+ Internet of Things (IoT) technology made it possible to monitor and compare the activity pattern of persons with mild cognitive impairment (MCI) with healthy elderly persons over a period of several months and diagnosed with Alzheimer’s disease based on pattern changes [20, 66, 73]. + MCI can be diagnosed in a person using IoT based on how the computer is used [29, 56, 57]. + Early diagnosis of neurological disorders became possible by comparing the amount of activity during different periods using IoT [26, 29, 63, 66, 73]. + MCI can be diagnosed using IoT by comparing the daily activities of a healthy person with a patient with Alzheimer disease (AD), diagnosing the type of activities based on the person’s presence in different places of the house, and comparing the results with those of conventional tests [20, 21, 26, 63, 73]. + MCI can be diagnosed by IoT-based machine learning techniques; e.g., in a study, random forest (RF) showed the best performance (precision-recall ,,) for this purpose [73]. + The outcomes of a scenario (carrying several tasks) had high correlation () with the mini mental state examination (MMSE) score. Alzheimer’s disease can be diagnosed with high accuracy by observing the activities through IoT (the area under the ROC (receiver operating characteristic) curve is =0.98, and the sensitivity and specificity are 100% and 94%). Monitoring tasks with IoT can assist in detecting MCI with high accuracy (the area under the ROC , and the sensitivity and specificity are 74% and 89%) [29].
+ IoT technology made it possible to extract behavioral patterns by examining the person’s ADLs and recognizing unusual and dangerous behaviors [20, 54, 55, 57, 65, 67, 71]. + Using IoT technology improved daily activities such as ironing, cooking, and personal hygiene [47, 54, 62, 73]. + IoT technology made it possible to compare the two healthy and Alzheimer’s groups in displacing kitchen utensils. The number of completed activities among people with Alzheimer’s was significantly lower [29, 49, 53, 75]. + Using the IoT system, when employing a single sensor or multiple sensors, the mean -measure for daily activity detection are high ( and ) [53]. +From examining the degree of correlation between various activities by IoT system, obtained at night, there was 0.71 correlation between getting up and going to the bathroom. The correlation between waking up in the middle of the night and doing daily activities was -0.35, and the correlation between getting up at night and performing the usual activities of the elderly and medical staff was 0.47 [55]. + Activities could be precisely recognized by an IoT system. Sensitivity to identifying certain activities was reported as follows: cooking (84.29%), eating (87.78%), toileting (94.79%), and getting ready for bed (92.38%). Specificity of this system was reported as sleeping (85.77%), getting ready for bed (94.48%), eating (94.83%), seated activity (94.98%), and grooming (96.98%) [57].
+ Sleep monitoring using IoT is a good criterion for the early diagnosis of dementia [21, 23, 26, 28, 58, 67, 72]. + Using IoT showed that the sleep quality, quantity, and rhythm of a healthy person are better than that of a patient. Thus, IoT can be used to monitor sleep and diagnose AD [21, 26, 28, 58, 72]. + Using IoT made it possible to study changes in the sleep patterns of the elderly [23, 28, 47, 58]. + Using IoT improved patients’ sleep duration and reduced frequent waking-up during the night [21, 58, 67]. + In monitoring sleeping by IoT system, the average sleep efficiency was found to be between 32.1% and 80% [58]. + Automatic IoT-based sleep analysis had similar findings to the individual’s self-reports (,,) [68]. + The participants’ sleep patterns changed after the IoT intervention. For example: (i) Increased the deep sleep time from 7% to 38% (ii) Reduced the amount of time that spend awake at night (e.g., from 11 times to 6 times) (iii) Reduced light sleep from 74% to 58% of total sleep time (iv) Decreased the overall amount of time spent awake in bed at night from 3.79% to 22.1% [21] + The correlation between light sleep length and cognitive tests was 0.836, while the correlation between total sleep duration and cognitive tests was 0.843, according to sleep monitoring by an IoT system [47].
+ IoT technology can monitor adherence to the medication regimen and gives the necessary alerts to healthcare providers or informal caregivers in the shortest time [20, 21, 29, 53, 67]. + Using IoT made it possible to monitor medication use and irregularities [47, 60, 63, 68].
+ IoT technology enabled users to make alerts in the event of observing abnormalities in the patient’s vital signs, along with a list of necessary actions in the case of encountering any alert [26, 47, 58, 63, 68]. + Using IoT and monitoring data such as heart rate or respiratory rate can predict agitation in AD/dementia patients [21, 24, 69, 70]. + IoT system is capable of measuring heart rate and respiratory rate (correlation between nurses and system was 0.874 without Charite Dome (ChD), 0.608 with ChD for heart rate, and 0.840 and 0.602 without ChD and with ChD for the respiratory rate [70].
+ IoT technology can reduce the stress and anxiety of the patient and their caregivers using warning messages appropriate to the patient’s wandering and agitation state as well as severity of the danger that threatens them [24, 27, 69, 70]. + Acceptable results were obtained in using robots and systems equipped with IoT for helping the patient’s wandering and agitation state [20, 22, 62]. + IoT system can detect persistent vocalizations by measuring the heart rate; it is found that the heart rate is around 40 beats per minute at times of persistent vocalizations [69].
+ IoT can remind people of different types of activities and reduce dependence on others to help them remember to do ADLs correctly [47, 63, 73]. + The effect of using IoT technology on helping the independence and security of people with memory impairment showed patients’ independence in performing activities such as daily walking [48, 49, 63]. + Correlation between traditional memory test (face name test) and an IoT system was 0.597, whereas the correlation between response time and face name test was 0.341. Therefore, IoT system can measure memory abilities in AD/dementia [49].
+ Accurate and continuous monitoring of social interactions using IoT showed individuals with better social interactions obtained higher scores in executive function tests [21, 26, 67]. + Using IoT can help better understand the relationship between social interactions and cognitive decline [21, 63, 67]. + IoT reduced depression, isolation symptoms, and watching television as well as increased social interactions with others and participation in various social programs [26, 63, 67].
+ Behavioral patterns of patients can be examined to monitor normal and abnormal behaviors as a sign of apathy in individuals using the data obtained from the IoT-enabled systems [20, 59, 62]. + IoT-based systems can differentiate normal and abnormal behaviors accurately (,,) [59].
+ Using IoT for continuously assessing walking speed at home provided a better understanding of changes in people’s speed over time [22, 55, 58, 62]. + Walking speed and its daily variability obtained using IoT can be among the first symptoms of MCI progression [21, 62, 64, 65]. + Movement disorders and disability in ADLs, MCI, and dementia caused by AD can be diagnosed at an earlier time using sensor technology and gait analysis [55, 62, 64, 65, 68, 74]. - Walking in the home environment was generally at low acceleration and rarely in a straight line; therefore, it is difficult to develop algorithms that use accelerometers to detect changes in walking [72, 73]. + An IoT-based machine learning method can detect sitting, sleeping, or standing positions accurately (97% accuracy by Naive Bayes (NB) and 99.1% accuracy by support vector machine (SVM) [60]. + By using IoT system and detection of movements, it is possible to predict MCI and Alzheimer’s disease. For example, after 3.2 years, 25% of persons were diagnosed by MCI, and two of the 12 movements previously detected by IoT system were useful in predicting MCI. This figure for Alzheimer’s was 9.4% of the participants using four types of movements [65].
+ Patients who go out of the safe area could be identified using IoT, and their location could be tracked and informed to their caregivers using mobile phones [24, 50, 61]. + An IoT-based monitoring system for the detection of patient’s presence yielded positive results, as follows: (i) The average detection time of patient’s departure from the safe region was 1.97 sec, which was enough to make a notification [61] (ii) In 82% of the cases, the presence of a patient could be detected at the distance of 48-55 m [61]
