Feature group partitioning: an approach for depression severity prediction with class balancing using machine learning algorithms.

In contemporary society, depression has emerged as a prominent mental disorder that exhibits exponential growth and exerts a substantial influence on premature mortality. Although numerous research applied machine learning methods to forecast signs of depression. Nevertheless, only a limited number of research have taken into account the severity level as a multiclass variable. Besides, maintaining the equality of data distribution among all the classes rarely happens in practical communities. So, the inevitable class imbalance for multiple variables is considered a substantial challenge in this domain. Furthermore, this research emphasizes the significance of addressing class imbalance issues in the context of multiple classes. We introduced a new approach Feature group partitioning (FGP) in the data preprocessing phase which effectively reduces the dimensionality of features to a minimum. This study utilized synthetic oversampling techniques, specifically Synthetic Minority Over-sampling Technique (SMOTE) and Adaptive Synthetic (ADASYN), for class balancing. The dataset used in this research was collected from university students by administering the Burn Depression Checklist (BDC). For methodological modifications, we implemented heterogeneous ensemble learning stacking, homogeneous ensemble bagging, and five distinct supervised machine learning algorithms. The issue of overfitting was mitigated by evaluating the accuracy of the training, validation, and testing datasets. To justify the effectiveness of the prediction models, balanced accuracy, sensitivity, specificity, precision, and f1-score indices are used. Overall, comprehensive analysis demonstrates the discrimination between the Conventional Depression Screening (CDS) and FGP approach. In summary, the results show that the stacking classifier for FGP with SMOTE approach yields the highest balanced accuracy, with a rate of 92.81%. The empirical evidence has demonstrated that the FGP approach, when combined with the SMOTE, able to produce better performance in predicting the severity of depression. Most importantly the optimization of the training time of the FGP approach for all of the classifiers is a significant achievement of this research.

Validation of Smartphone Accelerometry for the Evaluation of Sit-To-Stand Performance and Lower-Extremity Function in Older Adults.

This study tested the concurrent and construct validity of smartphone accelerometry measurement of sit-to-stand (STS) performance in individuals aged 65-89 years. Normal and fast STS times were recorded by smartphone accelerometer, force plate, and video motion systems concurrently, and isokinetic knee extension power and STS whole-body power were obtained. Normal and fast speed STS times from a smartphone accelerometer agreed closely with force plate and motion system methods (mean difference = 0.04 s). Normal and fast STS times were inversely related to isokinetic knee extension power (r = -.93, p < .001 and r = -.82, p < .001, respectively) and STS whole-body power (r = -.76, p < .001 and r = -.70, p < .001, respectively). The STS time obtained from a smartphone accelerometer was equivalent to the established, precise measures of STS time and was related to lower-extremity power, making it a potentially useful metric of lower-extremity function.

Stimulus-Driven Retinal Intrinsic Signal Optical Imaging in Mouse Demonstrates a Dominant Rod-Driven Component.

Purpose: The primary hypotheses tested are that (1) there exist stimulus-driven intrinsic optical signals in the mouse retina similar to those previously observed in other species, and (2) these optical signals require an intact rod photoreceptor phototransduction cascade. Methods: We used 38 wild-type C57BL6J mice and 18 genetic knockout Gnat1-/- mice to study the light-evoked retinal intrinsic response. A custom mouse fundus camera delivered visual stimuli and collected mouse retinal imaging data of changes in retinal reflectance for further analysis. The retina was stimulated in the high-mesopic range with a 505-nm light-emitting diode while also being illuminated with 780-nm near-infrared light. Results: Wild-type C57BL6J mice yielded retinal imaging signals that typically showed a stimulus-driven decrease in retinal reflectance of ∼0.1%, with a time course of several seconds. The signals exhibit spatial specificity in the retina. Overall, the mouse imaging signals are similar in sign and time course to those reported in other mammalian species but are of lower amplitude. In contrast, functional retinal imaging of Gnat1-/- mice that lack a functional rod transducin yielded no such stimulus-driven signals. Conclusions: Previous studies have not shown which pathway component is essential for the generation of these imaged signals. The absence of the intrinsic signal responses in Gnat1-/- knockout mice indicates that a functional rod transducin is likely to be necessary for generating the retinal intrinsic signals. These studies, to the best of our knowledge, demonstrate for the first time in vivo mouse retinal functional imaging signals similar to those previously shown in other mammalian species.

Caregiver perspectives on a smart home-based socially assistive robot for individuals with Alzheimer's disease and related dementia.

Purpose: Innovative assistive technology can address aging-in-place and caregiving needs of individuals with Alzheimer's disease and related dementia (ADRD). The purpose of this study was to beta-test a novel socially assistive robot (SAR) with a cohort of ADRD caregivers and gather their perspectives on its potential integration in the home context.Methods: The SAR involved a programmable research robot linked with commercially available Internet of things sensors to receive and respond to care recipient's behaviour. Eight caregivers observed the SAR perform two care protocols concerning the care recipient's daily routine and home safety, and then participated in a focus group and phone interview. The researchers used grounded theory and the Unified Theory of Acceptance and Use of Technology as a framework to gather and analyse the data.Results: The caregivers' asserted the potential of the SAR to relieve care burden and envisioned it as a next-generation technology for caregivers. Adoption of the SAR, as an identified theme, was subject to the SAR's navigability, care recipient engagement, adaptability, humanoid features, and interface design. In contrast, barriers leading to potential rejection were technological complexity, system failure, exasperation of burden, and failure to address digital divide.Conclusion: From a broader outlook, success of SARs as a home-health technology for ADRD is reliant on the timing of their integration, commercial viability, funding provisions, and their bonding with the care recipient. Long-term research in the home settings is required to verify the usability and impact of SARs in mediating aging-in-place of individuals with ADRD.IMPLICATIONS FOR REHABILITATIONSocially assistive robots (SARs), an emerging domain of assistive technology, are projected to have a crucial role in supporting aging-in-place of individuals with Alzheimer's disease and related dementia (ADRD).Caregivers of individuals with ADRD who observed and interacted with a novel SAR asserted their acceptance of the technology as well as its scope and feasibility for the upcoming generation of caregivers.Navigability, care recipient engagement, adaptability, humanoid features, and interface design were stated to be critical factors for SAR's acceptance by caregiver and care recipient dyads.In contrast, technological complexity, system failure, exasperation of burden, and failure to address digital divide are detrimental to SAR's adoption.Several design and implementation requirements must be considered towards the full-scale development and deployment of the SARs in the home context.

A learning-based agent for home neurorehabilitation.

This paper presents the iterative development of an artificially intelligent system to promote home-based neurorehabilitation. Although proper, structured practice of rehabilitation exercises at home is the key to successful recovery of motor functions, there is no home-program out there which can monitor a patient's exercise-related activities and provide corrective feedback in real time. To this end, we designed a Learning from Demonstration (LfD) based home-rehabilitation framework that combines advanced robot learning algorithms with commercially available wearable technologies. The proposed system uses exercise-related motion information and electromyography signals (EMG) of a patient to train a Markov Decision Process (MDP). The trained MDP model can enable an agent to serve as a coach for a patient. On a system level, this is the first initiative, to the best of our knowledge, to employ LfD in an health-care application to enable lay users to program an intelligent system. From a rehabilitation research perspective, this is a completely novel initiative to employ machine learning to provide interactive corrective feedback to a patient in home settings.

Robots to assist daily activities: views of older adults with Alzheimer's disease and their caregivers.

BACKGROUND: Robots have the potential to both enable older adults with dementia to perform daily activities with greater independence, and provide support to caregivers. This study explored perspectives of older adults with Alzheimer's disease (AD) and their caregivers on robots that provide stepwise prompting to complete activities in the home. METHODS: Ten dyads participated: Older adults with mild-to-moderate AD and difficulty completing activity steps, and their family caregivers. Older adults were prompted by a tele-operated robot to wash their hands in the bathroom and make a cup of tea in the kitchen. Caregivers observed interactions. Semi-structured interviews were conducted individually. Transcribed interviews were thematically analyzed. RESULTS: Three themes summarized responses to robot interactions: contemplating a future with assistive robots, considering opportunities with assistive robots, and reflecting on implications for social relationships. Older adults expressed opportunities for robots to help in daily activities, were open to the idea of robotic assistance, but did not want a robot. Caregivers identified numerous opportunities and were more open to robots. Several wanted a robot, if available. Positive consequences of robots in caregiving scenarios could include decreased frustration, stress, and relationship strain, and increased social interaction via the robot. A negative consequence could be decreased interaction with caregivers. CONCLUSIONS: Few studies have investigated in-depth perspectives of older adults with dementia and their caregivers following direct interaction with an assistive prompting robot. To fulfill the potential of robots, continued dialogue between users and developers, and consideration of robot design and caregiving relationship factors are necessary.

Performance of daily activities by older adults with dementia: the role of an assistive robot.

Older adults with cognitive impairment often have difficulties in remembering the proper sequence of activities of daily living (ADLs) or how to use the tools necessary to perform ADLs. They, therefore, require reminders in a timely fashion while performing ADLs. This is a very stressful situation for the caregivers of people with dementia. In this paper we describe a pilot study where a tele-operated assistive robot helps a group of older adults with dementia (OAwD) to perform an ADL, namely making a cup of tea in the kitchen. Five OAwD along with their caregivers participated in this study which took place in a simulated-home setting. The purpose of this study was to investigate the feasibility and usability of a robotic system in assisting the OAwD to perform ADL in a home setting. The findings from this study will contribute to achieve our ultimate goal of designing a full-fledged assistive robot that assists OAwD aging in their own homes. The assistive robots designed for people with dementia mostly focus on companionship. This is, to the best of our knowledge, the first attempt to design an assistive robot which will provide step-by-step guidance to people with dementia in their activities of daily living.

A probabilistic model of overt visual attention for cognitive robots.

Visual attention is one of the major requirements for a robot to serve as a cognitive companion for human. The robotic visual attention is mostly concerned with overt attention which accompanies head and eye movements of a robot. In this case, each movement of the camera head triggers a number of events, namely transformation of the camera and the image coordinate systems, change of content of the visual field, and partial appearance of the stimuli. All of these events contribute to the reduction in probability of meaningful identification of the next focus of attention. These events are specific to overt attention with head movement and, therefore, their effects are not addressed in the classical models of covert visual attention. This paper proposes a Bayesian model as a robot-centric solution for the overt visual attention problem. The proposed model, while taking inspiration from the primates visual attention mechanism, guides a robot to direct its camera toward behaviorally relevant and/or visually demanding stimuli. A particle filter implementation of this model addresses the challenges involved in overt attention with head movement. Experimental results demonstrate the performance of the proposed model.

Macronutrient intake and glycemic control in a population-based sample of American Indians with diabetes: the Strong Heart Study.

BACKGROUND: Little research has explored the association of macronutrient intake and glycated hemoglobin (HbA(1c)) in adults with diabetes. OBJECTIVE: The objective of the study was to examine the cross-sectional association between macronutrient intake and HbA(1c) in diabetic American Indians. DESIGN: A total of 1284 participants aged 47-80 y who had diabetes for >or=1 y at the second examination (1993-1995) of the Strong Heart Study were included in this study. Dietary intake was assessed by using a 24-h recall. Logistic regression models were used to evaluate the odds of poor glycemic control (HbA(1c) >or= 7%) among sex-specific quintiles of macronutrient intake, after adjustment for the possible confounders age, sex, study center, body mass index, duration of diabetes, diabetes treatment, smoking, alcohol drinking, total energy intake, and physical activity. RESULTS: Higher total fat (>25-30% of energy), saturated fatty acids (>13% of energy), and monounsaturated fatty acids (>10% of energy) and lower carbohydrate intake (<35-40% of energy) were associated with poor glycemic control. Lower fiber intake and higher protein intake were marginally associated with poor glycemic control (P for trend = 0.06 and 0.09, respectively). No significant association was found between polyunsaturated fatty acids or trans fatty acids and glycemic control in this population. CONCLUSIONS: These data suggest that a higher consumption of total fat and saturated and monounsaturated fatty acids and a lower intake of carbohydrates are associated with poor glycemic control in diabetic American Indians. Clinical trials focusing on whether modifications of macronutrient composition improve glycemic control in persons with diabetes are needed.

Albuminuria within the "normal" range and risk of cardiovascular disease and death in American Indians: the Strong Heart Study.

BACKGROUND: "Normal" albuminuria has been defined as urinary albumin-creatinine ratio (UACR) less than 30 mg/g (3.4 mg/mmol). Whether higher UACR within this range independently predicts cardiovascular disease (CVD) and CVD death is uncertain. METHODS: A total of 3,000 participants aged 45 to 74 years with a UACR less than 30 mg/g and free of CVD at the baseline examination of the Strong Heart Study (SHS) were evaluated. Survival time was calculated from the baseline examination to the first nonfatal CVD, fatal CVD, or December 31, 2002. RESULTS: During follow-up (average, 10.4 years), 383 incident nonfatal CVD and 145 fatal CVD cases were ascertained. After adjustment for conventional CVD risk factors, participants with a UACR in the third (UACR >or= 5.4 to <10.2 mg/g [>or=0.6 to <1.1 mg/mmol] in men, >or=7.6 to <12.9 mg/g [>or=0.9 to <1.4 mg/mmol] in women) and the fourth (UACR >or=10.2 to <30 mg/g in men, >or=12.9 to <30 mg/g in women) quartiles had 41% and 72% greater risks of all CVD events and 118% and 199% greater risks of CVD mortality than those in the lowest quartile (UACR < 2.7 mg/g [<0.3 mg/mmol] in men, <4.3 mg/g [<0.5 mg/mmol] in women), respectively. In subgroup analysis, these associations were more pronounced in persons with diabetes. CONCLUSION: In the SHS cohort of middle-aged to elderly American Indians, albuminuria levels less than the traditional cutoff value predict CVD. Our findings agree with a growing number of studies questioning the concept that UACR less than 30 mg/g is normal.

Type 2 diabetes and impaired fasting glucose in American Indians aged 5-40 years: the Cherokee diabetes study.

PURPOSE: To estimate the prevalence of type 2 diabetes and impaired fasting glucose (IFG) and to study several potential risk factors for type 2 diabetes among Oklahoma Cherokees aged 5 to 40 years. METHODS: A random sample of 2205 members of the Cherokee Nation of Oklahoma aged 5 to 40 years was recruited. Demographic, clinical, and laboratory data were collected. Type 2 diabetes and IFG were determined using the 1997 American Diabetes Association (ADA) criteria. Relationships between type 2 diabetes and potential risk factors were examined by univariate and multivariate regression methods. RESULTS: According to ADA criteria, the age-adjusted prevalence proportions of type 2 diabetes were 4.3% in females and 4.8% in males. Among the 89 individuals who had type 2 diabetes, 31 were newly diagnosed. Thirty-two (1.5%, 18 females and 14 males) were found to have IFG. The prevalence of type 2 diabetes and IFG increased with age, number of parents with diabetes, obesity, degree of Indian heritage, high triglyceride value, and low HDL cholesterol. CONCLUSIONS: The increasing prevalence of type 2 diabetes in young American Indians is alarming. The findings must be disseminated to the Indian communities and their health care providers. Preventive measures and early detection programs must be designed and implemented for children and adolescents in this population.