Artificial neural network to classify cognitive impairment using gait and clinical variables

BackgroundApproximately 30% of individuals with stroke experience impending cognitive decline, often leading to dementia over the subsequent years. Despite an increased risk of dementia among individuals with stroke, there is no predictive tool to identify stroke individuals who are at higher risk of future cognitive decline. The purpose of the Come‐Drive study is to identify whether dual impairments in cognitive and motor function may provide behavioral biomarkers to predict future cognitive decline in stroke.MethodThirty‐two stroke survivors (cognitively impaired; N = 16 and cognitively normal; N = 16) participated. All participants performed (1) Cognitive tasks: A divided attention task provided mean processing time. A simple reaction time task with the unaffected upper extremity provided an index of variability in processing time. (2) Motor tasks: Overground walking to provide gait speed and gait (stride time) variability measurements. Hierarchical logistic regression analyses were performed to determine whether processing speed and/or gait performance predicted participants’ cognitive status.ResultThe cognitively impaired stroke group showed significantly increased mean processing time (p = 0.001), and increased variability in processing time (p = 0.005), compared to the cognitively normal stroke group. The cognitively impaired stroke group showed slower gait speed and amplified gait variability, but these differences failed to reach statistical significance (p > 0.05). The logistic regression showed that the mean and variability of processing time significantly predicted the cognitive status in the individuals with stroke (X2 (4) = 17.29, p = 0.002). Gait speed and gait variability did not emerge as significant predictors of cognitive status in individuals with stroke. The regression model accounted for 58% of the variance in determining the cognitive status post‐stroke with 83% accuracy.ConclusionDual decline in processing speed and gait did not emerge to be better predictors of cognitive status in individuals with stroke than the decline in processing speed alone. Perhaps a more complex motor task may serve as a better predictor of cognitive decline than a simple overground walking task. Overall, unique impairments in cognitive processing speed may serve as indicators of poor cognitive status in individuals with stroke.

Advanced age is a well-recognized risk factor for adverse outcomes after trauma. A substantial body of literature, much of it cited within this article, demonstrates increased morbidity and mortality in geriatric trauma patients compared with their younger counterparts. Whether this outcome difference is because of the decreased physiologic reserve that accompanies aging, a higher incidence of preexisting medical conditions in the geriatric patient, or other factors yet to be identified remains unclear. It is clear, however, that good outcomes can be achieved in this patient population when appropriately aggressive trauma care is directed toward geriatric patients with survivable injuries. Implicit in the above statement is the need to identify, as soon as possible after injury, those patients who will benefit from aggressive resuscitation, timely injury management, and posttrauma rehabilitation. It is equally important, however, to limit these intensive and expensive treatment modalities to patients whose injuries are not only survivable but also compatible with an acceptable quality of life. Our purpose in developing this guideline was to provide the trauma practitioner with some evidence-based recommendations that could be used to guide decision-making in the care of the geriatric trauma patient. We began this process by first developing a series of questions, the answers to which we hoped could be supported by the existing scientific literature. The initial set of questions were as follows: 1. Is age itself a marker of increased morbidity/ mortality? If so, what age should be used? 2. Is age instead a surrogate for increased preexisting conditions (PECs)? If so, which premorbid conditions are particularly predictive of poor outcomes? 3. Should age itself be a criterion for triage from the field directly to a trauma center, regardless of Glasgow Coma Scale (GCS) score, Trauma Score (TS), and so forth? If so, what age should be used? 4. Do trauma centers have better outcomes with geriatric trauma than nontrauma centers? 5. Are there specific injuries, scores (e.g., Injury Severity Score [ISS], TS, GCS score), or PEC/age combinations in geriatric trauma patients that are so unlikely to be survivable that a nonaggressive approach from the outset could be justified? 6. What resuscitation end-points should be used for the geriatric trauma patient? 7. Should all geriatric trauma patients receive invasive hemodynamic monitoring? If so, what specific types of monitoring should be used? If not, which geriatric patients benefit from invasive monitoring? Unfortunately, after examining the available literature, it is clear that evidence-based responses to all of the questions raised above are not possible. As the evidentiary tables demonstrate, there are few, if any, prospective, randomized, controlled trials that definitively address any of the above issues. Second, there is a lack of uniformity as to a specific age criterion for geriatric trauma. As shown in the evidentiary tables, geriatric trauma is variously defined in the literature as age greater than or equal to 55, 60, 65, 70, 75, and even 80 years of age. There is even literature support for increased mortality from trauma beginning at age 45! Furthermore, because age is a continuous variable, and not a dichotomous one, adverse outcomes associated with geriatric trauma are likely to increase in a continuous fashion with age as opposed to a stepwise leap as a given patient reaches a specific age. Third, there is no concise definition of a geriatric trauma patient. In some studies, all patients over a given age are included, whereas in others, patients with penetrating injuries, burns, and minor injuries, such as slip-and-falls, are excluded. Some studies include all patients regardless of hemodynamic instability or injury severity, whereas others impose strict entrance criteria or exclude patients who do not survive for a predetermined period of time after admission. Such lack of uniformity regarding inclusion criteria makes it Submitted for publication October 3, 2001. Accepted for publication September 16, 2002. Copyright © 2003 by Lippincott Williams & Wilkins, Inc. From the Carolinas Medical Center (D.G.J.), Charlotte, North Carolina, Trauma Service, Bronson Hospital (B.R.P.), Kalamazoo, and Trauma Burn Center, University of Michigan Health System (W.W.), Ann Arbor, Michigan, New York Hospital–Cornell Medical Center (P.S.B.), New York, New York, Robert Wood Johnson Medical School (J.S.H.), New Brunswick, New Jersey, Mt. Sinai Hospital (M.R.H.), Chicago, Illinois, Morehouse School of Medicine (K.E.S.), Atlanta, Georgia, and R Adams Cowley Shock Trauma Center, University of Maryland Medical Center (T.M.S.), Baltimore, Maryland. Address for reprints: David G. Jacobs, MD, Carolinas Medical Center, P.O. Box 32861, Charlotte, NC 28232; email: djacobs@carolinas.org.

The identification of altered gait and its progression over time is important to gaining a better understanding of the clinical aspects of mild cognitive impairment (MCI) in older adults. The aim of the present systematic review was to determine changes in gait variables over time among older adults with MCI. The PubMed, Web of Science, Scopus, and Science Direct databases were searched for relevant articles using the following keywords and Medical Subject Headings: Aged AND "Mild cognitive impairment" AND (gait OR locomotion). A hand search was also performed of the reference lists of the selected articles in an attempt to find additional records. The following were the inclusion criteria: longitudinal studies and clinical trials involving a control group without intervention; samples of individuals 65 years or older; and characterization of gait using a single or dual task. The initial search led to the retrieval of 6979 studies, 9 of which met the inclusion criteria. The duration of follow-up among the studies ranged from 6 months to 2 years. Most trials investigated gait speed. Other gait variables were step length, time required to walk a given distance, and mean weekly gait speed. Altered gait progressed in older adults with MCI. The main alterations were gait speed and variability in daily number of steps in follow-up periods lasting more than 1 year. No significant changes in gait variables were found in shorter follow-up periods (up to 6 months). The progression of gait changes in older adults with MCI has been underinvestigated. MCI leads to reduced gait speed in longer follow-up periods. Such information can contribute to the determination of motor interventions for older adults with MCI, especially in the early stages.

BackgroundThe benefit of physical exercise in ageing and particularly in frailty has been the aim of recent research. Moreover, physical activity in the elderly is associated with a decreased risk of mortality, of common chronic illnesses (i.e. cardiovascular disease or osteoarthritis) and of institutionalization as well as with a delay in functional decline. Additionally, very recent research has shown that, despite its limitations, physical exercise is associated with a reduced risk of dementia, Alzheimer disease or mild cognitive decline. Nevertheless, the effect of physical exercise as a systematic, structured and repetitive type of physical activity, in the reduction of risk of cognitive decline in the elderly, is not very clear. The purpose of this study aims to examine whether an innovative multicomponent exercise programme called VIVIFRAIL has benefits for functional and cognitive status among pre-frail/frail patients with mild cognitive impairment or dementia.Methods/designThis study is a multicentre randomized clinical trial to be conducted in the outpatient geriatrics clinics of three tertiary hospitals in Spain. Altogether, 240 patients aged 75 years or older being capable of and willing to provide informed consent, with a Barthel Index ≥ 60 and mild cognitive impairment or mild dementia, pre-frail or frail and having someone to help to supervise them when conducting the exercises will be randomly assigned to the intervention or control group. Participants randomly assigned to the usual care group will receive normal outpatient care, including physical rehabilitation when needed. The VIVIFRAIL multicomponent exercise intervention programme consists of resistance training, gait re-training and balance training, which appear to be the best strategy for improving gait, balance and strength, as well as reducing the rate of falls in older individuals and consequently maintaining their functional capacity during ageing.The primary endpoint is the change in functional capacity, assessed with the Short Physical Performance Battery (1 point as clinically significant). Secondary endpoints are changes in cognitive and mood status, quality of life (EQ-5D), 6-m gait velocity and changes in gait parameters (i.e. gait velocity and gait variability) while performing a dual-task test (verbal and counting), handgrip, maximal strength and power of the lower limbs as well as Barthel Index of independence (5 points as clinically significant) at baseline and at the 1-month and 3-month follow-up.DiscussionFrailty and cognitive impairment are two very common geriatric syndromes in elderly patients and are frequently related and overlapped. Functional decline and disability are major adverse outcomes of these conditions. Exercise is a potential intervention for both syndromes. If our hypothesis is correct, the relevance of this project is that the results can contribute to understanding that an individualized multicomponent exercise programme (VIVIFRAIL) for frail elderly patients with cognitive impairment is more effective in reducing functional and cognitive impairment than conventional care. Moreover, our study may be able to show that an innovative individualized multicomponent exercise prescription for these high-risk populations is plausible, having at least similar therapeutic effects to other pharmacological and medical prescriptions.Trial registrationClinicalTrials.gov, NCT03657940. Registered on 5 September 2018.

BackgroundDespite the large neuroanatomic measurements obtained from brain MRI segmentation, we do not have yet a definitive neuroanatomic or vascular feature/even that can precisely delineate a cognitively normal subject from cognitively impaired and Alzheimer’s disease (AD). Increasing lifespan of population accompanied with increased prevalence of cardiovascular risks across globe demands for precise early detection of AD and cognitive status (impaired or normal) in the aging individuals. Here, we have employed machine learning (ML) method on neuroanatomic volume and white matter hyperintensity (WMH) volume obtained from MRI analyses of data from National Alzheimer’s Coordinating Center (NACC) to identify participants as cognitively normal (CN), cognitively impaired with etiological diagnosis as AD and cognitively impaired (CI) subjects.MethodBrain segmentation outputs for neuroanatomic‐volume and White matter hyperintensity (WMH) was obtained from NACC for 1290 unique subjects with total of 3684 longitudinal‐MRI. Supervised ML was performed using the MRI parameters for identifying the subjects as AD, CI, and CN. 50% of the MRI were used as training set while another 50% was used as test. Various combinations of neuroanatomic volume with WMH quantity were tested for accuracy of delineating the AD, CI and CN groups using random forest method. Furthermore, various ML methods such as XGB classifiers, classification tree, bagging classifier and simple classification were tested for accuracy using the optimized neuroanatomic volume and WMH.ResultRandom forest method converged with a combination of 5‐MRI features, intracranial, gray‐matter, cerebral CSF, hippocampus, and White‐matter hyperintensity volume with highest accuracy (83.3%) in identifying AD, CI, and CN subjects (Fig. 1). Inclusion of normal appearing white‐matter volume during the ML optimization of MRI features led to significant reduction in the accuracy. Simple random forest ML method showed highest accuracy using the optimized MRI features (Fig. 2).ConclusionAll the neuroanatomic events obtained from brain MRI may not be specific in providing diagnostic features for evaluating brain health. Here we show that a combination of 5 MRI features that includes microvascular pathology and neuroanatomic volume may provide a specific and efficient non‐invasive imaging biomarker to evaluate brain health in aging population.

Current assessments of fatigue and sleepiness rely on patient reported outcomes (PROs), which are subjective and prone to recall bias. The current study investigated the use of gait variability in the "real world" to identify patient fatigue and daytime sleepiness. Inertial measurement units were worn on the lower backs of 159 participants (117 with six different immune and neurodegenerative disorders and 42 healthy controls) for up to 20 days, whom completed regular PROs. To address walking bouts that were short and sparse, four feature groups were considered: sequence-independent variability (SIV), sequence-dependant variability (SDV), padded SDV (PSDV), and typical gait variability (TGV) measures. These gait variability measures were extracted from step, stride, stance, and swing time, step length, and step velocity. These different approaches were compared using correlations and four machine learning classifiers to separate low/high fatigue and sleepiness.Most balanced accuracies were above 50%, the highest was 57.04% from TGV measures. The strongest correlation was 0.262 from an SDV feature against sleepiness. Overall, TGV measures had lower correlations and classification accuracies.Identifying fatigue or sleepiness from gait variability is extremely complex and requires more investigation with a larger data set, but these measures have shown performances that could contribute to a larger feature set.Clinical relevance- Gait variability has been repeatedly used to assess fatigue in the lab. The current study, however, explores gait variability for fatigue and daytime sleepiness in real-world scenarios with multiple gait-impacted disorders.

Abstracts 8th PCNE working symposium ‘Navigating research on pharmaceutical care’. 11–12 February 2022, Lisbon, Portugal

Artificial Neural Network (ANN) Is Superior to the Pre-Endoscopic Rockall Score in Predicting Which Patients Will Benefit From Urgent EGD in Acute Upper GI Bleeding (UGIB)

Radiomics and machine learning have been extensively utilized in the realm of urinary stones, particularly in forecasting stone treatment outcomes. The objective of this study was to integrate clinical variables and radiomic features to develop a machine learning model for predicting the stone-free rate (SFR) following percutaneous nephrolithotomy (PCNL).A total of 212 eligible patients who underwent PCNL surgery at the Second Affiliated Hospital of Nanchang University were included in a retrospective analysis. Preoperative clinical variables and non-contrast-enhanced CT images of all patients were collected, and radiomic features were extracted after delineating the stone ROI. Univariate analysis was conducted to identify clinical variables strongly correlated with the stone-free rate after PCNL, and the least absolute shrinkage and selection operator algorithm (lasso regression) was utilized to screen radiomic features. Four supervised machine learning algorithms, including Logistic Regression, Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Gradient Boosting Decision Tree (GBDT), were employed. The clinical variables with strong correlation and screened radiomic features were integrated into the four machine learning algorithms to construct a prediction model, and the receiver operating curve was plotted. The area under the receiver operating curve (AUC), the accuracy rate, the specificity, etc., were used to evaluate the predictive performance of the four models.After analyzing postoperative statistics, the stone-free rate following the procedure was found to be 70.3% (n = 149). Among the various clinical variables examined, factors, such as stone number, stone diameter, stone CT value, stone location, and history of stone surgery, were identified as statistically significant in relation to the stone-free rate after PCNL. A total of 121 radiomic features were extracted, and through lasso regression, 7 features most closely associated with the stone-free rate post-PCNL were identified. The predictive accuracy of different models (Logistic Regression, RF, XGBoost, and GBDT) for determining the stone-free rate after PCNL was evaluated, yielding accuracies of 78.1%, 76.6%, 75.0%, and 73.4%, respectively. The corresponding area under the curve AUC (95%CI) were 0.85 (0.83-0.89), 0.81 (0.76-0.85), 0.82 (0.78-0.85), and 0.77 (0.73-0.81), positioning these models among the top performers in logistic regression prediction. In terms of predictive importance scores, the key factors identified by the logistic regression model were number of stone, zone percentage, stone diameter, and surface area. Similarly, the RF model highlighted number of stone, stone CT value, stone diameter, and surface area as the top predictors.Among the four machine learning models, the logistic regression model demonstrated the highest accuracy and discrimination ability in predicting the stone-free rate following PCNL. In comparison to XGBoost and GBDT, RF also exhibited superior accuracy and a certain level of discrimination ability. However, based on the performance of all four models, logistic regression is more likely to aid in clinical decision-making by assisting clinicians in diagnosing PCNL in patients. This enables us to effectively predict the presence of residual stones post-surgery and ultimately select patients who are suitable candidates for PCNL.

Background Glioblastoma multiforme (GBM) is a highly aggressive type of brain cancer with short survival time, poor prognosis and high mortality. Although research based on molecular data plays a crucial role in predicting cancer, few studies have been developed to combine high-throughput molecular data with clinical variables. Our goal is to establish machine learning based models to predict GBM survival. Methods The molecular data and clinical information of GBM were downloaded from The Cancer Genome Atlas (TCGA) database. The molecular data include somatic copy-number alteration (SCNA) and microRNA. Logistic regression and support vector machine (SVM) were used to establish the predictive models using molecular data, clinical variables and their combinations. Receiver operating characteristic (ROC) was conducted to estimate the sensitivity and specificity. Results The AUC scores are 0.72, 0.69 and 0.82, respectively using clinical variables, SCNA data and their combinations. The AUC scores are 0.97 and 0.98, respectively using microRNA data and the combinations of microRNA data and clinical variables. Conclusion Our study developed machine learning based models to predict GBM survival. The combination of molecular data and clinical variables can improve prediction accuracy. Moreover, it also brings us new insights into the molecular mechanisms underlying GBM cancer. Acknowledgements Supported by the National Natural Science Foundation of China (61471078, 31270903), the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry, the Program for Liaoning Excellent Talents in University (LJQ2015011) and the Fundamental Research Funds for the Central Universities (3132016330, 3132014306, 3132015213, and 3132017075).

BackgroundAbout one-third of people older than 65 years fall at least once a year. Physical exercise has been previously demonstrated to improve gait, enhance physical fitness, and prevent falls. Nonetheless, the addition of cognitive training components may potentially increase these effects, since cognitive impairment is related to gait irregularities and fall risk. We hypothesized that simultaneous cognitive–physical training would lead to greater improvements in dual-task (DT) gait compared to exclusive physical training.MethodsElderly persons older than 70 years and without cognitive impairment were randomly assigned to the following groups: 1) virtual reality video game dancing (DANCE), 2) treadmill walking with simultaneous verbal memory training (MEMORY), or 3) treadmill walking (PHYS). Each program was complemented with strength and balance exercises. Two 1-hour training sessions per week over 6 months were applied. Gait variables, functional fitness (Short Physical Performance Battery, 6-minute walk), and fall frequencies were assessed at baseline, after 3 months and 6 months, and at 1-year follow-up. Multiple regression analyses with planned comparisons were carried out.ResultsEighty-nine participants were randomized to three groups initially; 71 completed the training and 47 were available at 1-year follow-up. DANCE/MEMORY showed a significant advantage compared to PHYS in DT costs of step time variability at fast walking (P=0.044). Training-specific gait adaptations were found on comparing DANCE and MEMORY: DANCE reduced step time at fast walking (P=0.007) and MEMORY reduced gait variability in DT and DT costs at preferred walking speed (both trend P=0.062). Global linear time effects showed improved gait (P<0.05), functional fitness (P<0.05), and reduced fall frequency (−77%, P<0.001). Only single-task fast walking, gait variability at preferred walking speed, and Short Physical Performance Battery were reduced at follow-up (all P<0.05 or trend).ConclusionLong-term multicomponent cognitive–physical and exclusive physical training programs demonstrated similar potential to counteract age-related decline in physical functioning.

Tolerance and Outcomes of Neoadjuvant Chemotherapy in Geriatric Breast Cancer Patients

Predicting cognitive impairment among geriatric patients at Asir central hospital, Saudi Arabia.

Artificial intelligence has been applied in wildfire science and management since the 1990s, with early applications including neural networks and expert systems. Since then, the field has rapidly progressed congruently with the wide adoption of machine learning (ML) methods in the environmental sciences. Here, we present a scoping review of ML applications in wildfire science and management. Our overall objective is to improve awareness of ML methods among wildfire researchers and managers, as well as illustrate the diverse and challenging range of problems in wildfire science available to ML data scientists. To that end, we first present an overview of popular ML approaches used in wildfire science to date and then review the use of ML in wildfire science as broadly categorized into six problem domains, including (i) fuels characterization, fire detection, and mapping; (ii) fire weather and climate change; (iii) fire occurrence, susceptibility, and risk; (iv) fire behavior prediction; (v) fire effects; and (vi) fire management. Furthermore, we discuss the advantages and limitations of various ML approaches relating to data size, computational requirements, generalizability, and interpretability, as well as identify opportunities for future advances in the science and management of wildfires within a data science context. In total, to the end of 2019, we identified 300 relevant publications in which the most frequently used ML methods across problem domains included random forests, MaxEnt, artificial neural networks, decision trees, support vector machines, and genetic algorithms. As such, there exists opportunities to apply more current ML methods — including deep learning and agent-based learning — in the wildfire sciences, especially in instances involving very large multivariate datasets. We must recognize, however, that despite the ability of ML models to learn on their own, expertise in wildfire science is necessary to ensure realistic modelling of fire processes across multiple scales, while the complexity of some ML methods such as deep learning requires a dedicated and sophisticated knowledge of their application. Finally, we stress that the wildfire research and management communities play an active role in providing relevant, high-quality, and freely available wildfire data for use by practitioners of ML methods.

Aim Older inpatients have reduced physical function and walking ability with a higher risk of falls after being discharged home. Gait variability can assess ambulation and is strongly related to the risk of falls. However, the clinical factors affecting gait variability in inpatients have not been identified. The purpose of this study was to investigate the predictive factors affecting gait variability in older inpatients. Methods A total of 42 older orthopedic inpatients with fractures of the hip, spine, and other segments and 18 healthy volunteers as the control group were enrolled in this study. Inpatients wore tri-axial accelerometers for a 10m walk before discharge. Gait variability was assessed by the coefficient of variation (CV) based on five consecutive stride times. Clinical assessment of muscle strength, joint mobility, balance, pain, and activities of daily living were also evaluated. Results The CV in inpatients was higher than that in healthy elderly. Quadriceps muscle strength, ankle dorsiflexion range of motion, and balance described the CV. When model 2 (adjusted R2 = 0.473) was compared with model 1 (adjusted R2 = 0.293), the quadriceps muscle strength and ankle dorsiflexion range of motion had a major effect on CV, while balance had not a greater influence than these two factors when compared with model 3 (adjusted R2 = 0.537). Conclusions Poor knee extension strength, balance, and restriction of ankle dorsiflexion mobility have influenced gait variability in older inpatients.