Are GPs under-investigating older patients presenting with symptoms of ovarian cancer? Observational study using General Practice Research Database.

BACKGROUND: Recent studies suggest that older patients in the United Kingdom are not benefiting as much from improvements in cancer treatments as their younger counterparts. We investigate whether this might be partly due to differential referral rates using ovarian cancer as an example. METHODS: From the General Practice Research Database (GPRD), we identified all women aged 40-80 years on 1 June 2002 with a Read code for ovarian cancer between 1 June 2002 and 31 May 2007. Using these records, we compared the GPRD incidence of ovarian cancer with rates compiled from the UK cancer registries and investigated the relationship between age and coded investigations for suspected ovarian cancer. RESULTS: The GPRD rates peaked earlier, at 70-74, and were lower than registry rates for nearly all ages particularly for patients over 59. The proportion investigated or referred by the GP decreased significantly with age and delays between first coded symptom and investigation showed a U-shaped distribution by age. CONCLUSIONS: GPs appear to be less likely to recognise and to refer patients presenting with ovarian cancer as they get older. If our findings extend to other cancers, lack of or delays in referral to secondary care may partly explain poor UK cancer mortality rates of older people.

Is infant growth changing?

Weight gain between birth and 9 months of 12 903 term Millennium Cohort Study infants was investigated in order to determine differences according to sex, ethnicity and country of birth. The standardised weights and weight gains were also compared with a cohort of mainly white infants born 10 years earlier to determine whether weight gain has changed over the last decade. There were significant differences between ethnic groups, with black infants showing the largest weight gain and Asians the smallest. White boys born in England and Scotland grew relatively faster than girls, but there were no significant gender differences among the other ethnic groups or among infants born in Ireland and Wales. There was very little difference in weight gain between white English Millennium cohort infants and the earlier cohort, suggesting that the current epidemic of childhood obesity starts after 9 months of age.

Classification of brain tumours using short echo time 1H MR spectra.

The purpose was to objectively compare the application of several techniques and the use of several input features for brain tumour classification using Magnetic Resonance Spectroscopy (MRS). Short echo time 1H MRS signals from patients with glioblastomas (n = 87), meningiomas (n = 57), metastases (n = 39), and astrocytomas grade II (n = 22) were provided by six centres in the European Union funded INTERPRET project. Linear discriminant analysis, least squares support vector machines (LS-SVM) with a linear kernel and LS-SVM with radial basis function kernel were applied and evaluated over 100 stratified random splittings of the dataset into training and test sets. The area under the receiver operating characteristic curve (AUC) was used to measure the performance of binary classifiers, while the percentage of correct classifications was used to evaluate the multiclass classifiers. The influence of several factors on the classification performance has been tested: L2- vs. water normalization, magnitude vs. real spectra and baseline correction. The effect of input feature reduction was also investigated by using only the selected frequency regions containing the most discriminatory information, and peak integrated values. Using L2-normalized complete spectra the automated binary classifiers reached a mean test AUC of more than 0.95, except for glioblastomas vs. metastases. Similar results were obtained for all classification techniques and input features except for water normalized spectra, where classification performance was lower. This indicates that data acquisition and processing can be simplified for classification purposes, excluding the need for separate water signal acquisition, baseline correction or phasing.

Brain tumor classification based on long echo proton MRS signals.

There has been a growing research interest in brain tumor classification based on proton magnetic resonance spectroscopy (1H MRS) signals. Four research centers within the EU funded INTERPRET project have acquired a significant number of long echo 1H MRS signals for brain tumor classification. In this paper, we present an objective comparison of several classification techniques applied to the discrimination of four types of brain tumors: meningiomas, glioblastomas, astrocytomas grade II and metastases. Linear and non-linear classifiers are compared: linear discriminant analysis (LDA), support vector machines (SVM) and least squares SVM (LS-SVM) with a linear kernel as linear techniques and LS-SVM with a radial basis function (RBF) kernel as a non-linear technique. Kernel-based methods can perform well in processing high dimensional data. This motivates the inclusion of SVM and LS-SVM in this study. The analysis includes optimal input variable selection, (hyper-) parameter estimation, followed by performance evaluation. The classification performance is evaluated over 200 stratified random samplings of the dataset into training and test sets. Receiver operating characteristic (ROC) curve analysis measures the performance of binary classification, while for multiclass classification, we consider the accuracy as performance measure. Based on the complete magnitude spectra, automated binary classifiers are able to reach an area under the ROC curve (AUC) of more than 0.9 except for the hard case glioblastomas versus metastases. Although, based on the available long echo 1H MRS data, we did not find any statistically significant difference between the performances of LDA and the kernel-based methods, the latter have the strength that no dimensionality reduction is required to obtain such a high performance.

The contribution of proton magnetic resonance spectroscopy (1HMRS) to clinical brain tumour diagnosis.

Proton magnetic resonance spectroscopy (1HMRS) provides biochemical information from tissue non-invasively, and has an evolving role in brain tumour diagnosis and management. We present 100 consecutive patients with brain tumours who had single voxel 1HMRS as part of their preoperative investigations. We report the histopathological findings and the diagnostic contribution of spectroscopy in an adjunctive role. On the basis of clinical and radiological information the preoperative diagnosis was unclear or inaccurate in 26 out of 100 cases. The discrepancy was of lesion grade in 17 cases and lesion type in 9 cases. In 6 of 100 patients with brain tumours 1HMRS could have made a significant contribution to the preoperative diagnosis if used as part of the routine assessment. There is therefore a useful role for 1HMRS in the evaluation of intracranial mass lesions.

Magnetic Resonance Spectroscopy of cancer-practicalities of multi-centre trials and early results in non-Hodgkin's lymphoma.

This review describes problems and solutions encountered in large scale multicentre trials of Magnetic Resonance Methods for monitoring cancer. It is illustrated with reference to the Multi-Institutional Group on Magnetic Resonance Spectroscopy (MRS) Applications to Cancer which was set up to perform a trial of 31P MRS for monitoring non-invasively chemotherapy of solid tumours. 31P MR spectra of non-Hodgkin's lymphoma (NHL) pre- and posttreatment, across nine Institutions, were acquired on either General Electric (GE) or Siemens 1.5T Clinical MR instruments. Development of the trial protocol, design of the Radio Frequency (RF) coils and Quality Control procedures necessary to ensure that the datasets acquired at each centre were comparable, are described. The data revealed that phosphomonoesters (PME)/nucleotide triphosphates (NTP) ratio decreased significantly after treatment in the Complete (P<0.001) and Partial (P<0.05) Responders but not in the Non-Responders (P>0.1). In addition, the PME/NTP ratio in the pre-treatment spectra correlated with the subsequent outcome of treatment indicating that PME/NTP levels are significant predictors of long-term clinical response and time-to-treatment failure in NHL.

A prototype decision support system for MR spectroscopy-assisted diagnosis of brain tumours.

Our objective is to develop a decision support system that improves the accuracy of non-invasive brain tumour diagnosis and grading by enabling radiologists to use data from Magnetic Resonance Spectroscopy (MRS). The system, which uses pattern recognition techniques, is trained on a validated database of spectra and associated clinical information to provide automated classification of spectra from brain tumours. An innovative user-interface presents classification results as a two-dimensional overview plot in which points representing cases of different diseases form distinct clusters. Users can inspect any cases in these plots and compare them with the new, unknown spectrum. Hence, the overview plot can both communicate the classification of a case and help provide explanation for that classification in part by supporting human case-based reasoning. This paper describes the development of a prototype system implemented in JAVA.

Investigation of the metabolite variation in control rat urine using (1)H NMR spectroscopy.

An exploratory statistical analysis has been undertaken of 640 (1)H NMR spectra of rat urine, obtained from predose and control animals during the course of eight separate toxicology studies. The aim was to determine the degree and type of variation between (1)H NMR spectra from such control animals and to investigate the variations in the spectral descriptors based on averaged peak intensities. The results showed that many of the spectral descriptors had skew and/or multimodal distributions, and that it was possible to distinguish between samples of urine collected at different times of day with a success rate of (89%) and to classify 90% of the predose spectra into their correct study group using principal component and linear discriminant analyses. The results show that successful classification can be achieved of NMR spectra of control rat urine, which exhibit more subtle changes than those previously reported when treated and control animals were compared. The results presented here suggest that it will be possible to identify very subtle toxicological changes if care is taken to standardize the experimental conditions used during toxicity screens.

Magnetic resonance spectroscopy of brain hemangiopericytomas: high myoinositol concentrations and discrimination from meningiomas.

OBJECT: Hemangiopericytomas are a rare type of brain tumor that are very similar to meningiomas in appearance and symptoms but require different treatment. It is not normally possible to distinguish between them by using magnetic resonance (MR) imaging and computerized tomography studies. However, discrimination may be possible by using in vivo MR spectroscopy (MRS) because the biochemical composition of these two lesions is different. The goal of this study was to describe the use of MRS in discriminating between these similar tumor types. METHODS: In vivo MRS spectra were acquired in 27 patients (three with hemangiopericytomas and 24 with meningiomas) by using a single-voxel proton brain examination system at 1.5 teslas with short- (20-msec) and long- (135-msec) echo times. In addition, brain biopsy specimens obtained by open craniotomy were frozen within 5 minutes of resection and stored in liquid nitrogen until they were used. The specimens were powdered, extracted with perchloric acid, redissolved in 2H2O2 and high-resolution in vitro MRS was used at 9.4 teslas to record their spectra. CONCLUSIONS: In this study the authors show that hemangiopericytomas could be clearly distinguished from meningiomas because they have a larger peak at 3.56 ppm. Measurements of extracts of the tumors and comparison of spectra acquired with MRS at long- (135-msec) and short- (20-msec) echo times established that this was due to the much higher levels of myoinositol in the hemangiopericytomas.

The influence of medical history on restorative procedure failure rates following dental rehabilitation.

PURPOSE: This study evaluated the association between patient medical history and the outcomes of restorative procedures performed under general anesthesia. METHODS: The dental records of patients who had dental rehabilitation under general anesthesia at Children's Hospital in Boston (1990-1992) and Children's National Medical Center in Washington, DC (1994-1998) were examined. Data regarding restorative outcomes and the association between patient medical history and restorative failures were assessed using chi-square tests with correction for continuity. T-tests were performed on parametric data. RESULTS: Significantly higher stainless steel crown failure rates were found in young patients diagnosed with developmental disabilities when compared to patients without such disabilities (p<0.025, x2 = 5.50). However, there was no difference in the failure rates of SSCs in young patients with significant medical histories compared to patients without significant medical histories. Regarding amalgam and composite restorations, there were no differences in failure rates among patients with and without significant medical histories, including developmental disabilities. CONCLUSIONS: SSC failures were higher in young children with developmental disabilities compared to children without these disabilities.

Distinction between normal and renal cell carcinoma kidney cortical biopsy samples using pattern recognition of (1)H magic angle spinning (MAS) NMR spectra.

The technique of magic angle spinning (MAS) high resolution (1)H NMR spectroscopy applied to intact tissues provides excellent peak resolution and thus much biochemical information. The use of computer-based pattern recognition techniques to classify human renal cortex tissue samples as normal or tumour based on their (1)H MAS NMR spectra has been investigated. In this preliminary study of 22 paired control and tumour samples, exploratory data analysis using principal components based on NMR spectral intensities showed clear separation of the two classes. Furthermore, using the supervised method of linear discriminant analysis, based on individual data point intensities or on integrated spectral regions, it was possible to distinguish between the normal and tumour kidney cortex tissue with 100% accuracy, including a single example of a metastatic tumour from a primary lung carcinoma. A tumour sample from the collecting duct of the kidney showed a different NMR spectral profile, and pattern recognition indicated that this sample did not classify with the cortical tumours.

Towards a method for automated classification of 1H MRS spectra from brain tumours.

Recent studies have shown that MRS can substantially improve the non-invasive categorization of human brain tumours. However, in order for MRS to be used routinely by clinicians, it will be necessary to develop reliable automated classification methods that can be fully validated. This paper is in two parts: the first part reviews the progress that has been made towards this goal, together with the problems that are involved in the design of automated methods to process and classify the spectra. The second part describes the development of a simple prototype system for classifying 1H single voxel spectra, obtained at an echo time (TE) of 135 ms, of the four most common types of brain tumour (meningioma (MM), astrocytic (AST), oligodendroglioma (OD) and metastasis (ME)) and cysts. This system was developed in two stages: firstly, an initial database of spectra was used to develop a prototype classifier, based on a linear discriminant analysis (LDA) of selected data points. Secondly, this classifier was tested on an independent test set of 15 newly acquired spectra, and the system was refined on the basis of these results. The system correctly classified all the non-astrocytic tumours. However, the results for the the astrocytic group were poorer (between 55 and 100%, depending on the binary comparison). Approximately 50% of high grade astrocytoma (glioblastoma) spectra in our data base showed very little lipid signal, which may account for the poorer results for this class. Consequently, for the refined system, the astrocytomas were subdivided into two subgroups for comparison against other tumour classes: those with high lipid content and those without.

Automated feature extraction for the classification of human in vivo 13C NMR spectra using statistical pattern recognition and wavelets.

If magnetic resonance spectroscopy (MRS) is to become a useful tool in clinical medicine, it will be necessary to find reliable methods for analyzing and classifying MRS data. Automated methods are desirable because they can remove user bias and can deal with large amounts of data, allowing the use of all the available information. In this study, techniques for automatically extracting features for the classification of MRS in vivo data are investigated. Among the techniques used were wavelets, principal component analysis, and linear discriminant function analysis. These techniques were tested on a set of 75 in vivo 13C spectra of human adipose tissue from subjects from three different dietary groups (vegan, vegetarian, and omnivore). It was found that it was possible to assign automatically 94% of the vegans and omnivores to their correct dietary groups, without the need for explicit identification or measurement of peaks.

Lipid metabolite peaks in pattern recognition analysis of tumour in vivo MR spectra.

The ability to classify spectra of tumours according to their stage and type will be essential if magnetic resonance spectroscopy (MRS) is to be used as an aid in the diagnosis of cancer. MRS data are normally classified on the basis of selected peak measurements but these may be difficult to extract automatically. We present two alternative methods of feature extraction which we used to discriminate between spectra from tumours and normal tissues. Discrimination could be achieved either using features from the whole spectrum, or from a selected region containing the peaks from the phospholipid precursors in the phosphomonoester region.