Data Mining and Fusion Framework for In-Home Monitoring Applications

Sensor Data Fusion (SDT) algorithms and models have been widely used in diverse applications. One of the main challenges of SDT includes how to deal with heterogeneous and complex datasets with different formats. The present work utilised both homogenous and heterogeneous datasets to propose a novel SDT framework. It compares data mining-based fusion software packages such as RapidMiner Studio, Anaconda, Weka, and Orange, and proposes a data fusion framework suitable for in-home applications. A total of 574 privacy-friendly (binary) images and 1722 datasets gleaned from thermal and Radar sensing solutions, respectively, were fused using the software packages on instances of homogeneous and heterogeneous data aggregation. Experimental results indicated that the proposed fusion framework achieved an average Classification Accuracy of 84.7% and 95.7% on homogeneous and heterogeneous datasets, respectively, with the help of data mining and machine learning models such as Naïve Bayes, Decision Tree, Neural Network, Random Forest, Stochastic Gradient Descent, Support Vector Machine, and CN2 Induction. Further evaluation of the Sensor Data Fusion framework based on cross-validation of features indicated average values of 94.4% for Classification Accuracy, 95.7% for Precision, and 96.4% for Recall. The novelty of the proposed framework includes cost and timesaving advantages for data labelling and preparation, and feature extraction.

ST6GAL1-mediated aberrant sialylation promotes prostate cancer progression

Aberrant glycosylation is a universal feature of cancer cells, and cancer-associated glycans have been detected in virtually every cancer type. A common change in tumour cell glycosylation is an increase in α2,6 sialylation of N-glycans, a modification driven by the sialyltransferase ST6GAL1. ST6GAL1 is overexpressed in numerous cancer types, and sialylated glycans are fundamental for tumour growth, metastasis, immune evasion, and drug resistance, but the role of ST6GAL1 in prostate cancer is poorly understood. Here, we analyse matched cancer and normal tissue samples from 200 patients and verify that ST6GAL1 is upregulated in prostate cancer tissue. Using MALDI imaging mass spectrometry (MALDI-IMS), we identify larger branched α2,6 sialylated N-glycans that show specificity to prostate tumour tissue. We also monitored ST6GAL1 in plasma samples from >400 patients and reveal ST6GAL1 levels are significantly increased in the blood of men with prostate cancer. Using both in vitro and in vivo studies, we demonstrate that ST6GAL1 promotes prostate tumour growth and invasion. Our findings show ST6GAL1 introduces α2,6 sialylated N-glycans on prostate cancer cells and raise the possibility that prostate cancer cells can secrete active ST6GAL1 enzyme capable of remodelling glycans on the surface of other cells. Furthermore, we find α2,6 sialylated N-glycans expressed by prostate cancer cells can be targeted using the sialyltransferase inhibitor P-3FAX -Neu5Ac. Our study identifies an important role for ST6GAL1 and α2,6 sialylated N-glycans in prostate cancer progression and highlights the opportunity to inhibit abnormal sialylation for the development of new prostate cancer therapeutics. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Upregulation of GALNT7 in prostate cancer modifies O-glycosylation and promotes tumour growth

Prostate cancer is the most common cancer in men and it is estimated that over 350,000 men worldwide die of prostate cancer every year. There remains an unmet clinical need to improve how clinically significant prostate cancer is diagnosed and develop new treatments for advanced disease. Aberrant glycosylation is a hallmark of cancer implicated in tumour growth, metastasis, and immune evasion. One of the key drivers of aberrant glycosylation is the dysregulated expression of glycosylation enzymes within the cancer cell. Here, we demonstrate using multiple independent clinical cohorts that the glycosyltransferase enzyme GALNT7 is upregulated in prostate cancer tissue. We show GALNT7 can identify men with prostate cancer, using urine and blood samples, with improved diagnostic accuracy than serum PSA alone. We also show that GALNT7 levels remain high in progression to castrate-resistant disease, and using in vitro and in vivo models, reveal that GALNT7 promotes prostate tumour growth. Mechanistically, GALNT7 can modify O-glycosylation in prostate cancer cells and correlates with cell cycle and immune signalling pathways. Our study provides a new biomarker to aid the diagnosis of clinically significant disease and cements GALNT7-mediated O-glycosylation as an important driver of prostate cancer progression.

Gene expression analysis reveals a 5-gene signature for progression-free survival in prostate cancer

Prostate cancer (PCa) is the second most common male cancer worldwide, but effective biomarkers for the presence or progression risk of disease are currently elusive. In a series of nine matched histologically confirmed PCa and benign samples, we carried out an integrated transcriptome-wide gene expression analysis, including differential gene expression analysis and weighted gene co-expression network analysis (WGCNA), which identified a set of potential gene markers highly associated with tumour status (malignant vs. benign). We then used these genes to establish a minimal progression-free survival (PFS)-associated gene signature (GS) (PCBP1, PABPN1, PTPRF, DANCR, and MYC) using least absolute shrinkage and selection operator (LASSO) and stepwise multivariate Cox regression analyses from The Cancer Genome Atlas prostate adenocarcinoma (TCGA-PRAD) dataset. Our signature was able to predict PFS over 1, 3, and 5 years in TCGA-PRAD dataset, with area under the curve (AUC) of 0.64–0.78, and our signature remained as a prognostic factor independent of age, Gleason score, and pathological T and N stages. A nomogram combining the signature and Gleason score demonstrated improved predictive capability for PFS (AUC: 0.71–0.85) and was superior to the Cambridge Prognostic Group (CPG) model alone and some conventionally used clinicopathological factors in predicting PFS. In conclusion, we have identified and validated a novel five-gene signature and established a nomogram that effectively predicted PFS in patients with PCa. Findings may improve current prognosis tools for PFS and contribute to clinical decision-making in PCa treatment.

A likelihood analysis of quantile-matching transformations

Quantile matching is a strictly monotone transformation that sends the observed response values to the quantiles of a given target distribution. A profile likelihood-based criterion is developed for comparing one target distribution with another in a linear-model setting.

Wearable technology-based metrics for predicting operator performance during cardiac catheterisation

Introduction

Unobtrusive metrics that can auto-assess performance during clinical procedures are of value. Three approaches to deriving wearable technology-based metrics are explored: (1) eye tracking, (2) psychophysiological measurements [e.g. electrodermal activity (EDA)] and (3) arm and hand movement via accelerometry. We also measure attentional capacity by tasking the operator with an additional task to track an unrelated object during the procedure.

Methods

Two aspects of performance are measured: (1) using eye gaze and psychophysiology metrics and (2) measuring attentional capacity via an additional unrelated task (to monitor a visual stimulus/playing cards). The aim was to identify metrics that can be used to automatically discriminate between levels of performance or at least between novices and experts. The study was conducted using two groups: (1) novice operators and (2) expert operators. Both groups made two attempts at a coronary angiography procedure using a full-physics virtual reality simulator. Participants wore eye tracking glasses and an E4 wearable wristband. Areas of interest were defined to track visual attention on display screens, including: (1) X-ray, (2) vital signs, (3) instruments and (4) the stimulus screen (for measuring attentional capacity).

Results

Experts provided greater dwell time (63% vs 42%, p = 0.03) and fixations (50% vs 34%, p = 0.04) on display screens. They also provided greater dwell time (11% vs 5%, p = 0.006) and fixations (9% vs 4%, p = 0.007) when selecting instruments. The experts’ performance for tracking the unrelated object during the visual stimulus task negatively correlated with total errors (r = − 0.95, p = 0.0009). Experts also had a higher standard deviation of EDA (2.52 µS vs 0.89 µS, p = 0.04).

Conclusions

Eye tracking metrics may help discriminate between a novice and expert operator, by showing that experts maintain greater visual attention on the display screens. In addition, the visual stimulus study shows that an unrelated task can measure attentional capacity.

Trial registration This work is registered through clinicaltrials.gov, a service of the U.S. National Health Institute, and is identified by the trial reference: NCT02928796.

Androgen-regulated transcription of ESRP2 drives alternative splicing patterns in prostate cancer

Prostate is the most frequent cancer in men. Prostate cancer progression is driven by androgen steroid hormones, and delayed by androgen deprivation therapy (ADT). Androgens control transcription by stimulating androgen receptor (AR) activity, yet also control pre-mRNA splicing through less clear mechanisms. Here we find androgens regulate splicing through AR-mediated transcriptional control of the epithelial-specific splicing regulator ESRP2. Both ESRP2 and its close paralog ESRP1 are highly expressed in primary prostate cancer. Androgen stimulation induces splicing switches in many endogenous ESRP2-controlled mRNA isoforms, including splicing switches correlating with disease progression. ESRP2 expression in clinical prostate cancer is repressed by ADT, which may thus inadvertently dampen epithelial splice programmes. Supporting this, treatment with the AR antagonist bicalutamide (Casodex) induced mesenchymal splicing patterns of genes including FLNB and CTNND1. Our data reveals a new mechanism of splicing control in prostate cancer with important implications for disease progression.

Rich context information for just-in-time adaptive intervention promoting physical activity

Sedentary lifestyle and inadequate levels of physical activity represent two serious health risk factors. Nevertheless, within developed countries, 60% of people aged over 60 are deemed to be sedentary. Consequently, interest in behavior change to promote physical activity is increasing. In particular, the role of emerging mobile apps to facilitate behavior change has shown promising results. Smart technologies can help in providing rich context information including an objective assessment of the level of physical activity and information on the emotional and physiological state of the person. Collectively, this can be used to develop innovative persuasive solutions for adaptive behavior change. Such solutions offer potential in reducing levels of sedentary behavior. This work presents a study exploring new ways of employing smart technologies to facilitate behavior change. It is achieved by means of (i) developing a knowledge base on sedentary behaviors and recommended physical activity guidelines, and (ii) a context model able to combine information on physical activity, location, and a user's diary to develop a context-aware virtual coach with the ability to select the most appropriate behavior change strategy on a case by case basis.

A Personalized Self-Management Rehabilitation System for Stroke Survivors: A Quantitative Gait Analysis Using a Smart Insole

BACKGROUND

In the United Kingdom, stroke is the single largest cause of adult disability and results in a cost to the economy of £8.9 billion per annum. Service needs are currently not being met; therefore, initiatives that focus on patient-centered care that promote long-term self-management for chronic conditions should be at the forefront of service redesign. The use of innovative technologies and the ability to apply these effectively to promote behavior change are paramount in meeting the current challenges.

OBJECTIVE

Our objective was to gain a deeper insight into the impact of innovative technologies in support of home-based, self-managed rehabilitation for stroke survivors. An intervention of daily walks can assist with improving lower limb motor function, and this can be measured by using technology. This paper focuses on assessing the usage of self-management technologies on poststroke survivors while undergoing rehabilitation at home.

METHODS

A realist evaluation of a personalized self-management rehabilitation system was undertaken in the homes of stroke survivors (N=5) over a period of approximately two months. Context, mechanisms, and outcomes were developed and explored using theories relating to motor recovery. Participants were encouraged to self-manage their daily walking activity; this was achieved through goal setting and motivational feedback. Gait data were collected and analyzed to produce metrics such as speed, heel strikes, and symmetry. This was achieved using a "smart insole" to facilitate measurement of walking activities in a free-living, nonrestrictive environment.

RESULTS

Initial findings indicated that 4 out of 5 participants performed better during the second half of the evaluation. Performance increase was evident through improved heel strikes on participants' affected limb. Additionally, increase in performance in relation to speed was also evident for all 5 participants. A common strategy emerged across all but one participant as symmetry performance was sacrificed in favor of improved heel strikes. This paper evaluates compliance and intensity of use.

CONCLUSION

Our findings suggested that 4 out of the 5 participants improved their ability to heel strike on their affected limb. All participants showed improvements in their speed of gait measured in steps per minute with an average increase of 9.8% during the rehabilitation program. Performance in relation to symmetry showed an 8.5% average decline across participants, although 1 participant improved by 4%. Context, mechanism, and outcomes indicated that dual motor learning and compensatory strategies were deployed by the participants.

The androgen receptor controls expression of the cancer-associated sTn antigen and cell adhesion through induction of ST6GalNAc1 in prostate cancer

Patterns of glycosylation are important in cancer, but the molecular mechanisms that drive changes are often poorly understood. The androgen receptor drives prostate cancer (PCa) development and progression to lethal metastatic castration-resistant disease. Here we used RNA-Seq coupled with bioinformatic analyses of androgen-receptor (AR) binding sites and clinical PCa expression array data to identify ST6GalNAc1 as a direct and rapidly activated target gene of the AR in PCa cells. ST6GalNAc1 encodes a sialytransferase that catalyses formation of the cancer-associated sialyl-Tn antigen (sTn), which we find is also induced by androgen exposure. Androgens induce expression of a novel splice variant of the ST6GalNAc1 protein in PCa cells. This splice variant encodes a shorter protein isoform that is still fully functional as a sialyltransferase and able to induce expression of the sTn-antigen. Surprisingly, given its high expression in tumours, stable expression of ST6GalNAc1 in PCa cells reduced formation of stable tumours in mice, reduced cell adhesion and induced a switch towards a more mesenchymal-like cell phenotype in vitro. ST6GalNAc1 has a dynamic expression pattern in clinical datasets, beingsignificantly up-regulated in primary prostate carcinoma but relatively down-regulated in established metastatic tissue. ST6GalNAc1 is frequently upregulated concurrently with another important glycosylation enzyme GCNT1 previously associated with prostate cancer progression and implicated in Sialyl Lewis X antigen synthesis. Together our data establishes an androgen-dependent mechanism for sTn antigen expression in PCa, and are consistent with a general role for the androgen receptor in driving important coordinate changes to the glycoproteome during PCa progression.

The PI3K regulatory subunit gene PIK3R1 is under direct control of androgens and repressed in prostate cancer cells

Androgen receptor (AR) signalling and the PI3K pathway mediate survival signals in prostate cancer, and have been shown to regulate each other by reciprocal negative feedback, such that inhibition of one activates the other. Understanding the reciprocal regulation of these pathways is important for disease management as tumour cells can adapt and survive when either single pathway is inhibited pharmacologically. We recently carried out genome-wide exon-specific profiling of prostate cancer cells to identify novel androgen-regulated transcriptional events. Here we interrogated this dataset for novel androgen-regulated genes associated with the PI3K pathway. We find that the PI3K regulatory subunits PIK3R1 (p85α) and PIK3R3 (p55γ) are direct targets of the AR which are rapidly repressed by androgens in LNCaP cells. Further characterisation revealed that the PIK3CA p110α catalytic subunit is also indirectly regulated by androgens at the protein level. We show that PIK3R1 mRNA is significantly under-expressed in prostate cancer (PCa) tissue, and provide data to suggest a context-dependent regulatory mechanism whereby repression of the p85α protein by the AR results in destabilisation of the PI3K p110α catalytic subunit and downstream PI3K pathway inhibition that functionally affects the properties of prostate cancer cells.

Flexible context aware interface for ambient assisted living

A Multi Agent System that provides a (cared for) person, the subject, with assistance and support through an Ambient Assisted Living Flexible Interface (AALFI) during the day while complementing the night time assistance offered by NOCTURNAL with feedback assistance, is presented. It has been tailored to the subject’s requirements profile and takes into account factors associated with the time of day; hence it attempts to overcome shortcomings of current Ambient Assisted Living Systems. The subject is provided with feedback that highlights important criteria such as quality of sleep during the night and possible breeches of safety during the day. This may help the subject carry out corrective measures and/or seek further assistance. AALFI provides tailored interaction that is either visual or auditory so that the subject is able to understand the interactions and this process is driven by a Multi-Agent System. User feedback gathered from a relevant user group through a workshop validated the ideas underpinning the research, the Multi-agent system and the adaptable interface.

Qualitative evaluation of the SMART2 self-management system for people in chronic pain

PURPOSE

Technology could support the self-management of long-term health conditions such as chronic pain. This article describes an evaluation of SMART2, a personalised self-management system incorporating activity planning and review, feedback on behaviour- and acceptance-based therapeutic exercises.

METHOD

The SMART2 system was evaluated over a four-week trial in the homes of people in chronic pain. At conclusion, participants were interviewed to understand the experience of using and living with the SMART2 system as a therapeutic tool.

RESULTS

Qualitative analysis of the interviews found that participants liked the system and reported making associated changes to their behaviour. Goal setting and feedback were the most useful elements of the system. A third key and unexpected element was that by simulating some of the functions of a therapist, SMART2 also simulated some of the process of interacting with a therapist.

CONCLUSIONS

People in chronic pain may experience positive outcomes when using a self-management system designed for behaviour change. Furthermore, some of the supportive aspects of the therapeutic context can be elicited by self-management technologies. Implications of Rehabilitation Self-management technology has the potential to assist rehabilitation by supporting goal setting and providing feedback. By simulating some of the functions of a therapist, technology can simulate some of the process of therapy during rehabilitation. People in chronic pain liked using the technology in their own home and thought it could augment services delivered by clinical practitioners. Complex programmes of therapeutic exercises delivered by technology had limited success in engaging people in chronic pain.

Science versus design; comparable, contrastive or conducive?

Science and design are two completely separated areas of expertise with their own specialists. Science analyses the existing world to create new knowledge, design uses existing knowledge to create a new world. This tunnel-vision mentality and narrow-minded approach is dangerous for problem solving, where a broad view on potential solutions is required to realise a high-quality answer on the defined problem. We state that design benefits from scientific methods, resulting in a more effective design process and in better products, while science benefits from a design approach, resulting in more efficient and effective results. Our philosophy is illustrated using examples from the field of biomedical engineering. Both methods can benefit tremendously from each other. By applying scientific methods, superior choices will be made in the design process. With design, more accurate, effective and efficient science will be performed.

Data and Information Quality Issues in Ambient Assisted Living Systems

Demographic aging, as a result of people living for longer, has put an increased burden on health and social care provision across most of the economies of the developed and developing world. In order to cope with the greater numbers of older people, together with increasing prevalence of chronic diseases, governments are looking to new ways to provide care and support to older people and their care providers. A growing trend is where health and social care providers are moving towards the use of assisted living technologies to provide care and assistance in the home. In this article, the research area of Ambient Assisted Living (AAL) systems is examined and the data, information and the higher-level contextual knowledge quality issues in relation to these systems, is discussed. Lack of quality control may result in an AAL system providing assistance and support based upon incorrect data, information and knowledge inputs, and this may have a detrimental effect on the person making use of the system. We propose a model whereby contextual knowledge gained during the AAL system’s reasoning cycle can be fed back to aid in further quality checking at the various architectural layers, and a realistic AAL scenario is provided to support this. Future research should be conducted in these areas, with the requirement of building quality criteria into the design and implementation of AAL systems.

Evaluation of connected health technology

With current advances in sensing technology, communication networks and software applications, the use of connected health technology within the home environment has become both more affordable and widespread. Nevertheless, the introduction of this new care paradigm has brought with it many challenges, with one of the most notable being assessing of the impact or otherwise of its usage. The assessment of efficiency, benefit and utility of such technology is recognised as still being in its infancy. Traditional evaluation protocols may fail to address the specific challenges associated with increased use of networks, databases and home deployments, in addition to the multitude of factors influencing successful adoption. This article aims to delineate the required steps of connected health technology evaluations and move towards a common framework that can be used to support future evaluations. A series of recommendations are presented based on previous experience in the domain.

Professional development of health informatics in Northern Ireland

This paper addresses the assessment and verification of health informatics professional competencies. Postgraduate provision in Health Informatics was targeted at informatics professionals working full-time in the National Health Service, in Northern Ireland, United Kingdom. Many informatics health service positions do not require a formal informatics background, and as we strive for professionalism, a recognized qualification provides important underpinning. The course, delivered from a computing perspective, builds upon work-based achievement and provides insight into emerging technologies associated with the 'connected health' paradigm. The curriculum was designed with collaboration from the Northern Ireland Health and Social Care ICT Training Group. Material was delivered by blended learning using a virtual learning environment and face-to-face sessions. Professional accreditation was of high importance. The aim was to provide concurrent qualifications: a postgraduate certificate, awarded by the University of Ulster and a professional certificate validated and accredited by a professional body comprising experienced health informatics professionals. Providing both qualifications puts significant demands upon part-time students, and a balance must be achieved for successful completion.

Feature selection and classification in supporting report-based self-management for people with chronic pain

Chronic pain is a common long-term condition that affects a person's physical and emotional functioning. Currently, the integrated biopsychosocial approach is the mainstay treatment for people with chronic pain. Self-reporting (the use of questionnaires) is one of the most common methods to evaluate treatment outcome. The questionnaires can consist of more than 300 questions, which is tedious for people to complete at home. This paper presents a machine learning approach to analyze self-reporting data collected from the integrated biopsychosocial treatment, in order to identify an optimal set of features for supporting self-management. In addition, a classification model is proposed to differentiate the treatment stages. Four different feature selection methods were applied to rank the questions. In addition, four supervised learning classifiers were used to investigate the relationships between the numbers of questions and classification performance. There were no significant differences between the feature ranking methods for each classifier in overall classification accuracy or AUC ( p > 0.05); however, there were significant differences between the classifiers for each ranking method ( p < 0.001). The results showed the multilayer perceptron classifier had the best classification performance on an optimized subset of questions, which consisted of ten questions. Its overall classification accuracy and AUC were 100% and 1, respectively.

Alterations in LMTK2, MSMB and HNF1B gene expression are associated with the development of prostate cancer

Background

Genome wide association studies (GWAS) have identified several genetic variants that are associated with prostate cancer. Most of these variants, like other GWAS association signals, are located in non-coding regions of potential candidate genes, and thus could act at the level of the mRNA transcript.

Methods

We measured the expression and isoform usage of seven prostate cancer candidate genes in benign and malignant prostate by real-time PCR, and correlated these factors with cancer status and genotype at the GWAS risk variants.

Results

We determined that levels of LMTK2 transcripts in prostate adenocarcinomas were only 32% of those in benign tissues (p = 3.2 × 10^-7), and that an independent effect of genotype at variant rs6465657 on LMTK2 expression in benign (n = 39) and malignant tissues (n = 21) was also evident (P = 0.002). We also identified that whilst HNF1B(C) and MSMB2 comprised the predominant isoforms in benign tissues (90% and 98% of total HNF1B or MSMB expression), HNF1B(B) and MSMB1 were predominant in malignant tissue (95% and 96% of total HNF1B or MSMB expression; P = 1.7 × 10^-7 and 4 × 10^-4 respectively), indicating major shifts in isoform usage.

Conclusions

Our results indicate that the amount or nature of mRNA transcripts expressed from the LMTK2, HNF1B and MSMB candidate genes is altered in prostate cancer, and provides further evidence for a role for these genes in this disorder. The alterations in isoform usage we detect highlights the potential importance of alternative mRNA processing and moderation of mRNA stability as potentially important disease mechanisms.

Smart self management: assistive technology to support people with chronic disease

We have developed a personalised self management system to support self management of chronic conditions with support from health-care professionals. Accelerometers are used to measure gross levels of activity, for example walking around the house, and used to infer higher level activity states, such as standing, sitting and lying. A smart phone containing an accelerometer and a global positioning system (GPS) module can be used to monitor outdoor activity, providing both activity and location based information. Heart rate, blood pressure and weight are recorded and input to the system by the user. A decision support system (DSS) detects abnormal activity and distinguishes life style patterns. The DSS is used to assess the self management process, and automates feedback to the user, consistent with the achievement of their life goals. We have found that telecare and assistive technology is feasible to support self management for chronic conditions within the home and local community environments.

Can brain computer interfaces become practical assistive devices in the community?

A Brain Computer Interface (BCI) provides direct communication from the brain to a computer or electronic device. In order for BCIs to become practical assistive devices it is necessary to develop robust systems, which can be used outside of the laboratory. This paper appraises the technical challenges, and outlines the design of an intuitive user interface, which can be used for smart device control and entertainment applications, of specific interest to users. We adopted a user-centred approach, surveying two groups of participants: fifteen volunteers who could use BCI as an additional technology and six users with complex communication and assistive technology needs. Interaction is based on a four way choice, parsing a hierarchical menu structure which allows selection of room location and then device (e.g. light, television) within a smart home. The interface promotes ease of use which aim to improve the BCI communication rate.

An unusual case of Epstein-Barr virus driven lymphoproliferative disorder of the conjunctiva which mimicked a high grade lymphoma: a sheep in wolf's clothing

Lymphoid proliferations represent 25-33% of acquired sub-epithelial conjunctival lesions which are excised or biopsied in patients over 15 years of age. These lesions are reported in association with Epstein-Barr virus (EBV). One such case of EBV associated spontaneously regressed monoclonal B cell infiltrate in conjunctiva that mimicked a large B cell lymphoma is reported.

Anti-parental lymphocyte reactions in neonatal F1 hybrid rats

The subpopulation of parental-strain lymphocytes responsible for the recognition of a particular F1 hybrid strain as foreign has been shown to be subject to specific, reversible inactivation after its injection into neonatal rats of that F1 hybrid strain. Neonates born to mothers that were syngeneic with the parental-strain lymphocytes under test acquired the capacity to inactivate these lymphocytes at an earlier age than did the genotypically identical reciprocal F1 hybrids. Neonates had little capacity to inactivate completely allogeneic lymphocytes. It is inferred from the difference in behavior between reciprocal F1 hybrids that the augmented ability to inactivate anti-F1 hybrid maternal-strain lymphocytes follows exposure to such cells in utero and to antibodies with anti-F1 hybrid activity in colostrum. Specific inactivation of those marauding maternal lymphocytes with anti-fetal activity is envisaged as an important means of protection of the fetus from immunological attack by the mother. On the basis of the results presented in this and the preceding paper, it has been proposed that many of the sequelae of the transfer of immunocompetent parental-strain cells to F1 hybrid animals result not from graft anti-host activity but from an F1 hybrid anti-parental lymphocyte response that has eluded normal regulatory mechanisms. These experiments also raise the possibility that regulation of auto-immune responses may be achieved by the inactivation of lymphocytes with anti-self reactivity by other lymphocytes that respond to the recognition structure required for such reactivity.

Feature selection and classification model construction on type 2 diabetic patients’ data

OBJECTIVE

Diabetes affects between 2% and 4% of the global population (up to 10% in the over 65 age group), and its avoidance and effective treatment are undoubtedly crucial public health and health economics issues in the 21st century. The aim of this research was to identify significant factors influencing diabetes control, by applying feature selection to a working patient management system to assist with ranking, classification and knowledge discovery. The classification models can be used to determine individuals in the population with poor diabetes control status based on physiological and examination factors.

METHODS

The diabetic patients' information was collected by Ulster Community and Hospitals Trust (UCHT) from year 2000 to 2004 as part of clinical management. In order to discover key predictors and latent knowledge, data mining techniques were applied. To improve computational efficiency, a feature selection technique, feature selection via supervised model construction (FSSMC), an optimisation of ReliefF, was used to rank the important attributes affecting diabetic control. After selecting suitable features, three complementary classification techniques (Naïve Bayes, IB1 and C4.5) were applied to the data to predict how well the patients' condition was controlled.

RESULTS

FSSMC identified patients' 'age', 'diagnosis duration', the need for 'insulin treatment', 'random blood glucose' measurement and 'diet treatment' as the most important factors influencing blood glucose control. Using the reduced features, a best predictive accuracy of 95% and sensitivity of 98% was achieved. The influence of factors, such as 'type of care' delivered, the use of 'home monitoring', and the importance of 'smoking' on outcome can contribute to domain knowledge in diabetes control.

CONCLUSION

In the care of patients with diabetes, the more important factors identified: patients' 'age', 'diagnosis duration' and 'family history', are beyond the control of physicians. Treatment methods such as 'insulin', 'diet' and 'tablets' (a variety of oral medicines) may be controlled. However lifestyle indicators such as 'body mass index' and 'smoking status' are also important and may be controlled by the patient. This further underlines the need for public health education to aid awareness and prevention. More subtle data interactions need to be better understood and data mining can contribute to the clinical evidence base. The research confirms and to a lesser extent challenges current thinking. Whilst fully appreciating the requirement for clinical verification and interpretation, this work supports the use of data mining as an exploratory tool, particularly as the domain is suffering from a data explosion due to enhanced monitoring and the (potential) storage of this data in the electronic health record. FSSMC has proved a useful feature estimator for large data sets, where processing efficiency is an important factor.

Neuroaffective processing in criminal psychopaths: brain event-related potentials reveal task-specific anomalies

This study aimed to confirm neuroaffective processing deficits in psychopaths by measuring late brain event-related potential (ERP) components and behavior in groups of psychopathic and nonpsychopathic inmates of a Singaporean prison while they performed two tasks. In a Categorization task, affective stimuli were task-relevant and required focused attention, while in a Vigilance task, affective pictures were presented in the background while participants discriminated vertical from oblique lines. Psychopaths showed differences in late positive ERPs that were sensitive to affective stimulus properties (valence and arousal) in the Categorization, but not in the Vigilance task, suggesting that only under conditions of focused attention did psychopaths show a neuroaffective processing deficit. In the Categorization task, psychopaths also showed a significantly larger prefrontal negative ERP (N350) whose amplitude correlated positively with the behavioral facet of psychopathy. In the Vigilance task, psychopaths both missed more targets and showed significantly smaller target-evoked parietal ERPs when viewing arousing pictures, suggesting their attentional focus was disrupted by the affective background.

A comparison of modelling and imagery in the acquisition and retention of motor skills

Although many researchers have examined the effects of imagery and/or modelling interventions, it is unclear which of the two interventions is more effective. In two experiments, novice learners assessed over multiple trials of a free weight squat lifting or a stabilometer balancing task were given modelling, imagery, a combination of modelling and imagery, or control interventions. Group differences indicated, in general, that groups receiving modelling (modelling, combination) evidenced a more appropriate form than groups that did not receive modelling (imagery, control). When apparent, these differences were already in place after the first of several interventions. Practical implications are that even a single bout of modelling can have immediate beneficial effects on movement form (Experiments 1 and 2) and outcome (Experiment 1).

Auditory brainstem response classification: A hybrid model using time and frequency features

OBJECTIVE

The auditory brainstem response (ABR) is an evoked response obtained from brain electrical activity when an auditory stimulus is applied to the ear. An audiologist can determine the threshold level of hearing by applying stimuli at reducing levels of intensity, and can also diagnose various otological, audiological, and neurological abnormalities by examining the morphology of the waveform and the latencies of the individual waves. This is a subjective process requiring considerable expertise. The aim of this research was to develop software classification models to assist the audiologist with an automated detection of the ABR waveform and also to provide objectivity and consistency in this detection.

MATERIALS AND METHODS

The dataset used in this study consisted of 550 waveforms derived from tests using a range of stimulus levels applied to 85 subjects ranging in hearing ability. Each waveform had been classified by a human expert as 'response=Yes' or 'response=No'. Individual software classification models were generated using time, frequency and cross-correlation measures. Classification employed both artificial neural networks (NNs) and the C5.0 decision tree algorithm. Accuracies were validated using six-fold cross-validation, and by randomising training, validation and test datasets.

RESULTS

The result was a two stage classification process whereby strong responses were classified to an accuracy of 95.6% in the first stage. This used a ratio of post-stimulus to pre-stimulus power in the time domain, with power measures at 200, 500 and 900Hz in the frequency domain. In the second stage, outputs from time, frequency and cross-correlation classifiers were combined using the Dempster-Shafer method to produce a hybrid model with an accuracy of 85% (126 repeat waveforms).

CONCLUSION

By combining the different approaches a hybrid system has been created that emulates the approach used by an audiologist in analysing an ABR waveform. Interpretation did not rely on one particular feature but brought together power and frequency analysis as well as consistency of subaverages. This provided a system that enhanced robustness to artefacts while maintaining classification accuracy.

A comparison of supervised classification methods for auditory brainstem response determination

The ABR is commonly used in the Audiology clinic to determine and quantify hearing loss. Its interpretation is subjective, dependent upon the expertise and experience of the clinical scientist. In this study we investigated the role of machine learning for pattern classification in this domain. We extracted features from the ABRs of 85 test subjects (550 waveforms) and compared four complimentary supervised classification methods: Naïve Bayes, Support Vector Machine Multi-Layer Perceptron and KStar. The Abr dataset comprised both high level and near threshold recordings, labeled as 'response' or 'no response' by the human expert. Features were extracted from single averaged recordings to make the classification process straightforward. A best classification accuracy of 83.4% was obtained using Naïve Bayes and five relevant features extracted from time and wavelet domains. Naïve Bayes also achieved the highest specificity (86.3%). The highest sensitivity (93.1%) was obtained with Support Vector Machine-based classification models. In terms of the overall classification accuracy, four classifiers have shown the consistent, relatively high performance, indicating the relevance of selected features and the feasibility of using machine learning and statistical classification models in the analysis of ABR.

Sirolimus-Induced Remission of Posttransplantation Lymphoproliferative Disorder

Posttransplantation lymphoproliferative disorder (PTLD) is one of the most serious complications of solid-organ transplantation. It potentially is treatable in most cases, but current methods involve withdrawal or reduction of immunosuppression and the consequent risk for graft rejection. Sirolimus was shown in vivo and in vitro to limit proliferation of a number of malignant cell lines, including those of PTLD-derived cells. We present a case of disseminated PTLD in a patient with a renal transplant that resolved completely with conversion of immunosuppression to sirolimus. Graft function was maintained and improved with treatment. This offers a novel means of treating these patients and minimizing transplant loss.

Differentiation of the follicle-associated epithelium in ileal Peyer?s patch and production of 50-nm particles are maintained in B-cell-depleted fetal sheep

To evaluate the dependence of the differentiation of the follicle-associated epithelium (FAE) on the presence of follicular B-cells, the FAE of ileal Peyer's patch follicles was examined in B-cell-depleted fetal lambs. The FAE of these rudimentary follicles, which are devoid of lymphocytes, showed normal differentiation, including carbonic anhydrase reactivity and ultrastructural characteristics of transcytosis, extensive interdigitation of the lateral plasma membrane and the shedding of membrane-bounded particles, approximately 50 nm in size, resembling exosomes. These 50-nm membrane-bounded particles were abundant in the extracellular space of the epithelium and the dome but no particles were found in the rudimentary follicles. This study confirms that the rudimentary follicles consist of clusters of follicular dendritic cells. Our findings suggest that the differentiation of FAE of ileal Peyer's patch and the production of the 50-nm particles constitute features that appear to be independent of B-cells.

Self directed and lifelong learning

Given the many changes that occur in medicine, health care and information technologies we need to prepare all our students to engage in self directed and life long learning. There is considerable opportunity for self-directed and lifelong learning in health informatics bringing together students in exciting global learning environments, where they have much greater freedom and flexibility in their studies and potentially a wider variety of resources available to them. Self-directed learning focuses on the process by which adults take control of their own learning, in particular how they set their own learning goals, locate appropriate resources, decide on which learning methods to use and evaluate their progress. Lifelong learning happens in a variety of formal and informal settings building on both intentional and incidental learning experiences. In a lifelong learning situation the tutor must relinquish the role of expert and assume the role of facilitator, guiding learners to uncover their own knowledge. Against a back drop of rapid advances in technology which can be used to both deliver course materials and provide enhanced learning opportunities, this chapter outlines the pedagogic principles and practices which underpin self-directed and lifelong learning.

The effect of dosage, gestational age and splenectomy on anti-IgM interception of prenatal B-cell development in sheep

The administration of a single bolus of anti-IgM antibody to foetal lambs early in pregnancy produces prolonged B-cell depletion. The present study investigated this depletion by examining the effect, on B-cell development in the ileal Peyer's patches, of varying the timing and dosage of antibody administration and by supplementing anti-IgM with surgical splenectomy. The capacity of a 1 mg bolus of anti-IgM to deplete Peyer's patches of B cells was lost if its administration was deferred until two thirds of the way through pregnancy, but persisted beyond this time if weight-adjusted doses were used. Splenectomy of the foetus performed at an earlier age failed to extend the age at which a 1 mg dose of antibody remained effective. As the concentration of murine immunoglobulin in foetal serum was greatly reduced after 21 days, it is inferred that ongoing suppression of B-cell development is not dependent on the continued presence of murine immunoglobulin. The enduring nature of suppression could be attributable to a limited period during which differentiation of B cells from stem cells normally occurs, although further studies will be needed to investigate this and other possible explanations for the effect of anti-IgM treatment on prenatal B-cell development in sheep.

An application of linear output error modelling for studying lymphocyte migration in peripheral lymphoid tissues

Lymphocyte recirculation between lymphatic and blood vessels and migration through tissues are essential mechanisms underlying immunological surveillance. However, the kinetics of lymphocyte migration through lymphoid tissues remains poorly understood. The present study of lymphocyte migration, based on a sheep model and entailing the long term cannulation of blood vessels and lymphatic vessels efferent from lymph nodes, represents the first attempt to apply control engineering based models to overcome some of the experimental impediments to understanding the complex phenomena involved in lymphocyte migration. An output error model order (1,2,nk) was systematically selected under given criteria from four classes of Linear Time-Invariant Single-Input Single-Output, (LTI-SISO) systems to represent the peripheral lymph node system. The unit impulse responses were simulated under noise free conditions and their features were extracted to describe the dynamics of the system. The findings from this study revealed novel information about several aspects of the dynamics of lymphocyte migration.