Classification of brain tumours from MR spectra: the INTERPRET collaboration and its outcomes.

The INTERPRET project was a multicentre European collaboration, carried out from 2000 to 2002, which developed a decision-support system (DSS) for helping neuroradiologists with no experience of MRS to utilize spectroscopic data for the diagnosis and grading of human brain tumours. INTERPRET gathered a large collection of MR spectra of brain tumours and pseudo-tumoural lesions from seven centres. Consensus acquisition protocols, a standard processing pipeline and strict methods for quality control of the aquired data were put in place. Particular emphasis was placed on ensuring the diagnostic certainty of each case, for which all cases were evaluated by a clinical data validation committee. One outcome of the project is a database of 304 fully validated spectra from brain tumours, pseudotumoural lesions and normal brains, along with their associated images and clinical data, which remains available to the scientific and medical community. The second is the INTERPRET DSS, which has continued to be developed and clinically evaluated since the project ended. We also review here the results of the post-INTERPRET period. We evaluate the results of the studies with the INTERPRET database by other consortia or research groups. A summary of the clinical evaluations that have been performed on the post-INTERPRET DSS versions is also presented. Several have shown that diagnostic certainty can be improved for certain tumour types when the INTERPRET DSS is used in conjunction with conventional radiological image interpretation. About 30 papers concerned with the INTERPRET single-voxel dataset have so far been published. We discuss stengths and weaknesses of the DSS and the lessons learned. Finally we speculate on how the INTERPRET concept might be carried into the future.

Determining the date of diagnosis--is it a simple matter? The impact of different approaches to dating diagnosis on estimates of delayed care for ovarian cancer in UK primary care.

BACKGROUND: Studies of cancer incidence and early management will increasingly draw on routine electronic patient records. However, data may be incomplete or inaccurate. We developed a generalizable strategy for investigating presenting symptoms and delays in diagnosis using ovarian cancer as an example. METHODS: The General Practice Research Database was used to investigate the time between first report of symptom and diagnosis of 344 women diagnosed with ovarian cancer between 01/06/2002 and 31/05/2008. Effects of possible inaccuracies in dating of diagnosis on the frequencies and timing of the most commonly reported symptoms were investigated using four increasingly inclusive definitions of first diagnosis/suspicion: 1. "Definite diagnosis" 2. "Ambiguous diagnosis" 3. "First treatment or complication suggesting pre-existing diagnosis", 4 "First relevant test or referral". RESULTS: The most commonly coded symptoms before a definite diagnosis of ovarian cancer, were abdominal pain (41%), urogenital problems(25%), abdominal distension (24%), constipation/change in bowel habits (23%) with 70% of cases reporting at least one of these. The median time between first reporting each of these symptoms and diagnosis was 13, 21, 9.5 and 8.5 weeks respectively. 19% had a code for definitions 2 or 3 prior to definite diagnosis and 73% a code for 4. However, the proportion with symptoms and the delays were similar for all four definitions except 4, where the median delay was 8, 8, 3, 10 and 0 weeks respectively. CONCLUSION: Symptoms recorded in the General Practice Research Database are similar to those reported in the literature, although their frequency is lower than in studies based on self-report. Generalizable strategies for exploring the impact of recording practice on date of diagnosis in electronic patient records are recommended, and studies which date diagnoses in GP records need to present sensitivity analyses based on investigation, referral and diagnosis data. Free text information may be essential in obtaining accurate estimates of incidence, and for accurate dating of diagnoses.

The health of UK military personnel who deployed to the 2003 Iraq war: a cohort study.

BACKGROUND: Concerns have been raised about the mental and physical health of UK military personnel who deployed to the 2003 war in Iraq and subsequent tours of duty in the country. METHODS: We compared health outcomes in a random sample of UK armed forces personnel who were deployed to the 2003 Iraq war with those in personnel who were not deployed. Participants completed a questionnaire covering the nature of the deployment and health outcomes, which included symptoms of post-traumatic stress disorder, common mental disorders, general wellbeing, alcohol consumption, physical symptoms, and fatigue. FINDINGS: The participation rate was 62.3% (n=4722) in the deployed sample, and 56.3% (n=5550) in the non-deployed sample. Differences in health outcomes between groups were slight. There was a modest increase in the number of individuals with multiple physical symptoms (odds ratio 1.33; 95% CI 1.15-1.54). No other differences between groups were noted. The effect of deployment was different for reservists compared with regulars. In regulars, only presence of multiple physical symptoms was weakly associated with deployment (1.32; 1.14-1.53), whereas for reservists deployment was associated with common mental disorders (2.47, 1.35-4.52) and fatigue (1.78; 1.09-2.91). There was no evidence that later deployments, which were associated with escalating insurgency and UK casualties, were associated with poorer mental health outcomes. INTERPRETATION: For regular personnel in the UK armed forces, deployment to the Iraq war has not, so far, been associated with significantly worse health outcomes, apart from a modest effect on multiple physical symptoms. There is evidence of a clinically and statistically significant effect on health in reservists.

How many mailouts? Could attempts to increase the response rate in the Iraq war cohort study be counterproductive?

BACKGROUND: Low response and reporting errors are major concerns for survey epidemiologists. However, while nonresponse is commonly investigated, the effects of misclassification are often ignored, possibly because they are hard to quantify. We investigate both sources of bias in a recent study of the effects of deployment to the 2003 Iraq war on the health of UK military personnel, and attempt to determine whether improving response rates by multiple mailouts was associated with increased misclassification error and hence increased bias in the results. METHODS: Data for 17,162 UK military personnel were used to determine factors related to response and inverse probability weights were used to assess nonresponse bias. The percentages of inconsistent and missing answers to health questions from the 10,234 responders were used as measures of misclassification in a simulation of the 'true' relative risks that would have been observed if misclassification had not been present. Simulated and observed relative risks of multiple physical symptoms and post-traumatic stress disorder (PTSD) were compared across response waves (number of contact attempts). RESULTS: Age, rank, gender, ethnic group, enlistment type (regular/reservist) and contact address (military or civilian), but not fitness, were significantly related to response. Weighting for nonresponse had little effect on the relative risks. Of the respondents, 88% had responded by wave 2. Missing answers (total 3%) increased significantly (p < 0.001) between waves 1 and 4 from 2.4% to 7.3%, and the percentage with discrepant answers (total 14%) increased from 12.8% to 16.3% (p = 0.007). However, the adjusted relative risks decreased only slightly from 1.24 to 1.22 for multiple physical symptoms and from 1.12 to 1.09 for PTSD, and showed a similar pattern to those simulated. CONCLUSION: Bias due to nonresponse appears to be small in this study, and increasing the response rates had little effect on the results. Although misclassification is difficult to assess, the results suggest that bias due to reporting errors could be greater than bias caused by nonresponse. Resources might be better spent on improving and validating the data, rather than on increasing the response rate.

Quality of recording of diabetes in the UK: how does the GP's method of coding clinical data affect incidence estimates? Cross-sectional study using the CPRD database.

OBJECTIVE: To assess the effect of coding quality on estimates of the incidence of diabetes in the UK between 1995 and 2014. DESIGN: A cross-sectional analysis examining diabetes coding from 1995 to 2014 and how the choice of codes (diagnosis codes vs codes which suggest diagnosis) and quality of coding affect estimated incidence. SETTING: Routine primary care data from 684 practices contributing to the UK Clinical Practice Research Datalink (data contributed from Vision (INPS) practices). MAIN OUTCOME MEASURE: Incidence rates of diabetes and how they are affected by (1) GP coding and (2) excluding 'poor' quality practices with at least 10% incident patients inaccurately coded between 2004 and 2014. RESULTS: Incidence rates and accuracy of coding varied widely between practices and the trends differed according to selected category of code. If diagnosis codes were used, the incidence of type 2 increased sharply until 2004 (when the UK Quality Outcomes Framework was introduced), and then flattened off, until 2009, after which they decreased. If non-diagnosis codes were included, the numbers continued to increase until 2012. Although coding quality improved over time, 15% of the 666 practices that contributed data between 2004 and 2014 were labelled 'poor' quality. When these practices were dropped from the analyses, the downward trend in the incidence of type 2 after 2009 became less marked and incidence rates were higher. CONCLUSIONS: In contrast to some previous reports, diabetes incidence (based on diagnostic codes) appears not to have increased since 2004 in the UK. Choice of codes can make a significant difference to incidence estimates, as can quality of recording. Codes and data quality should be checked when assessing incidence rates using GP data.

Mother's consent to linkage of survey data with her child's birth records in a multi-ethnic national cohort study.

BACKGROUND: The increased use of computer-based records has facilitated linkage of routine data with that obtained for research. When children are involved, parental consent for linkage is usually required. The Millennium Cohort Study, of 18,819 UK babies born in 2000-02, over-sampled families from disadvantaged and ethnic wards, providing the opportunity to investigate factors associated with mother's consent to access her child's birth records. METHODS: Factors considered included ward type and mother's socioeconomic status, ethnicity, education, age, and language. Logistic regression was used to investigate the relationship of these factors with consent. RESULTS: Consent for linkage to birth register and/or hospital maternity data was obtained from 92% of the cohort mothers. The proportions consenting differed according to the mother's country of residence, age, and education, with consent being less likely among minority ethnic group mothers, lone parents, and those with higher degrees or no qualifications. Where interviews had been translated, consent was significantly less likely if the interpreter was a male. CONCLUSION: A large proportion of mothers who were interviewed gave permission for linkage. However, there were some groups who were less likely to do so, particularly those from minority ethnic groups. These sources of non-consent bias should be taken into account when analysing linked data from socially and ethnically mixed populations. Efforts should be made to understand the reasons for non-consent, which in turn will help determine the best ways to encourage more mothers to consent in future.

Are breastfeeding rates higher among mothers delivering in Baby Friendly accredited maternity units in the UK?

BACKGROUND: The promotion and support of breastfeeding is a global priority with benefits for maternal and infant health, particularly in low-income and middle-income countries where its relevance for child survival is undisputed. However, breastfeeding rates are strikingly low in some higher-income countries, including the UK. Evidence to support the effectiveness of public health interventions to increase rates of breastfeeding initiation and duration in this setting is limited. We examined whether mothers were more likely to start and continue to breastfeed if they delivered in a UNICEF UK Baby Friendly accredited maternity unit, in a cohort with a high representation of disadvantaged and lower socioeconomic groups with traditionally low rates of breastfeeding. METHODS: We analysed maternally reported breastfeeding initiation and prevalence of any breastfeeding at 1 month for 17 359 singleton infants according to maternity unit Baby Friendly Initiative participation status at birth (accredited, certificated, or neither award). RESULTS: Mothers delivering in accredited maternity units were more likely to start breastfeeding than those delivering in units with neither award [adjusted rate ratio: 1.10, 95% confidence interval (CI) 1.05-1.15], but were not more likely to breastfeed at 1 month (0.96, 95% CI 0.84-1.09), after adjustment for social, demographic, and obstetric factors. Antenatal class attendance (1.14, 95% CI 1.11-1.17), vaginal delivery (1.05, 1.03-1.08), a companion at delivery (1.09, 1.04-1.16), and maternal post-partum hospital stay >24 h (1.06, 1.04-1.09) were also independently associated with breastfeeding initiation. CONCLUSIONS: Policies to increase the proportion of maternity units participating in the UNICEF UK Baby Friendly Initiative are likely to increase breastfeeding initiation but not duration. Other strategies are required in order to support UK mothers to breastfeed for the recommended duration.

Exploring practical approaches to maximising data quality in electronic healthcare records in the primary care setting and associated benefits. Report of panel-led discussion held at SAPC in July 2014.

BACKGROUND: Electronic healthcare records provide information about patient care over time which not only affords the opportunity to improve patient care directly through effective monitoring and identification of care requirements but also offers a unique platform for both clinical and service-model research essential to the longer-term development of the health service. The quality of the recorded data can, however, be variable and can compromise the validity of data use both for primary and secondary purposes. OBJECTIVES: In order to explore the challenges and benefits of and approaches to recording high quality primary care electronic records, a Clinical Practice Research Datalink (CPRD) sponsored workshop was held at the Society of Academic Primary Care (SAPC) conference in 2014 with the aim of engaging GPs and other data users. METHODS: The workshop was held as a structured discussion, led by an expert panel and focused around three questions: (1) What are the data quality priorities for clinicians and researchers? How do these priorities differ or overlap? (2) What challenges might GPs face in provision of good data quality both for treating their patients and for research? Do these aims conflict? (3) What tools (such as data metrics and visualisations or software components) could assist the GP in improving data quality and patient management and could this tie in with analytical processes occurring at the research stage? RESULTS: The discussion highlighted both overlap and differences in the perceived data quality priorities and challenges for different user groups. Five key areas of focus were agreed upon and recommendations determined for moving forward in improving quality. CONCLUSIONS: The importance of good high quality electronic healthcare records has been set forth along with the need for a practical user-considered and collaborative approach to its improvement.

The contribution of parental and community ethnicity to breastfeeding practices: evidence from the Millennium Cohort Study.

BACKGROUND: The association of ethnic and social factors at the individual and community level with inequalities in starting, and continuing, to breastfeed remains unclear. We explored these factors using data from the Millennium Cohort Study. METHODS: We obtained data for 11 286 natural mothers [8207 (85%) white] of singleton infants, living in England at age 9 months. Breastfeeding mothers were defined as the proportion of all mothers who reported: putting their baby to the breast at least once and giving any breastmilk (initiation); and having started, continuing for at least 1 month (continuation). RESULTS: In England, 72% of all mothers started breastfeeding, and of these 70% continued for at least 1 month. White mothers were the least likely to start breastfeeding (70%), as were multiparous mothers (69%), younger mothers, those with no academic qualifications (51%), in routine occupations (59%), or living in disadvantaged communities (60%). For white mothers, having a partner of a different ethnic group was independently and positively associated with breastfeeding initiation and continuation to 1 month [adjusted rate ratios (95% CI): 1.14 (1.07-1.21) and 1.09 (1.03-1.16), respectively]. White lone mothers were more likely to initiate breastfeeding if they lived in high ethnic minority communities [adjusted rate ratio (95% CI): 1.42 (1.15-1.76)] rather than disadvantaged areas. For all mothers, maternal age at first motherhood was positively associated with breastfeeding [adjusted rate ratio (95% CI): 1.06 (1.04-1.08) per 5 year increase]. CONCLUSIONS: Significant inequalities in breastfeeding practices remain within the UK. White women are less likely to breastfeed and, for these women, partner and community ethnicity have an important relation to starting and continuing to breastfeed. Our findings suggest that public health strategies to increase breastfeeding need to be focussed on mothers who are young at first motherhood and address support offered by partners and the communities in which women live. Measures to evaluate the effectiveness of these strategies over time and between places should take account of changes in ethnic composition of the child-bearing population.

Factors affecting a mother's recall of her baby's birth weight.

BACKGROUND: The Millennium Cohort Study of UK babies born this century obtained maternal report of birth weight and data on the family's characteristics, including parental ethnicity, education, and social circumstances. Parental permission to link babies to their birth registration data provided the opportunity to investigate factors affecting accuracy of maternal recall of birth weight and to determine possible causes of error. METHODS: Logistic regression was used to investigate the relationship between maternal factors and recall of birth weight. Numerical and graphical methods were used to identify potential causes for birth weight discrepancies. RESULTS: Data were obtained from the birth registry and Millennium Cohort Study for 11 890 of the 14 294 cohort children born in England and Wales. Weight was reported in imperial units by 84% of mothers and this was more common in younger mothers. Accuracy within 100 g was 92% overall, varying from 94% among British/Irish white mothers to 69-89% for other ethnic groups and was lower among the long-term unemployed and those living in disadvantaged or ethnic wards. Explanations (mostly rounding and transcription errors) were identified for 27% of the discrepancies of 100 g or more. Conclusion Mothers' reports of their infants' birth weight showed high level of agreement with registration data, the mean discrepancy being consistently close to zero. However, the variance of the discrepancy differed according to ethnic group, ward type, and socioeconomic status. These sources of differential variability should be taken into account in analyses using birth weight, and possibly other reported data, from socially mixed populations.

Exploiting the potential of large databases of electronic health records for research using rapid search algorithms and an intuitive query interface.

OBJECTIVE: UK primary care databases, which contain diagnostic, demographic and prescribing information for millions of patients geographically representative of the UK, represent a significant resource for health services and clinical research. They can be used to identify patients with a specified disease or condition (phenotyping) and to investigate patterns of diagnosis and symptoms. Currently, extracting such information manually is time-consuming and requires considerable expertise. In order to exploit more fully the potential of these large and complex databases, our interdisciplinary team developed generic methods allowing access to different types of user. MATERIALS AND METHODS: Using the Clinical Practice Research Datalink database, we have developed an online user-focused system (TrialViz), which enables users interactively to select suitable medical general practices based on two criteria: suitability of the patient base for the intended study (phenotyping) and measures of data quality. RESULTS: An end-to-end system, underpinned by an innovative search algorithm, allows the user to extract information in near real-time via an intuitive query interface and to explore this information using interactive visualization tools. A usability evaluation of this system produced positive results. DISCUSSION: We present the challenges and results in the development of TrialViz and our plans for its extension for wider applications of clinical research. CONCLUSIONS: Our fast search algorithms and simple query algorithms represent a significant advance for users of clinical research databases.

Optimising use of electronic health records to describe the presentation of rheumatoid arthritis in primary care: a strategy for developing code lists.

BACKGROUND: Research using electronic health records (EHRs) relies heavily on coded clinical data. Due to variation in coding practices, it can be difficult to aggregate the codes for a condition in order to define cases. This paper describes a methodology to develop 'indicator markers' found in patients with early rheumatoid arthritis (RA); these are a broader range of codes which may allow a probabilistic case definition to use in cases where no diagnostic code is yet recorded. METHODS: We examined EHRs of 5,843 patients in the General Practice Research Database, aged ≥ 30 y, with a first coded diagnosis of RA between 2005 and 2008. Lists of indicator markers for RA were developed initially by panels of clinicians drawing up code-lists and then modified based on scrutiny of available data. The prevalence of indicator markers, and their temporal relationship to RA codes, was examined in patients from 3 y before to 14 d after recorded RA diagnosis. FINDINGS: Indicator markers were common throughout EHRs of RA patients, with 83.5% having 2 or more markers. 34% of patients received a disease-specific prescription before RA was coded; 42% had a referral to rheumatology, and 63% had a test for rheumatoid factor. 65% had at least one joint symptom or sign recorded and in 44% this was at least 6-months before recorded RA diagnosis. CONCLUSION: Indicator markers of RA may be valuable for case definition in cases which do not yet have a diagnostic code. The clinical diagnosis of RA is likely to occur some months before it is coded, shown by markers frequently occurring ≥ 6 months before recorded diagnosis. It is difficult to differentiate delay in diagnosis from delay in recording. Information concealed in free text may be required for the accurate identification of patients and to assess the quality of care in general practice.

Extracting diagnoses and investigation results from unstructured text in electronic health records by semi-supervised machine learning.

BACKGROUND: Electronic health records are invaluable for medical research, but much of the information is recorded as unstructured free text which is time-consuming to review manually. AIM: To develop an algorithm to identify relevant free texts automatically based on labelled examples. METHODS: We developed a novel machine learning algorithm, the 'Semi-supervised Set Covering Machine' (S3CM), and tested its ability to detect the presence of coronary angiogram results and ovarian cancer diagnoses in free text in the General Practice Research Database. For training the algorithm, we used texts classified as positive and negative according to their associated Read diagnostic codes, rather than by manual annotation. We evaluated the precision (positive predictive value) and recall (sensitivity) of S3CM in classifying unlabelled texts against the gold standard of manual review. We compared the performance of S3CM with the Transductive Vector Support Machine (TVSM), the original fully-supervised Set Covering Machine (SCM) and our 'Freetext Matching Algorithm' natural language processor. RESULTS: Only 60% of texts with Read codes for angiogram actually contained angiogram results. However, the S3CM algorithm achieved 87% recall with 64% precision on detecting coronary angiogram results, outperforming the fully-supervised SCM (recall 78%, precision 60%) and TSVM (recall 2%, precision 3%). For ovarian cancer diagnoses, S3CM had higher recall than the other algorithms tested (86%). The Freetext Matching Algorithm had better precision than S3CM (85% versus 74%) but lower recall (62%). CONCLUSIONS: Our novel S3CM machine learning algorithm effectively detected free texts in primary care records associated with angiogram results and ovarian cancer diagnoses, after training on pre-classified test sets. It should be easy to adapt to other disease areas as it does not rely on linguistic rules, but needs further testing in other electronic health record datasets.

Using free text information to explore how and when GPs code a diagnosis of ovarian cancer: an observational study using primary care records of patients with ovarian cancer.

BACKGROUND: Primary care databases provide a unique resource for healthcare research, but most researchers currently use only the Read codes for their studies, ignoring information in the free text, which is much harder to access. OBJECTIVES: To investigate how much information on ovarian cancer diagnosis is 'hidden' in the free text and the time lag between a diagnosis being described in the text or in a hospital letter and the patient being given a Read code for that diagnosis. DESIGN: Anonymised free text records from the General Practice Research Database of 344 women with a Read code indicating ovarian cancer between 1 June 2002 and 31 May 2007 were used to compare the date at which the diagnosis was first coded with the date at which the diagnosis was recorded in the free text. Free text relating to a diagnosis was identified (a) from the date of coded diagnosis and (b) by searching for words relating to the ovary. RESULTS: 90% of cases had information relating to their ovary in the free text. 45% had text indicating a definite diagnosis of ovarian cancer. 22% had text confirming a diagnosis before the coded date; 10% over 4 weeks previously. Four patients did not have ovarian cancer and 10% had only ambiguous or suspected diagnoses associated with the ovarian cancer code. CONCLUSIONS: There was a vast amount of extra information relating to diagnoses in the free text. Although in most cases text confirmed the coded diagnosis, it also showed that in some cases GPs do not code a definite diagnosis on the date that it is confirmed. For diseases which rely on hospital consultants for diagnosis, free text (particularly letters) is invaluable for accurate dating of diagnosis and referrals and also for identifying misclassified cases.

Do early infant feeding practices vary by maternal ethnic group?

OBJECTIVE: To examine UK country and ethnic variations in infant feeding practices. DESIGN: Cohort study. SETTING: Infants enrolled in the Millennium Cohort Study, born between September 2000 and January 2002. SUBJECTS: A total of 18 150 natural mothers (11 286 (8207 white) living in England) of singleton infants. OUTCOME MEASURES: Breast-feeding initiation, breast-feeding discontinuation and introduction of solid foods before 4 months. EXPLANATORY VARIABLES: Maternal ethnic group, education and social class. RESULTS: Seventy per cent of UK mothers started to breast-feed, of whom 62% stopped before 4 months. Median age at discontinuing breast-feeding was 14, 13, 10 and 6 weeks in Scotland, England, Wales and Northern Ireland, respectively. Thirty-six per cent of UK mothers (34% in England) introduced solids before 4 months. White mothers were more likely to discontinue breast-feeding (62%) and introduce solids early (37%) than most other ethnic minority groups; those stopping before 4 months were more likely to introduce solids early compared with those continuing to breast-feed beyond this age (adjusted rate ratio (95% confidence interval): 1.3 (1.1-1.2)). Educated mothers were less likely to stop breast-feeding before 4 months (white mothers, 0.8 (0.8-0.9); non-white mothers, 0.9 (0.8-1.0)) than those with no/minimal qualifications but, among ethnic minorities, were more likely to introduce solids early (1.3 (1.0-1.6)). Socio-economic status was positively associated with breast-feeding continuation among white women, and with age at introduction of solids among non-white women. CONCLUSIONS: We have identified important geographic, ethnic and social inequalities in breast-feeding continuation and introduction of solids within the UK, many of which have not been reported previously. The factors mediating these associations are complex and merit further study to ensure that interventions proposed to promote maternal adherence to current infant feeding recommendations are appropriate and effective.

Differences in risk factors for partial and no immunisation in the first year of life: prospective cohort study.

OBJECTIVE: To compare demographic, social, maternal, and infant related factors associated with partial immunisation and no immunisation in the first year of life in the United Kingdom. DESIGN: Prospective cohort study. SETTING: Sample of electoral wards in England, Wales, Scotland, and Northern Ireland, stratified by measures of ethnic composition and social disadvantage. PARTICIPANTS: 18,488 infants born between September 2000 and January 2002, resident in the UK and eligible to receive child benefit (a universal benefit available to all families) at age 9 months. MAIN OUTCOME MEASURE: Immunisation status at 9 months of age, defined as fully immunised, partially immunised, or not immunised. RESULTS: Overall in the UK, 3.3% of infants were partially immunised and 1.1% were unimmunised; these rates were highest in England (3.6% and 1.3%, respectively; P < 0.01). Residence in ethnic or disadvantaged wards, larger family size, lone or teenaged parenthood, maternal smoking in pregnancy, and admission to hospital by 9 months of age were independently associated with partial immunisation status. In contrast, a higher proportion of mothers of unimmunised infants were educated to degree level or above (1.9%), were older (3.1%), or were of black Caribbean ethnicity (4.7%). CONCLUSIONS: Mothers of unimmunised infants differ in terms of age and education from those of partially immunised infants. Interventions to reduce incomplete immunisation in infancy need different approaches.

Independent component analysis for automated decomposition of in vivo magnetic resonance spectra.

Fully automated methods for analyzing MR spectra would be of great benefit for clinical diagnosis, in particular for the extraction of relevant information from large databases for subsequent pattern recognition analysis. Independent component analysis (ICA) provides a means of decomposing signals into their constituent components. This work investigates the use of ICA for automatically extracting features from in vivo MR spectra. After its limits are assessed on artificial data, the method is applied to a set of brain tumor spectra. ICA automatically, and in an unsupervised fashion, decomposes the signals into interpretable components. Moreover, the spectral decomposition achieved by the ICA leads to the separation of some tissue types, which confirms the biochemical relevance of the components.

Automated classification of short echo time in in vivo 1H brain tumor spectra: a multicenter study.

Automated pattern recognition techniques are needed to help radiologists categorize MRS data of brain tumors according to histological type and grade. A major question is whether a computer program "trained" on spectra from one hospital will be able to classify those from another, particularly if the acquisition protocol is different. A subset of 144 histopathologically validated brain tumor spectra in the INTERPRET database, obtained from three of the collaborating centers, was grouped into meningiomas, low-grade astrocytomas, and "aggressive tumors" (glioblastomas and metastases). Spectra from two centers formed the training set (94 spectra) while the third acted as the test set (50 spectra). Linear discriminant analysis successfully classified 48/50 in the test set; the remaining two were atypical cases. When the training and test sets were combined, 133 of the 144 spectra were correctly classified using the leave-one-out procedure. These spectra had been obtained using different sequences (STEAM and PRESS), different echo times (20, 30, 31, and 32 ms), different repetition times (1600 and 2000 ms), and different manufacturers' instruments (GE and Philips). Pattern recognition algorithms are less sensitive to acquisition parameters than had been expected.