe, That Mon; Qiu, Qiang; Wai, Aung Aung Phyo; Jayachandran, M.; Biswas, Jit; Yap, Philip

doi:10.1109/health.2007.381605

Cited by 22 publications

(13 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Compared to these techniques, Kintense is a completely non-invasive and passive monitoring system. Remote monitoring of sleeping patients for agitation has been performed using regular video cameras [12]. However, their downside is that they do not provide skeleton level information to accurately classify an action, and compared to video images, the stick figures from Kinect are far less privacy invasive.…”

Section: False Alarmsmentioning

confidence: 99%

“…State-of-the-art agitation monitoring systems use two kinds of sensing techniques in general: invasive [26], [15], [6] and non-invasive [1], [12], [21]. Invasive techniques use physiological parameters such as EEG waveforms, skin temperature, skin conductance, pupil diameter, respiration rate, and accelerometer readings from wearable sensors to detect agitation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Kintense: A robust, accurate, real-time and evolving system for detecting aggressive actions from streaming 3D skeleton data

Nirjon

Greenwood

Torres

et al. 2014

2014 IEEE International Conference on Pervasive Computing and Communications (PerCom)

View full text Add to dashboard Cite

Abstract-Kintense is a robust, accurate, real-time, and evolving system for detecting aggressive actions such as hitting, kicking, pushing, and throwing from streaming 3D skeleton joint coordinates obtained from Kinect sensors. Kintense uses a combination of: (1) an array of supervised learners to recognize a predefined set of aggressive actions, (2) an unsupervised learner to discover new aggressive actions or refine existing actions, and (3) human feedback to reduce false alarms and to label potential aggressive actions. This paper describes the design and implementation of Kintense and provides empirical evidence that the system is 11% − 16% more accurate and 10% − 54% more robust to changes in distance, body orientation, speed, and person when compared to standard techniques such as dynamic time warping (DTW) and posture based gesture recognizers. We deploy Kintense in two multi-person households and demonstrate how it evolves to discover and learn unseen actions, achieves up to 90% accuracy, runs in real-time, and reduces false alarms with up to 13 times fewer user interactions than a typical system.

show abstract

Section: False Alarmsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Kintense: A robust, accurate, real-time and evolving system for detecting aggressive actions from streaming 3D skeleton data

Nirjon

Greenwood

Torres

et al. 2014

2014 IEEE International Conference on Pervasive Computing and Communications (PerCom)

View full text Add to dashboard Cite

show abstract

“…In fact, some of these possibilities are already being researched. Examples of automatic assessment of agitation in dementia include the work of Fook et al [291] which aim at measuring it with video streams, or the one by Bankole et al [292] where the use of body sensors was purposed instead of cameras. In the case of wondering detection, steps have already been given so as to detect it using activity recognition techniques in wireless sensor networks [293], as said before, by means of GPS tracking systems [294] and also by using RFID tags and antennas to track the walking patterns and ADLs [295] of the patients.…”

Section: Othersmentioning

confidence: 99%

On the early diagnosis of Alzheimer's Disease from multimodal signals: A survey

Alberdi

Aztiria

Basarab

2016

Artificial Intelligence in Medicine

167

View full text Add to dashboard Cite

“…Body Sensor Network [33] Peak Detection [34] StressSense [35] A sensor network can perform constant monitoring of activities.…”

Section: Anxiety and Agitationmentioning

confidence: 99%

Internet of Things Enabled Technologies for Behaviour Analytics in Elderly Person Care: A Survey

Newcombe

Yang

Carter

et al. 2017

2017 IEEE International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and I

View full text Add to dashboard Cite

The advances in sensor technology over recent years has provided new ways for researchers to monitor the elderly in uncontrolled environments. Sensors have become smaller, cheaper and can be worn on the body, potentially creating a network of sensors. Smart phones are also more common in the average household and can also provide some behavioural analysis due to the built in sensors. As a result of this, researchers are able to monitor behaviours in a more natural setting, which can lead to more useful data. This is important for those that may be suffering from mental illness as it allows for continuous, non-invasive monitoring in order to diagnose symptoms from different behaviours. However there are various challenges that need to be addressed ranging from issues with sensors to the involvement of human factors. It is vital that these challenges are taken into consideration along with the major behavioural symptoms that can appear in an Elderly Person. For a person suffering with Dementia, the application of sensor technologies can improve the quality of life of the person and also monitor the progress of the disease through behavioural analysis. This paper will consider the behaviours that can be associated with dementia and how these behaviours can be monitored through sensor technology. We will also provide an insight into some sensors and algorithms gathered through survey in order to provide advantages and disadvantages of these technologies as well as to present any challenges that may face future research.

show abstract

Automated Recognition of Complex Agitation Behavior of Dementia Patients Using Video Camera

Cited by 22 publications

References 9 publications

Kintense: A robust, accurate, real-time and evolving system for detecting aggressive actions from streaming 3D skeleton data

Kintense: A robust, accurate, real-time and evolving system for detecting aggressive actions from streaming 3D skeleton data

On the early diagnosis of Alzheimer's Disease from multimodal signals: A survey

Internet of Things Enabled Technologies for Behaviour Analytics in Elderly Person Care: A Survey

Contact Info

Product

Resources

About