Machine Learning and the Elderly
Machine learning has the potential to improve everyone’s quality of life, but could have the biggest impact with the elderly population’s quality of life. Using machine learning, the elderly have a chance to stay independent longer, be their own primary caretaker, and live fuller lives.
Self-driving cars are on many people’s minds nowadays with many companies like Google, Tesla, and even Apple rumored to be working on one, but not many people think about the elderly right away when they hear about self-driving cars. These automated vehicles would help the elderly immensely, giving them the freedom to get to where they need and want, without possibly endangering themselves or others on the road due to health complications.
Proactive home care for the elderly would allow them to stay in their homes longer without the need of an expensive in home caretaker. Hong Kong Communications International has began developing a proactive system that uses IR motion sensors placed in the elderly’s home and utilizes machine learning spatiotemporal reasoning to learn the elderly’s movement patterns. Then, if abnormal motion is detected, medical attention can quickly be dispatched to arrive on the scene, far within the golden eight hours of rescue.
Another advance in machine learning that will help the elderly is being heavily researched in Japan, caretaker robots, or Carebots. These robots would live with the elderly patient and help aid their independence as well. In 2015, nearly one third of the Japan government budget was allocated to developing these Carebots, as Japan has projected a shortage of over one million caretakers by 2025. Without proper caretakers, an elderly person’s quality of life can decrease drastically. These Carebots are being developed to aid the elderly around the house, as if a real human caretaker was with them.
The elderly often have complex health complications, making accurate and quick diagnoses important to their very survival. Machine learning is being developed to give six month term mortality scores to patients, using machine learning classifiers to rank risk accurately and quickly. In the long term, this could potentially lead to better ranking for organ transplant lists, quicker medical attention, and a generally healthier life for the elderly.
Home automation would give a boon to elderly independence as it would mimic having a caretaker at home, but at a much more affordable price. Machine learning could be used to predict and learn the elderly’s usage and automatically order additional food or medicine if necessary. Home automation could also be utilized as safety proofing potentially dangerous parts of the home, such as ovens, burners, AC, showers and tubs, etc. Knowing what is abnormal behavior would allow the home automation system to turn off the potentially dangerous system in question and call for emergency services, giving both preventive care and reactionary care.
Mortality Prediction: http://mghassem.mit.edu/wp-content/uploads/2013/02/Makar_Ghassemi_IJMLC_506.pdf