Multi-Modal Convolutional Parameterisation Network for Guided Image Inverse Problems.

There are several image inverse tasks, such as inpainting or super-resolution, which can be solved using deep internal learning, a paradigm that involves employing deep neural networks to find a solution by learning from the sample itself rather than a dataset. For example, Deep Image Prior is a technique based on fitting a convolutional neural network to output the known parts of the image (such as non-inpainted regions or a low-resolution version of the image). However, this approach is not well adjusted for samples composed of multiple modalities. In some domains, such as satellite image processing, accommodating multi-modal representations could be beneficial or even essential. In this work, Multi-Modal Convolutional Parameterisation Network (MCPN) is proposed, where a convolutional neural network approximates shared information between multiple modes by combining a core shared network with modality-specific head networks. The results demonstrate that these approaches can significantly outperform the single-mode adoption of a convolutional parameterisation network on guided image inverse problems of inpainting and super-resolution.

Driving following a diagnosis of dementia: Exploring the views and experiences of people with dementia-A UK survey.

BACKGROUND: Driving cessation can be one of the adjustments made following a diagnosis of dementia. Little is known about the views and opinions of people living with dementia about this. The study aimed to gather a broad idea of the expectations, impacts and the process of driving cessation from the perspective of those living with dementia. METHODS: 138 people with dementia and 91 relatives/friends (on behalf of an individual with dementia) took part in an online questionnaire. RESULTS: People living with dementia reported stopping driving following diagnosis can have negative psychological impacts particularly in relation to; feelings of isolation, depression, loss of freedom and feeling life isn't worth living. Age, gender and choice in the driving cessation process were related to the degree of negative experiences. CONCLUSIONS: The difficulties reported by people with dementia suggest a need to provide more structured post diagnostic support to aid decision making of driving continuation or cessation; with the view to reducing associated distress and enabling people with dementia to continue to live a meaningful life.

Behavioural Classification of Cattle Using Neck-Mounted Accelerometer-Equipped Collars.

Monitoring and classification of dairy cattle behaviours is essential for optimising milk yields. Early detection of illness, days before the critical conditions occur, together with automatic detection of the onset of oestrus cycles is crucial for obviating prolonged cattle treatments and improving the pregnancy rates. Accelerometer-based sensor systems are becoming increasingly popular, as they are automatically providing information about key cattle behaviours such as the level of restlessness and the time spent ruminating and eating, proxy measurements that indicate the onset of heat events and overall welfare, at an individual animal level. This paper reports on an approach to the development of algorithms that classify key cattle states based on a systematic dimensionality reduction process through two feature selection techniques. These are based on Mutual Information and Backward Feature Elimination and applied on knowledge-specific and generic time-series extracted from raw accelerometer data. The extracted features are then used to train classification models based on a Hidden Markov Model, Linear Discriminant Analysis and Partial Least Squares Discriminant Analysis. The proposed feature engineering methodology permits model deployment within the computing and memory restrictions imposed by operational settings. The models were based on measurement data from 18 steers, each animal equipped with an accelerometer-based neck-mounted collar and muzzle-mounted halter, the latter providing the truthing data. A total of 42 time-series features were initially extracted and the trade-off between model performance, computational complexity and memory footprint was explored. Results show that the classification model that best balances performance and computation complexity is based on Linear Discriminant Analysis using features selected through Backward Feature Elimination. The final model requires 1.83 ± 1.00 ms to perform feature extraction with 0.05 ± 0.01 ms for inference with an overall balanced accuracy of 0.83.

Classification of Cattle Behaviours Using Neck-Mounted Accelerometer-Equipped Collars and Convolutional Neural Networks.

Monitoring cattle behaviour is core to the early detection of health and welfare issues and to optimise the fertility of large herds. Accelerometer-based sensor systems that provide activity profiles are now used extensively on commercial farms and have evolved to identify behaviours such as the time spent ruminating and eating at an individual animal level. Acquiring this information at scale is central to informing on-farm management decisions. The paper presents the development of a Convolutional Neural Network (CNN) that classifies cattle behavioural states ('rumination', 'eating' and 'other') using data generated from neck-mounted accelerometer collars. During three farm trials in the United Kingdom (Easter Howgate Farm, Edinburgh, UK), 18 steers were monitored to provide raw acceleration measurements, with ground truth data provided by muzzle-mounted pressure sensor halters. A range of neural network architectures are explored and rigorous hyper-parameter searches are performed to optimise the network. The computational complexity and memory footprint of CNN models are not readily compatible with deployment on low-power processors which are both memory and energy constrained. Thus, progressive reductions of the CNN were executed with minimal loss of performance in order to address the practical implementation challenges, defining the trade-off between model performance versus computation complexity and memory footprint to permit deployment on micro-controller architectures. The proposed methodology achieves a compression of 14.30 compared to the unpruned architecture but is nevertheless able to accurately classify cattle behaviours with an overall F1 score of 0.82 for both FP32 and FP16 precision while achieving a reasonable battery lifetime in excess of 5.7 years.

The Internet of Things enhancing animal welfare and farm operational efficiency.

The growth in wirelessly enabled sensor network technologies has enabled the low cost deployment of sensor platforms with applications in a range of sectors and communities. In the agricultural domain such sensors have been the foundation for the creation of decision support tools that enhance farm operational efficiency. This Research Reflection illustrates how these advances are assisting dairy farmers to optimise performance and illustrates where emerging sensor technology can offer additional benefits. One of the early applications for sensor technology at an individual animal level was the accurate identification of cattle entering into heat (oestrus) to increase the rate of successful pregnancies and thus optimise milk yield per animal. This was achieved through the use of activity monitoring collars and leg tags. Additional information relating to the behaviour of the cattle, namely the time spent eating and ruminating, was subsequently derived from collars giving further insights of economic value into the wellbeing of the animal, thus an enhanced range of welfare related services have been provisioned. The integration of the information from neck-mounted collars with the compositional analysis data of milk measured at a robotic milking station facilitates the early diagnosis of specific illnesses such as mastitis. The combination of different data streams also serves to eliminate the generation of false alarms, improving the decision making capability. The principle of integrating more data streams from deployed on-farm systems, for example, with feed composition data measured at the point of delivery using instrumented feeding wagons, supports the optimisation of feeding strategies and identification of the most productive animals. Optimised feeding strategies reduce operational costs and minimise waste whilst ensuring high welfare standards. These IoT-inspired solutions, made possible through Internet-enabled cloud data exchange, have the potential to make a major impact within farming practices. This paper gives illustrative examples and considers where new sensor technology from the automotive industry may also have a role.

Contrasting effects of high-starch and high-sugar diets on ruminal function in cattle.

The experiment reported in this research paper aimed to determine whether clinical and subclinical effects on cattle were similar if provided with isoenergetic and isonitrogenous challenge diets in which carbohydrate sources were predominantly starch or sugar. The study was a 3 × 3 Latin square using six adult Jersey cows with rumen cannulae, over 9 weeks. In the first 2 weeks of each 3 week experimental period cows were fed with a maintenance diet and, in the last week, each animal was assigned to one of three diets: a control diet (CON), being a continuation of the maintenance diet; a high starch (HSt) or a high sugar (HSu) diet. Reticuloruminal pH and motility were recorded throughout the study period. Blood and ruminal samples were taken on day-1 (TP-1), day-2 (TP-2) and day-7 (TP-7) of each challenge week. Four clinical variables were recorded daily: diarrhoea, inappetence, depression and ruminal tympany. The effects of treatment, hour of day and day after treatment on clinical parameters were analysed using linear mixed effects (LME) models. Although both challenge diets resulted in a decline in pH, an increase in the absolute pH residuals and an increase in the number of minutes per day under pH 5.8, systemic inflammation was only detected with the HSt diet. The challenge diets differentially modified amplitude and period of reticuloruminal contractions compared with CON diet and both were associated with an increased probability of diarrhoea. The HSu diet reduced the probability of an animal consuming its complete allocation. Because the challenge diets were derived from complex natural materials (barley and molasses respectively), it is not possible to assign all the differential effects to the difference in starch and sugar concentration: non-starch components of barley or non-sugar components of molasses might have contributed to some of the observations. In conclusion, substituting much of the starch with sugar caused no substantial reduction in the acidosis load, but inflammatory response was reduced while feed rejection was increased.

Response Inhibition and Internet Gaming Disorder: A Meta-analysis.

Previous research has demonstrated that Internet Gaming Disorder (IGD) has multiple negative effects in psychological functioning and health. This makes the identification of its underpinnings, such as response inhibition, essential for the development of relevant interventions that target these core features of the disorder resulting in more effective treatment. Several empirical studies have evaluated the relationship between response inhibition deficits and IGD using neurocognitive tasks, but provided mixed results. In this study, we conducted a meta-analysis of studies using three neurocognitive tasks, the Go/No Go, the Stroop, and the Stop-Signal tasks, to integrate existing research and estimate the magnitude of this relationship. We found a medium overall effect size (d=0.56, 95% CI [0.32, 0.80]) indicating that compared with healthy individuals, individuals with IGD are more likely to exhibit impaired response inhibition. This finding is in alignment with literature on inhibition and addictive and impulsive behaviors, as well as with neuroimaging research. Theoretical implications regarding the conceptualization of IGD as a clinical disorder, shared commonalities with externalizing psychopathology, and clinical implications for treatment are discussed.

An evidence-based algorithm for the utility of FDG-PET for diagnosing Alzheimer's disease according to presence of medial temporal lobe atrophy.

BACKGROUND: Imaging biomarkers for Alzheimer's disease include medial temporal lobe atrophy (MTLA) depicted on computed tomography (CT) or magnetic resonance imaging (MRI) and patterns of reduced metabolism on fluorodeoxyglucose positron emission tomography (FDG-PET). AIMS: To investigate whether MTLA on head CT predicts the diagnostic usefulness of an additional FDG-PET scan. METHOD: Participants had a clinical diagnosis of Alzheimer's disease (n = 37) or dementia with Lewy bodies (DLB; n = 30) or were similarly aged controls (n = 30). We visually rated MTLA on coronally reconstructed CT scans and, separately and blind to CT ratings, abnormal appearances on FDG-PET scans. RESULTS: Using a pre-defined cut-off of MTLA ⩾5 on the Scheltens (0-8) scale, 0/30 controls, 6/30 DLB and 23/30 Alzheimer's disease had marked MTLA. FDG-PET performed well for diagnosing Alzheimer's disease v DLB in the low-MTLA group (sensitivity/specificity of 71%/79%), but in the high-MTLA group diagnostic performance of FDG-PET was not better than chance. CONCLUSIONS: In the presence of a high degree of MTLA, the most likely diagnosis is Alzheimer's disease, and an FDG-PET scan will probably not provide significant diagnostic information. However, in cases without MTLA, if the diagnosis is unclear, an FDG-PET scan may provide additional clinically useful diagnostic information.

A double-blind randomized placebo-controlled withdrawal trial comparing memantine and antipsychotics for the long-term treatment of function and neuropsychiatric symptoms in people with Alzheimer's disease (MAIN-AD).

BACKGROUND: Neuropsychiatric symptoms in Alzheimer disease (AD) cause significant distress and present a complex clinical challenge for treatment. Pharmacological treatment options are limited to antipsychotics, which carry extensive safety issues. There is emerging evidence to support the potential benefits of memantine, currently licensed for moderate to severe AD, in the prophylaxis of neuropsychiatric symptoms. METHODS: The MAIN-AD study is a double-blind randomized placebo-controlled withdrawal trial comparing memantine with antipsychotics for the treatment of neuropsychiatric symptoms over 24 weeks. A total of 199 people with probable AD living in care homes already receiving an antipsychotic were randomized to receive either memantine or to continue an antipsychotic. The primary outcomes were function (Bristol Activities of Daily Living Scale [BADLS]) and agitation (Cohen-Mansfield Agitation Inventory [CMAI]). Secondary outcomes were Neuropsychiatric Inventory (NPI), Mini-Mental State Examination (MMSE), and mortality. RESULTS: There was no significant difference between groups on the BADLS or CMAI. At 24 weeks, there was a nonsignificant adjusted difference in favor of memantine on the BADLS of 0.23 (95% CI -1.80-2.27; P = .82) and in favor of antipsychotic on the CMAI of 0.09 (95% CI -0.35-8.53; P = .07). Although there were no significant differences in total NPI, there were 5.01 (95% CI -1.68-11.70; P = .05) and 3.63 (95% CI -1.40-8.67; P = .16) point advantages favoring antipsychotics at weeks 12 and 24, respectively. In addition, in an exploratory analysis, individuals allocated to antipsychotics were significantly less likely to experience relapse of neuropsychiatric symptoms at all time points. The group receiving memantine had a nonsignificant 1.3-point advantage on the MMSE at 24 weeks. DISCUSSION: This study indicates no benefits for memantine in the long-term treatment and prophylaxis of clinically significant neuropsychiatric symptoms. The results did indicate some benefits for antipsychotic medications in reducing the relapse of neuropsychiatric symptoms, but this must be balanced against increased mortality risk.

18F-FDG PET and perfusion SPECT in the diagnosis of Alzheimer and Lewy body dementias.

UNLABELLED: Brain imaging with glucose ((18)F-FDG) PET or blood flow (hexamethylpropyleneamine oxime) SPECT is widely used for the differential diagnosis of dementia, though direct comparisons to clearly establish superiority of one method have not been undertaken. METHODS: Subjects with Alzheimer disease (AD; n = 38) and dementia with Lewy bodies (DLB; n = 30) and controls (n = 30) underwent (18)F-FDG PET and SPECT in balanced order. The main outcome measure was area under the curve (AUC) of receiver-operating-characteristic analysis of visual scan rating. RESULTS: Consensus diagnosis with (18)F-FDG PET was superior to SPECT for both dementia vs. no-dementia (AUC = 0.93 vs. 0.72, P = 0.001) and AD vs. DLB (AUC = 0.80 vs. 0.58, P = 0.005) comparisons. The sensitivity and specificity for dementia/no-dementia was 85% and 90%, respectively, for (18)F-FDG PET and 71% and 70%, respectively, for SPECT. CONCLUSION: (18)F-FDG PET was significantly superior to blood flow SPECT. We recommend (18)F-FDG PET be performed instead of perfusion SPECT for the differential diagnosis of degenerative dementia if functional imaging is indicated.

A comparison of FDG-PET and blood flow SPECT in the diagnosis of neurodegenerative dementias: a systematic review.

OBJECTIVE: Perfusion single photon emission computed tomography (SPECT) and 18F fluorodeoxyglucose positron emission tomography (FDG-PET) both have clinical utility for the differential diagnosis of dementia. Although PET is often viewed by some as more accurate and therefore preferential, the extent to which published evidence supports this is not clear. The aim of this review was to address the question by reviewing studies of SPECT and PET imaging in dementia diagnosis, with a particular focus on all published head-to-head studies. DESIGN: A MEDLINE search was carried out using the following keywords: "PET" and "SPECT" and "dementia" or "Mild Cognitive Impairment," together with "alzheimers" or "DLB" or "lewy body" or "frontotemporal" or "FTD" or "Picks." Articles were included up to February 2013, limited to human studies and in English language. RESULTS: Published studies of SPECT accuracy show that it is a useful tool for differential diagnosis, with sensitivities of 65-85% for diagnosing Alzheimer's disease (AD) and specificities (for other neurodegenerative dementias) of 72-87%. PET studies generally report higher accuracy, with sensitivities of 75-99% for AD and specificities of 71-93%. However, there have been few direct head-to-head comparisons, with some indicating SPECT and PET to be equally useful in dementia diagnosis and others favouring PET. Many of these studies are limited with respect to numbers and methodically with poorly matched control groups. CONCLUSIONS: Overall, although studies suggest superiority of PET over SPECT, the evidence base for this is actually quite limited. We suggest that further direct comparative studies, including health economic and patient preference evaluations, are needed to help direct future service provision.

Multivariate spatial covariance analysis of 99mTc-exametazime SPECT images in dementia with Lewy bodies and Alzheimer's disease: utility in differential diagnosis.

We examined (99m)Tc-exametazime brain blood flow single-photon emission computed tomography (SPECT) images using a spatial covariance analysis (SCA) approach to assess its diagnostic value in distinguishing dementia with Lewy bodies (DLB) from Alzheimer's disease (AD). Voxel SCA was simultaneously applied to a set of preprocessed images (AD, n=40; DLB, n=26), generating a series of eigenimages representing common intercorrelated voxels in AD and DLB. Linear regression derived a spatial covariance pattern (SCP) that discriminated DLB from AD. To investigate the diagnostic value of the model SCP, the SCP was validated by applying it to a second, independent, AD and DLB cohort (AD, n=34; DLB, n=29). Mean SCP expressions differed between AD and DLB (F(1,64)=36.2, P<0.001) with good diagnostic accuracy (receiver operating characteristic (ROC) curve area 0.87, sensitivity 81%, specificity 88%). Forward application of the model SCP to the independent cohort revealed similar differences between groups (F(1,61)=38.4, P<0.001), also with good diagnostic accuracy (ROC 0.86, sensitivity 80%, specificity 80%). Multivariate analysis of blood flow SPECT data appears to be robust and shows good diagnostic accuracy in two independent cohorts for distinguishing DLB from AD.

Low dose lipid formulations: effects on gastric emptying and biliary secretion.

PURPOSE: Food stimulates changes to gastrointestinal secretion and motility patterns, however, the effect of smaller quantities of lipid, such as that contained in a lipid-based drug formulation, has not been detailed. This study aimed to examine the effects of small quantities of lipid on gastric emptying and biliary secretion. METHODS: The influence of oral administration of three lipid-based formulations and a negative control formulation on gastric emptying and biliary secretion was evaluated in 16 healthy male subjects using gamma scintigraphy, ultrasonography and duodenal aspiration. RESULTS: Low quantities (2 g) of long chain lipid stimulated gall bladder contraction and elevated intestinal bile salt, phospholipid and cholesterol levels. Changes in gastric emptying were also evident, although these did not reach statistical significance. Administration of a similar quantity of medium chain lipid, however, had little effect on gastric emptying and gallbladder contraction and did not stimulate appreciable increases in intestinal concentrations of biliary-derived lipids. CONCLUSIONS: The quantities of long chain lipid that might be administered in a pharmaceutical formulation stimulate gallbladder contraction and elevate intestinal levels of bile salt and phospholipid. This effect is a likely contributor to the ability of lipid based formulations to enhance the absorption of poorly water-soluble drugs.