top of page
Writer's pictureThe Natural Philosopher

Smart Architecture for In-Home Healthcare

Updated: Aug 22, 2018

By Vanessa Palermo


The elderly population in the USA and Brazil is growing rapidly, and consequently, so is the need for in-home health care. Elderly populations are more susceptible to diseases than younger populations, so oftentimes, people in these populations are found needing in-home treatment following surgical procedures. For people with a wide range of diseases, disabilities, and disorders, in-home one-on-one care could be highly beneficial for both monitoring their physical and mental health. Unfortunately, 24/7 nurses come with a high price and a loss of independence for the patient. The loss of the ability to live an independent lifestyle can be a difficult transition for many and at times may be frustrating and discouraging.


Health Smart Homes (HSH) have been proposed as a potential solution for those who either cannot afford a nurse providing 24/7 in-home care or wish to hold onto their independence. Researchers from The University of Sao Carlos and the Vale Institute of Technology in Brazil have teamed up with researchers at the University of Southern California to create a model of an individualized HSH system called Smart Architecture for In-Home Healthcare (SAHHc). This system is unlike any of its kind in the way that it incorporates the patient’s emotional state into their treatment. According to psychologists, emotion plays a major role in recovering from diseases. Additionally, emotion is a critical component for many mental disorders such as schizophrenia, depression, autism and bipolar disorder. For this reason, SAHHc could be a promising solution for patients living with those disorders as well as those recovering from diseases.

Art by Vanessa Palermo

The SAHHc is customizable for the patient’s environment, running on the power supply of the patient’s home. SAHHc utilizes strategically placed sensors and a decision making element to optimize in-home patient care. Numerous sensors are planted throughout a person’s home and are programmed to detect changes in the environment such as the presence of people. The sensors are programmed to identify the patient, and if a person who is not the patient is present, the sensors will decide to not gather information about that person. Different types of sensors are placed with the intention of extracting different types of information.


Facial detection is no easy task. Since extracting facial information can be very difficult if a person is moving or not looking directly at a camera, smart cameras that can pan, tilt, and move in response to movement are used to optimize the image being used for emotion identification. Depressed people tend to look down more frequently, and for this reason the moving camera feature is critical to the success of this system. Some unresolved facial detection issues remain, such as the lighting of the environment, the hairstyle, and head motion of the patient, but researchers are working on improving the technology to account for these challenges.


After the face has been successfully detected, the facial features are extracted and the emotion is identified. The image of the patient portraying their current emotional state is captured and compared to a baseline image of the patient in a neutral state using an geometric algorithm. The algorithm compares 66 points on the face in relation to each other. Cross-culturally, when humans are experiencing different emotional states such as fear, anger, disgust, sadness, surprise or happiness, the distance between the 66 points on the face changes. The difference in distance between these is indicative of the emotion that person is actively experiencing. For example, when a person is surprised, their eyebrows may raise and the distance between their eyebrows and eyes will be different than it is in the baseline. The system also detects changes in the texture and color of the skin which also can serve as indicators of the person’s emotional state, such as when a person is mad. Emotional analysis is what sets the SAHHc apart from other HSHs because the technology incorporates highly researched psychological phenomenon, upholding the importance of patient experience in their medical care.


After the emotion has been identified and analyzed, the extracted information is sent to the decision making element. The decision making element has received input on the patient’s emotion as well as their treatment protocol. From there, it processes this information and makes a decision on what that patient needs. If the patient is in need of help, the decision making element is responsible for deciding who to contact to get that help. When the SAHHc is installed, the patient configures their preferences for who is contacted in what type of emergency using an Android application.


The sort customizability of the SAHHc is a major advantage of the smart home. For one, it allows for the patient’s treatment to be consistent with their treatment protocol and disorder, which are also configured when installed. Additionally, items that the patient may have already owned such as smart phones, smart watches, and clothing that contain wearable sensors can configured into the system to provide additional information about heart rate or body temperature to further inform care. What may be one of the most important aspects from a psychological perspective is that fact that by configuring the patient’s preferences via Android application, it also takes into account that the patient is a human with different types of relationships and is respectful towards who is contacted under differing circumstances. By maintaining an understanding of aspects that are fundamental to humans, such as emotions and relationships, the patient can feel comfortable that the smart home is fulfilling its goal of using technology to better their life.


The SAHHc provides hope for individuals recovering from surgeries, experiencing loss of ability with age, or suffering from lifelong disorders. An issue raised has been the concern of privacy, but the researchers argue their HSH allows for even more privacy for the patient than a 24/7 nurse would because SAHHc can be turned off by the patient at any time. Another concern is that the technology is not accurate enough in identifying individuals and emotions to be practical. Researchers argue that their technology had a 99.75% accuracy in identifying the patient and 80% accuracy in identifying emotion. Technology will continue to improve and the SAHHc has been designed to be able to evolve as technology does. The algorithms for facial and emotional recognition are being improved to take into account diseases that may affect the patient’s ability to show emotions on their face, such as in Parkinson’s disease.


The SAHHc could be the future of affordable in home care, but what does that mean for us? As citizens of this planet, we will likely see health smart home technologies becoming more prevalent— and perhaps even becoming the norm— by the time some of us move on to our elderly years. Health smart homes may be exciting to those who are thrilled by the prospect of freedom and accessible health care, but may be intimidating to others who are unsure about the boundaries between technology and human life. As technology continues to advance, we must consider what technologies like the SAHHc may mean not only to our parents and grandparents, but us, our children and the future generations. If health smart homes do become the norm, the need for in-home nurses will likely decrease due to cost. It is vital that we acknowledge that though health smart homes have the potential to benefit our communities at large, many unforeseen outcomes could come with it and finding the balance between consequences for the elderly and consequences for the future generations is a duty of us all.


References:


Mano, L. Y., Faical, B. S., Nakamura, L. H., Gomes, P. H., Libralon, G. L., Meneguete, R. I., . . . Ueyama, J. (2016, March 15). Exploiting IoT technologies for enhancing Health Smart Homes through patient identification and emotion recognition. Retrieved February 5, 2018, from https://www.sciencedirect.com/science/article/pii/S0140366416300688

24 views0 comments

コメント


bottom of page