Machine Learning Algorithm Helps Identify Non-Diagnosed Prodromal Alzheimer's Disease Patients in the General Population.

BACKGROUND: Recruiting patients for clinical trials of potential therapies for Alzheimer's disease (AD) remains a major challenge, with demand for trial participants at an all-time high. The AD treatment R and D pipeline includes around 112 agents. In the United States alone, 150 clinical trials are seeking 70,000 participants. Most people with early cognitive impairment consult primary care providers, who may lack time, diagnostic skills and awareness of local clinical trials. Machine learning and predictive analytics offer promise to boost enrollment by predicting which patients have prodromal AD, and which will go on to develop AD. OBJECTIVES: The authors set out to develop a machine learning predictive model that identifies prodromal AD patients in the general population, to aid early AD detection by primary care physicians and timely referral to expert sites for biomarker confirmation of diagnosis and clinical trial enrollment. DESIGN: The authors use a classification machine learning algorithm to extract patterns within healthcare claims and prescription data three years prior to AD diagnosis/AD drug initiation. SETTING: The study focused on subjects included within proprietary IQVIA US data assets (claims and prescription databases). Patient information was extracted from January 2010 to July 2018, for cohorts aged between 50 and 85 years. PARTICIPANTS: A total of 88,298,289 subjects aged between 50 and 85 years were identified. For the positive cohort, 667,288 subjects were identified who had 24 months of medical history and at least one record with AD or AD treatment. For the negative cohort, 3,670,254 patients were selected who had a similar length of medical history and who were matched to positive cohort subjects based on the prevalence rate. The scoring cohort was selected based on availability of recent medical data of 2-5 years and included 72,670,283 subjects between the ages of 50 and 85 years. Intervention (if any): None. MEASUREMENTS: A list of clinically-relevant and interpretable predictors was generated and extracted from the data sets for each subject, including pharmacological treatments (NDC/product), office/specialist visits (specialty), tests and procedures (HCPCS and CPT), and diagnosis (ICD). The positive cohort was defined as patients who have AD diagnosis/AD treatment with a 3 years offset as an estimate for prodromal AD diagnosis. Supervised ML techniques were used to develop algorithms to predict the occurrence of prodromal AD cases. The sample dataset was divided randomly into a training dataset and a test dataset. The classification models were trained and executed in the PySpark framework. Training and evaluation of LogisticRegression, DecisionTreeClassifier, RandomForestClassifier, and GBTClassifier were executed using PySpark's mllib module. The area under the precision-recall curve (AUCPR) was used to compare the results of the various models. RESULTS: The AUCPRs are 0.426, 0.157, 0.436, and 0.440 for LogisticRegression, DecisionTreeClassifier, RandomForestClassifier, and GBTClassifier, respectively, meaning that GBTClassifier (Gradient Boosted Tree) outperforms the other three classifiers. The GBT model identified 222,721 subjects in the prodromal AD stage with 80% precision. Some 76% of identified prodromal AD patients were in the primary care setting. CONCLUSIONS: Applying the developed predictive model to 72,670,283 U.S. residents, 222,721 prodromal AD patients were identified, the majority of whom were in the primary care setting. This could drive major advances in AD research by enabling more accurate and earlier prodromal AD diagnosis at the primary care physician level , which would facilitate timely referral to expert sites for in-depth assessment and potential enrolment in clinical trials.

Perceived state of self during motion can differentially modulate numerical magnitude allocation.

Although a direct relationship between numerical allocation and spatial attention has been proposed, recent research suggests that these processes are not directly coupled. In keeping with this, spatial attention shifts induced either via visual or vestibular motion can modulate numerical allocation in some circumstances but not in others. In addition to shifting spatial attention, visual or vestibular motion paradigms also (i) elicit compensatory eye movements which themselves can influence numerical processing and (ii) alter the perceptual state of 'self', inducing changes in bodily self-consciousness impacting upon cognitive mechanisms. Thus, the precise mechanism by which motion modulates numerical allocation remains unknown. We sought to investigate the influence that different perceptual experiences of motion have upon numerical magnitude allocation while controlling for both eye movements and task-related effects. We first used optokinetic visual motion stimulation (OKS) to elicit the perceptual experience of either 'visual world' or 'self'-motion during which eye movements were identical. In a second experiment, we used a vestibular protocol examining the effects of perceived and subliminal angular rotations in darkness, which also provoked identical eye movements. We observed that during the perceptual experience of 'visual world' motion, rightward OKS-biased judgments towards smaller numbers, whereas leftward OKS-biased judgments towards larger numbers. During the perceptual experience of 'self-motion', judgments were biased towards larger numbers irrespective of the OKS direction. Contrastingly, vestibular motion perception was found not to modulate numerical magnitude allocation, nor was there any differential modulation when comparing 'perceived' vs. 'subliminal' rotations. We provide a novel demonstration that numerical magnitude allocation can be differentially modulated by the perceptual state of self during visual but not vestibular mediated motion.

Right hemisphere dominance directly predicts both baseline V1 cortical excitability and the degree of top-down modulation exerted over low-level brain structures.

Right hemisphere dominance for visuo-spatial attention is characteristically observed in most right-handed individuals. This dominance has been attributed to both an anatomically larger right fronto-parietal network and the existence of asymmetric parietal interhemispheric connections. Previously it has been demonstrated that interhemispheric conflict, which induces left hemisphere inhibition, results in the modulation of both (i) the excitability of the early visual cortex (V1) and (ii) the brainstem-mediated vestibular-ocular reflex (VOR) via top-down control mechanisms. However to date, it remains unknown whether the degree of an individual's right hemisphere dominance for visuospatial function can influence, (i) the baseline excitability of the visual cortex and (ii) the extent to which the right hemisphere can exert top-down modulation. We directly tested this by correlating line bisection error (or pseudoneglect), taken as a measure of right hemisphere dominance, with both (i) visual cortical excitability measured using phosphene perception elicited via single-pulse occipital trans-cranial magnetic stimulation (TMS) and (ii) the degree of trans-cranial direct current stimulation (tDCS)-mediated VOR suppression, following left hemisphere inhibition. We found that those individuals with greater right hemisphere dominance had a less excitable early visual cortex at baseline and demonstrated a greater degree of vestibular nystagmus suppression following left hemisphere cathodal tDCS. To conclude, our results provide the first demonstration that individual differences in right hemisphere dominance can directly predict both the baseline excitability of low-level brain structures and the degree of top-down modulation exerted over them.

How imagery changes self-motion perception.

Imagery and perception are thought to be tightly linked, however, little is known about the interaction between imagery and the vestibular sense, in particular, self-motion perception. In this study, the observers were seated in the dark on a motorized chair that could rotate either to the right or to the left. Prior to the physical rotation, observers were asked to imagine themselves rotating leftward or rightward. We found that if the direction of imagined rotation was different to the physical rotation of the chair (incongruent trials), the velocity of the chair needed to be higher for observers to experience themselves rotating relative to when the imagined and the physical rotation matched (on congruent trials). Accordingly, the vividness of imagined rotations was reduced on incongruent relative to congruent trials. Notably, we found that similar effects of imagery were found at the earliest stages of vestibular processing, namely, the onset of the vestibular-ocular reflex was modulated by the congruency between physical and imagined rotations. Together, the results demonstrate that mental imagery influences self-motion perception by exerting top-down influences over the earliest vestibular response and subsequent perceptual decision-making.

Effects of prochlorperazine on normal vestibular ocular and perceptual responses: a randomised, double-blind, crossover, placebo-controlled study.

BACKGROUND: The present study investigated whether prochlorperazine affects vestibulo-ocular reflex (VOR) and vestibulo-perceptual function. METHODS: We studied 12 healthy naïve subjects 3 h after a single dose of oral prochlorperazine 5 mg in a randomised, placebo-controlled, double-blind, crossover study in healthy young subjects. Two rotational tests in yaw were used: (1) a threshold task investigating perceptual motion detection and nystagmic thresholds (acceleration steps of 0.5°/s(2)) and (2) suprathreshold responses to velocity steps of 90°/s in which vestibulo-ocular and vestibuloperceptual time constants of decay, as well as VOR gain, were measured. RESULTS: Prochlorperazine had no effect upon any measure of nystagmic or perceptual vestibular function compared to placebo. This lack of effects on vestibular-mediated motion perception suggests that the drug is likely to act more as an anti-emetic than as an antivertiginous agent.

Unidirectional visual motion adaptation induces reciprocal inhibition of human early visual cortex excitability.

OBJECTIVES: Behavioural observations provided by the waterfall illusion suggest that motion perception is mediated by a comparison of responsiveness of directional selective neurones. These are proposed to be optimally tuned for motion detection in different directions. Critically however, despite the behavioural observations, direct evidence of this relationship at a cortical level in humans is lacking. By utilising the state dependant properties of transcranial magnetic stimulation (TMS), one can probe the excitability of specific neuronal populations using the perceptual phenomenon of phosphenes. METHOD: We exposed subjects to unidirectional visual motion adaptation and subsequently simultaneously measured early visual cortex (V1) excitability whilst viewing motion in the adapted and non-adapted direction. RESULT: Following adaptation, the probability of perceiving a phosphene whilst viewing motion in the adapted direction was diminished reflecting a reduction in V1 excitability. Conversely, V1 excitability was enhanced whilst viewing motion in the opposite direction to that used for adaptation. CONCLUSION: Our results provide support that in humans a process of reciprocal inhibition between oppositely tuned directionally selective neurones in V1 facilitates motion perception. SIGNIFICANCE: This paradigm affords a unique opportunity to investigate changes in cortical excitability following peripheral vestibular disorders.

Applications of neuromodulation to explore vestibular cortical processing; new insights into the effects of direct current cortical modulation upon pursuit, VOR and VOR suppression.

Functional imaging, lesion studies and behavioural observations suggest that vestibular processing is lateralised to the non-dominant hemisphere. Moreover, disruption of interhemispheric balance via inhibition of left parietal cortex using transcranial direct current stimulation (tDCS) has been associated with an asymmetric suppression of the vestibulo-ocular reflex (VOR). However, the mechanism by which the VOR was modulated remains unknown. In this paper we review the literature on non-invasive brain stimulation techniques which have been used to probe vestibular function over the last decade. In addition, we investigate the mechanisms whereby tDCS may modulate VOR, e.g. by acting upon pursuit, VOR suppression mechanisms or direct VOR modulation. We applied bi-hemispheric parietal tDCS in 11 healthy subjects and only observed significant effects on VOR gain (tdcs * condition p=0.041) - namely a trend for VOR gain increase with right anodal/left cathodal stimulation, and a decrease with right cathodal/left anodal stimulation. Hence, we suggest that the modulation of the VOR observed both here and in previous reports, is directly caused by top-down cortical control of the VOR as a result of disruption to interhemispheric balance, likely parietal.

Suboptimal maternal nutrition, during early fetal liver development, promotes lipid accumulation in the liver of obese offspring.

Maternal nutrition during the period of early organ development can modulate the offspring's ability to metabolise excess fat as young adults when exposed to an obesogenic environment. This study examined the hypothesis that exposing offspring to nutrient restriction coincident with early hepatogenesis would result in endocrine and metabolic adaptations that subsequently lead to increased ectopic lipid accumulation within the liver. Pregnant sheep were fed either 50 or 100% of total metabolisable energy requirements from 30 to 80 days gestation and 100% thereafter. At weaning, offspring were made obese, and at ~1 year of age livers were sampled. Lipid infiltration and molecular indices of gluconeogenesis, lipid metabolism and mitochondrial function were measured. Although hepatic triglyceride accumulation was not affected by obesity per se, it was nearly doubled in obese offspring born to nutrient-restricted mothers. This adaptation was accompanied by elevated gene expression for peroxisome proliferator-activated receptor γ (PPARG) and its co-activator PGC1α, which may be indicative of changes in the rate of hepatic fatty acid oxidation. In contrast, maternal diet had no influence on the stimulatory effect of obesity on gene expression for a range of proteins involved in glucose metabolism and energy balance including glucokinase, glucocorticoid receptors and uncoupling protein 2. Similarly, although gene expressions for the insulin and IGF1 receptors were suppressed by obesity they were not influenced by the prenatal nutritional environment. In conclusion, excess hepatic lipid accumulation with juvenile obesity is promoted by suboptimal nutrition coincident with early development of the fetal liver.