Publications by Mr Daniele Soria

Individuals developing stroke have varying clinical characteristics, demographic, and biochemical profiles. This heterogeneity in phenotypic characteristics can impact on cardiovascular disease (CVD) morbidity and mortality outcomes. This study uses a novel clustering approach to stratify individuals with incident stroke into phenotypic clusters and evaluates the differential burden of recurrent stroke and other cardiovascular outcomes. We used linked clinical data from primary care, hospitalisations, and death records in the UK. A data-driven clustering analysis (kamila algorithm) was used in 48,114 patients aged ≥ 18 years with incident stroke, from 1-Jan-1998 to 31-Dec-2017 and no prior history of serious vascular events. Cox proportional hazards regression was used to estimate hazard ratios (HRs) for subsequent adverse outcomes, for each of the generated clusters. Adverse outcomes included coronary heart disease (CHD), recurrent stroke, peripheral vascular disease (PVD), heart failure, CVD-related and all-cause mortality. Four distinct phenotypes with varying underlying clinical characteristics were identified in patients with incident stroke. Compared with cluster 1 (n = 5,201, 10.8%), the risk of composite recurrent stroke and CVD-related mortality was higher in the other 3 clusters (cluster 2 [n = 18,655, 38.8%]: hazard ratio [HR], 1.07; 95% CI, 1.02–1.12; cluster 3 [n = 10,244, 21.3%]: HR, 1.20; 95% CI, 1.14–1.26; and cluster 4 [n = 14,014, 29.1%]: HR, 1.44; 95% CI: 1.37–1.50). Similar trends in risk were observed for composite recurrent stroke and all-cause mortality outcome, and subsequent recurrent stroke outcome. However, results were not consistent for subsequent risk in CHD, PVD, heart failure, CVD-related mortality, and all-cause mortality. In this proof of principle study, we demonstrated how a heterogenous population of patients with incident stroke can be stratified into four relatively homogenous phenotypes with differential risk of recurrent and major cardiovascular outcomes. This offers an opportunity to revisit the stratification of care for patients with incident stroke to improve patient outcomes.

Colorectal cancer (CRC) is one of the most common cancers worldwide. This study was designed to evaluate biological patterns, explore molecular classification and correlate with survival outcome in treatment naïve CRC patients. Methods Over 11 years consecutive series of 435 CRC patients were operated on as primary surgical therapy. A total of 201 CRC patients were included, whose complete set of clinical information was available, and their good quality tumour blocks were retrieved. Immunohistochemistry was used for tumour analysis, and partitional clustering was performed using R software for cluster analysis. Results The median age was 43 (range 10-85) years; adenocarcinoma was the most commonly seen histological type. The great majority had positive CK20, CEA, E-Cadherin, Ki67, CDX2, and p53 expression. There were four distinct molecular classes found, whereas Ki67, CDX2, and p53 play 3 the main role in partitioning. Younger age negatively impacted survival; overall and diseasespecific survival was 26 months only with 50 months' longest survival. Conclusion Colorectal cancer is a biologically heterogeneous disease with at least four distinct molecular patterns, where cell proliferation and gene repair mechanisms appear to play the key role.

Purpose: Some patients with breast cancer treated by surgery and radiation therapy experience clinically significant toxicity, which may adversely affect cosmesis and quality of life. There is a paucity of validated clinical prediction models for radiation toxicity. We used machine learning (ML) algorithms to develop and optimise a clinical prediction model for acute breast desquamation after whole breast external beam radiation therapy in the prospective multicenter REQUITE cohort study. Methods and Materials: Using demographic and treatment-related features (m = 122) from patients (n = 2058) at 26 centers, we trained 8 ML algorithms with 10-fold cross-validation in a 50:50 random-split data set with class stratification to predict acute breast desquamation. Based on performance in the validation data set, the logistic model tree, random forest, and naïve Bayes models were taken forward to cost-sensitive learning optimisation. Results: One hundred and ninety-two patients experienced acute desquamation. Resampling and cost-sensitive learning optimisation facilitated an improvement in classification performance. Based on maximising sensitivity (true positives), the "hero" model was the cost-sensitive random forest algorithm with a false-negative: false-positive misclassification penalty of 90:1 containing m = 114 predictive features. Model sensitivity and specificity were 0.77 and 0.66, respectively, with an area under the curve of 0.77 in the validation cohort. Conclusions: ML algorithms with resampling and cost-sensitive learning generated clinically valid prediction models for acute desquamation using patient demographic and treatment features. Further external validation and inclusion of genomic markers in ML prediction models are worthwhile, to identify patients at increased risk of toxicity who may benefit from supportive intervention or even a change in treatment plan

Testing a hypothesis for 'factors-outcome effect' is a common quest, but standard statistical regression analysis tools are rendered ineffective by data contaminated with too many noisy variables. Expert Systems (ES) can provide an alternative methodology in analysing data to identify variables with the highest correlation to the outcome. By applying their effective machine learning (ML) abilities, significant research time and costs can be saved. The study aims to systematically review the applications of ES in urological research and their methodological models for effective multi-variate analysis. Their domains, development and validity will be identified.

The prediction by classification of side effects incidence in a given medical treatment is a common challenge in medical research. Machine Learning (ML) methods are widely used in the areas of risk prediction and classification. The primary objective of such algorithms is to use several features to predict dichotomous responses (e.g., disease positive/negative). Similar to statistical inference modelling, ML modelling is subject to the class imbalance problem and is affected by the majority class, increasing the false-negative rate. In this study, seventynine ML models were built and evaluated to classify approximately 2000 participants from 26 hospitals in eight different countries into two groups of radiotherapy (RT) side effects incidence based on recorded observations from the international study of RT related toxicity "REQUITE". We also examined the effect of sampling techniques and cost-sensitive learning methods on the models when dealing with class imbalance. The combinations of such techniques used had a significant impact on the classification. They resulted in an improvement in incidence status prediction by shifting classifiers' attention to the minority group. The best classification model for RT acute toxicity prediction was identified based on domain experts' success criteria. The Area Under Receiver Operator Characteristic curve of the models tested with an isolated dataset ranged between 0.50 and 0.77. The scale of improved results is promising and will guide further development of models to predict RT acute toxicities. One model was optimised and found to be beneficial to identify patients who are at risk of developing acute RT early-stage toxicities as a result of undergoing breast RT ensuring relevant treatment interventions can be appropriately targeted. The design of the approach presented in this paper resulted in producing a preclinical-valid prediction model. The study was developed by a multi-disciplinary collaboration of data scientists, medical physicists, oncologists and surgeons in the UK Radiotherapy Machine Learning Network.

Background: Colorectal cancer (CRC) is a heterogeneous disease, however there is limited information available regarding molecular classification and correlation with long term clinical outcome. Methods: Over the period of 11 years (2008 to 2018) totals 435 patients of colorectal cancer were reported and their tumour blocks and complete set of clinical information was available for 201 patients. Immunohistochemistry was done for ER, PR, HER 2-neu, Ki-67, Bcl-2, E-Cadherin, P53, CEA, EGFR, and VEGF. PDL1, CDX-2 and CK 20. The biological pattern characterized by partitioned clustering method as described using R software. Survival analysis was done by using Kaplan Meier method. Results: There were 201 patients including 54.7% male and females were 45.3 %. Median survival was 28 months. Cluster analysis showed four novel clusters (Table), with major difference based on Ki67, CDX2 and p53. These classes showed difference in median survival, where common class 1 showed higher survival while common cluster 4 showed poor survival. Conclusions: There are at least four distinct molecular classes of colorectal cancer which can be potentially utilized in clinical practice. Pattern of Novel molecular classification of colorectal cancer and correlation with long term survival.

Classifying subjects into risk categories is a common challenge in medical research. Machine Learning (ML) methods are widely used in the areas of risk prediction and classification. The primary objective of such algorithms is to use several features to predict dichotomous responses (e.g., healthy/at risk). Similar to statistical inference modelling, ML modelling is subject to the problem of class imbalance and is affected by the majority class, increasing the false-negative rate. In this study, we built and evaluated thirty-six ML models to classify approximately 4300 female and 4100 male participants from the UK Biobank into three categorical risk statuses based on discretised visceral adipose tissue (VAT) measurements from magnetic resonance imaging. We also examined the effect of sampling techniques on the models when dealing with class imbalance. The sampling techniques used had a significant impact on the classification and resulted in an improvement in risk status prediction by facilitating an increase in the information contained within each variable. Based on domain expert criteria the best three classification models for the female and male cohort visceral fat prediction were identified. The Area Under Receiver Operator Characteristic curve of the models tested (with external data) was 0.78 to 0.89 for females and 0.75 to 0.86 for males. These encouraging results will be used to guide further development of models to enable prediction of VAT value. This will be useful to identify individuals with excess VAT volume who are at risk of developing metabolic disease ensuring relevant lifestyle interventions can be appropriately targeted.

Hierarchical fuzzy systems (HFSs) have been shown to have the potential to improve the interpretability of fuzzy logic systems (FLSs). However, challenges remain, such as: "How can we measure their interpretability?", "How can we make an informed assessment of how HFSs should be designed to enhance interpretability?". The challenges of measuring the interpretability of HFSs include issues such as their topological structure, the number of layers, the meaning of intermediate variables, and so on. In this paper, an initial framework to measure the interpretability of HFSs is proposed, combined with a participatory user design process to create a specific instance of the framework for an application context. This approach enables the subjective views of a range of practitioners, experts in the design and creation of FLSs, to be taken into account in shaping the design of a generic framework for measuring interpretability in HFSs. This design process and framework are demonstrated through two classification application examples, showing the ability of the resulting index to appropriately capture interpretability as perceived by system design experts.

Although driving behavior has been largely studied amongst private motor vehicles drivers, the literature addressing heavy goods vehicle (HGV) drivers is scarce. Identifying the existing groups of driving stereotypes and their proportions enables researchers, companies, and policy makers to establish group-specific strategies to improve safety and economy. In addition, insight into driving styles can help predict drivers' reactions and therefore enable the modeling of interactions between vehicles and the possible obstacles encountered on a journey. Consequently, there are also contributions to the research and development of autonomous vehicles and smart roads. In this paper, our interest lies in investigating driving behavior within the HGV community in the United Kingdom (U.K.). We conduct analysis of a telematics dataset containing the incident information on 21 193 HGV drivers across the U.K. We are interested in answering two research questions: 1) What groups of behavior are we able to uncover? 2) How do these groups complement current findings in the literature? To answer these questions, we apply a two-stage data analysis methodology involving consensus clustering and ensemble classification to the dataset. Through the analysis, eight patterns of behavior are uncovered. It is also observed that although our findings have similarities to those from previous work on driving behavior, further knowledge is obtained, such as extra patterns and driving traits arising from vehicle and road characteristics.

Breast Cancer is one of the most common causes of cancer death in women, representing a very complex disease with varied molecular alterations. To assist breast cancer prognosis, the classification of patients into biological groups is of great significance for treatment strategies. Recent studies have used an ensemble of multiple clustering algorithms to elucidate the most characteristic biological groups of breast cancer. However, the combination of various clustering methods resulted in a number of patients remaining unclustered. Therefore, a framework still needs to be developed which can assign as many unclustered (i.e. biologically diverse) patients to one of the identified groups in order to improve classification. Therefore, in this paper we develop a novel classification framework which introduces a new ensemble classification stage after the ensemble clustering stage to target the unclustered patients. Thus, a step-by-step pipeline is introduced which couples ensemble clustering with ensemble classification for the identification of core groups, data distribution in them and improvement in final classification results by targeting the unclustered data. The proposed pipeline is employed on a novel real world breast cancer dataset and subsequently its robustness and stability are examined by testing it on standard datasets. The results show that by using the presented framework, an improved classification is obtained. Finally, the results have been verified using statistical tests, visualisation techniques, cluster quality assessment and interpretation from clinical experts.

One of the methods for stratifying different molecular classes of breast cancer is the Nottingham Prognostic Index Plus (NPI+) which uses breast cancer relevant biomarkers to stain tumour tissues prepared on tissue microarray (TMA). To determine the molecular class of the tumour, pathologists will have to manually mark the nuclei activity biomarkers through a microscope and use a semi-quantitative assessment method to assign a histochemical score (H-Score) to each TMA core. Manually marking positively stained nuclei is a time consuming, imprecise and subjective process which will lead to inter-observer and intra-observer discrepancies. In this paper, we present an end-to-end deep learning system which directly predicts the H-Score automatically. Our system imitates the pathologists' decision process and uses one fully convolutional network (FCN) to extract all nuclei region (tumour and non-tumour), a second FCN to extract tumour nuclei region, and a multi-column convolutional neural network which takes the outputs of the first two FCNs and the stain intensity description image as input and acts as the high-level decision making mechanism to directly output the H-Score of the input TMA image. To the best of our knowledge, this is the first end-to-end system that takes a TMA image as input and directly outputs a clinical score. We will present experimental results which demonstrate that the H-Scores predicted by our model have very high and statistically significant correlation with experienced pathologists' scores and that the H-Score discrepancy between our algorithm and the pathologists is on par with the inter-subject discrepancy between the pathologists.

Purpose: Breast cancer (BC) is a heterogeneous disease characterised by variant biology, metabolic activity, and patient outcome. Glutamine availability for growth and progression of BC is important in several BC subtypes. This study aimed to evaluate the biological and prognostic role of the combined expression of key glutamine transporters, SLC1A5, SLC7A5 and SLC3A2 in BC with emphasis on the intrinsic molecular subtypes. Methods: SLC1A5, SLC7A5 and SLC3A2 were assessed at the protein level, using immunohistochemistry on tissue microarrays constructed from a large well characterised BC cohort (n=2,248). Patients were stratified into accredited clusters based on protein expression and correlated with clinicopathological parameters, molecular subtypes, and patient outcome. Results: Clustering analysis of SLC1A5, SLC7A5 and SLC3A2 identified three clusters Low SLCs (SLC1A5-/SLC7A5-/SLC3A2-), High SLC1A5 (SLC1A5+/SLC7A5-/SLC3A2-) and High SLCs (SLC1A5+/SLC7A5+/SLC3A2+) which had distinct correlations to known prognostic factors and patient outcome (p<0.001). The key regulator of tumour cell metabolism, c-MYC, was significantly expressed in tumours in the High SLCs cluster (p<0.001). When different BC subtypes were considered, the association with the poor outcome was observed in the ER+ high proliferation/luminal B class only (p= 0.003). In multivariate analysis, SLC clusters were independent risk factor for shorter breast cancer specific survival (p= 0.001). Conclusion: The co-operative expression of SLC1A5, SLC7A5 and SLC3A2 appears to play a role in the aggressive subclass of ER+ high proliferation/ luminal BC, driven by c-MYC, and therefore have the potential to act as therapeutic targets, particularly in synergism.

Background: Altered cellular metabolism is a hallmark of cancer and some are reliant on glutamine for sustained proliferation and survival. We hypothesise that the glutamine–proline regulatory axis has a key role in breast cancer (BC) in the highly proliferative classes. Methods: Glutaminase (GLS), pyrroline-5-carboxylate synthetase (ALDH18A1), and pyrroline-5-carboxylate reductase 1 (PYCR1) were assessed at DNA/mRNA/protein levels in large, well-characterised cohorts. Results: Gain of PYCR1 copy number and high PYCR1 mRNA was associated with Luminal B tumours. High ALDH18A1 and high GLS protein expression was observed in the oestrogen receptor (ER)+/human epidermal growth factor receptor (HER2)– high proliferation class (Luminal B) compared with ER+/HER2– low proliferation class (Luminal A) (P=0.030 and P=0.022 respectively), however this was not observed with mRNA. Cluster analysis of the glutamine–proline regulatory axis genes revealed significant associations with molecular subtypes of BC and patient outcome independent of standard clinicopathological parameters (P=0.012). High protein expression of the glutamine–proline enzymes were all associated with high MYC protein in Luminal B tumours only (P<0.001). Conclusions: We provide comprehensive clinical data indicating that the glutamine–proline regulatory axis plays an important role in the aggressive subclass of luminal BC and is therefore a potential therapeutic target.

Background: There remains a need to identify and validate biomarkers for predicting prostate cancer (CaP) outcomes using robust and routinely available pathology techniques to identify men at most risk of premature death due to prostate cancer. Previous immunohistochemical studies suggest the proliferation marker Ki67 might be a predictor of survival, independently of PSA and Gleason score. We performed a validation study of Ki67 as a marker of survival and disease progression and compared its performance against another candidate biomarker, DLX2, selected using artificial neural network analysis. Methods: A tissue microarray (TMA) was constructed from transurethral resected prostatectomy histology samples (n=192). Artificial neural network analysis was used to identify candidate markers conferring increased risk of death and metastasis in a public cDNA array. Immunohistochemical analysis of the TMA was carried out and univariate and multivariate tests performed to explore the association of tumour protein levels of Ki67 and DLX2 with time to death and metastasis. Results: Univariate analysis demonstrated Ki67 as predictive of CaP-specific survival (DSS; P=0.022), and both Ki67 (P=0.025) and DLX2 (P=0.001) as predictive of future metastases. Multivariate analysis demonstrated Ki67 as independent of PSA, Gleason score and D'Amico risk category for DSS (HR=2.436, P=0.029) and both Ki67 (HR=3.296, P=0.023) and DLX2 (HR=3.051, P=0.003) as independent for future metastases. Conclusions: High Ki67 expression is only present in 6.8% of CaP patients and is predictive of reduced survival and increased risk of metastasis, independent of PSA, Gleason score and D'Amico risk category. DLX2 is a novel marker of increased metastasis risk found in 73% patients and 8.2% showed co-expression with a high Ki67 score. Two cancer cell proliferation markers, Ki67 and DLX2, may be able to inform clinical decision-making when identifying patients for active surveillance.

The Nottingham prognostic index plus (NPI+) is based on the assessment of biological class combined with established clinicopathologic prognostic variables providing improved patient outcome stratification for breast cancer superior to the traditional NPI. This study aimed to determine prognostic capability of the NPI+ in predicting risk of development of distant disease. A well-characterised series of 1073 primary early-stage BC cases treated in Nottingham and 251 cases from Budapest were immunohistochemically assessed for cytokeratin (Ck)5/6, Ck18, EGFR, oestrogen receptor (ER), progesterone receptor, HER2, HER3, HER4, Mucin 1 and p53 expression. NPI+ biological class and prognostic scores were assigned using individual algorithms for each biological class incorporating clinicopathologic parameters and investigated in terms of prediction of distant metastases-free survival (MFS). The NPI+ identified distinct prognostic groups (PG) within each molecular class which were predictive of MFS providing improved patient outcome stratification superior to the traditional NPI. NPI+ PGs, between series, were comparable in predicting patient outcome between series in luminal A, basal p53 altered and HER2+/ER+ (p > 0.01) tumours. The low-risk groups were similarly validated in luminal B, luminal N, basal p53 normal tumours (p > 0.01). Due to small patient numbers the remaining PGs could not be validated. NPI+ was additionally able to predict a higher risk of metastases at certain distant sites. This study may indicate the NPI+ as a useful tool in predicting the risk of metastases. The NPI+ provides accurate risk stratification allowing improved individualised clinical decision making for breast cancer.

Background Patients' illness beliefs have been associated with glycaemic control in diabetes and survival in other conditions. Objective We examined whether illness beliefs independently predicted survival in patients with diabetes and foot ulceration. Methods Patients (n = 169) were recruited between 2002 and 2007. Data on illness beliefs were collected at baseline. Data on survival were extracted on 1st November 2011. Number of days survived reflected the number of days from date of recruitment to 1st November 2011. Results Cox regressions examined the predictors of time to death and identified ischemia and identity beliefs (beliefs regarding symptoms associated with foot ulceration) as significant predictors of time to death. Conclusions Our data indicate that illness beliefs have a significant independent effect on survival in patients with diabetes and foot ulceration. These findings suggest that illness beliefs could improve our understanding of mortality risk in this patient group and could also be the basis for future therapeutic interventions to improve survival.

The Nottingham Prognostic Index Plus (NPI+) is a clinical decision making tool in breast cancer (BC) that aims to provide improved patient outcome stratification superior to the traditional NPI. This study aimed to validate the NPI+ in an independent series of BC. 885 primary early stage BC cases from Edinburgh were semi-quantitatively assessed for 10 biomarkers [Estrogen Receptor (ER), Progesterone Receptor (PgR), cytokeratin (CK) 5/6, CK7/8, epidermal growth factor receptor (EGFR), HER2, HER3, HER4, p53, and Mucin 1] using immunohistochemistry and classified into biological classes by fuzzy logic-derived algorithms previously developed in the Nottingham series. Subsequently, NPI+ Prognostic Groups (PGs) were assigned for each class using bespoke NPI-like formulae, previously developed in each NPI+ biological class of the Nottingham series, utilising clinicopathological parameters: number of positive nodes, pathological tumour size, stage, tubule formation, nuclear pleomorphism and mitotic counts. Biological classes and PGs were compared between the Edinburgh and Nottingham series using Cramer's V and their role in patient outcome prediction using Kaplan–Meier curves and tested using Log Rank. The NPI+ biomarker panel classified the Edinburgh series into seven biological classes similar to the Nottingham series (p>0.01). The biological classes were significantly associated with patient outcome (p<0.001). PGs were comparable in predicting patient outcome between series in Luminal A, Basal p53 altered, HER2+/ER+ tumours (p>0.01). The good PGs were similarly validated in Luminal B, Basal p53 normal, HER2+/ER- tumours and the poor PG in the Luminal N class (p>0.01). Due to small patient numbers assigned to the remaining PGs, Luminal N, Luminal B, Basal p53 normal and HER2+/ER- classes could not be validated. This study demonstrates the reproducibility of NPI+ and confirmed its prognostic value in an independent cohort of primary breast cancer. Further validation in large randomised controlled trial material is warranted.

Accurate distant metastasis (DM) prediction is critical for risk stratification and effective treatment decisions in breast cancer (BC). Many prognostic markers/models based on tissue marker studies are continually emerging using conventional statistical approaches analysing complex/dimensional data association with DM/poor prognosis. However, few of them have fulfilled satisfactory evidences for clinical application. This study aimed at building DM risk assessment algorithm for BC patients. A well-characterised series of early invasive primary operable BC (n=1902), with immunohistochemical (IHC) expression of a panel of biomarkers (n=31) formed the material of this study. Decision tree algorithm was computed using WEKA software, utilising quantitative biomarkers' expression and the absence/presence of distant metastases. Fifteen biomarkers were significantly associated with DM, with six temporal subgroups characterised based on time-to-development of DM ranging from < 1 year to > 15 years of follow-up. Of these 15 biomarkers, 10 had a significant expression pattern where Ki67LI, HER2, p53, N-cadherin, P-cadherin, PIK3CA and TOMM34 showed significantly higher expressions with earlier development of DM. In contrast, higher expressions of ER, PR, and BCL2, were associated with delayed occurrence of DM. DM prediction algorithm was built utilising cases informative for the 15 significant markers. Four risk groups of patients were characterised. Three markers; p53, HER2 and BCL2 predicted the probability of DM, based on software-generated cut-offs, with a precision rate of 81.1% for positive predictive value and 77.3%, for the negative predictive value. This algorithm reiterates the reported prognostic values of these three markers and underscores their central biologic role in BC progression. Further independent validation of this pruned panel of biomarkers is therefore warranted.

Background: Current management of breast cancer (BC) relies on risk stratification based on well-defined clinicopathologic factors. Global gene expression profiling studies have demonstrated that BC comprises distinct molecular classes with clinical relevance. In this study, we hypothesised that molecular features of BC are a key driver of tumour behaviour and when coupled with a novel and bespoke application of established clinicopathologic prognostic variables can predict both clinical outcome and relevant therapeutic options more accurately than existing methods. Methods: In the current study, a comprehensive panel of biomarkers with relevance to BC was applied to a large and well-characterised series of BC, using immunohistochemistry and different multivariate clustering techniques, to identify the key molecular classes. Subsequently, each class was further stratified using a set of well-defined prognostic clinicopathologic variables. These variables were combined in formulae to prognostically stratify different molecular classes, collectively known as the Nottingham Prognostic Index Plus (NPI+). The NPI+ was then used to predict outcome in the different molecular classes. Results: Seven core molecular classes were identified using a selective panel of 10 biomarkers. Incorporation of clinicopathologic variables in a second-stage analysis resulted in identification of distinct prognostic groups within each molecular class (NPI+). Outcome analysis showed that using the bespoke NPI formulae for each biological BC class provides improved patient outcome stratification superior to the traditional NPI. Conclusion: This study provides proof-of-principle evidence for the use of NPI+ in supporting improved individualised clinical decision making.

The electronic healthcare databases are starting to become more readily available and are thought to have excellent potential for generating adverse drug reaction signals. The Health Improvement Network (THIN) database is an electronic healthcare database containing medical information on over 11 million patients that has excellent potential for detecting ADRs. In this paper we apply four existing electronic healthcare database signal detecting algorithms (MUTARA, HUNT, Temporal Pattern Discovery and modified ROR) on the THIN database for a selection of drugs from six chosen drug families. This is the first comparison of ADR signalling algorithms that includes MUTARA and HUNT and enabled us to set a benchmark for the adverse drug reaction signalling ability of the THIN database. The drugs were selectively chosen to enable a comparison with previous work and for variety. It was found that no algorithm was generally superior and the algorithms' natural thresholds act at variable stringencies. Furthermore, none of the algorithms perform well at detecting rare ADRs. © 2013 Springer-Verlag Berlin Heidelberg.

Objectives: Recent studies of breast cancer data have identified seven distinct clinical phenotypes (groups) using immunohistochemical analysis and a range of different clustering techniques Consensus between unsupervised classification algorithms has been successfully used to categorise patients into these specific groups, but often at the expenses of not classifying the whole set It is known that fuzzy methodologies can provide linguistic based classification rules The objective of this study was to investigate the use of fuzzy methodologies to create an easy to interpret set of classification rules, capable of placing the large majority of patients into one of the specified groups Materials and methods: In this paper, we extend a data-driven fuzzy rule-based system for classification purposes (called 'fuzzy quantification subsethood-based algorithm') and combine it with a novel class assignment procedure The whole approach is then applied to a well characterised breast cancer dataset consisting of ten protein markers for over 1000 patients to refine previously identified groups and to present clinicians with a linguistic ruleset A range of statistical approaches was used to compare the obtained classes to previously obtained groupings and to assess the proportion of unclassified patients Results: A rule set was obtained from the algorithm which features one classification rule per class, using labels of High, Low or Omit for each biomarker, to determine the most appropriate class for each patient When applied to the whole set of patients, the distribution of the obtained classes had an agreement of 0.9 when assessed using Kendall's Tau with the original reference class distribution In doing so, only 38 patients out of 1073 remain unclassified, representing a more clinically usable class assignment algorithm Conclusion: The fuzzy algorithm provides a simple to interpret, linguistic rule set which classifies over 95 of breast cancer patients into one of seven clinical groups © 2013 Elsevier B.V.

Breast cancer is recognised to be a heterogeneous disease and the second most common cause of morbidity and mortality worldwide in women. Basal-like breast cancer (BLBC) is associated with aggressive characteristics including development of recurrent disease and reduced survival. BLBC has been defined in some studies as tumours lacking both oestrogen receptor and progesterone receptor protein expression. Gene expression studies have shown that these tumours are also associated with expression of basal-type cytokeratins, the phenotypic patterns of basal cytokeratin expression in BLBC have not been widely studied. A well-characterised series of 995 invasive breast cancers with a long-term follow up were investigated using immunohistochemical staining for four basal cytokeratins (CK5, CK5/6, CK14 and CK17). The data were analysed using univariate and clustering analysis. As a result BLBC, as defined by negativity for ER and HER2 showed variable positivity for basal cytokeratin expression: 61.7 CK5, 50.5 CK5/6, 24.2 CK14 and 23 CK17. These characteristics were associated with poor outcome characteristics including high histological grade, mitosis, pleomorphism and tumour size >1.5 cm. CK5 positivity was more associated with ER-, PgR-, TN and double ER-PgR-, than the other cytokeratins. Four different clusters of basal cytokeratin expression patterns were identified: (1) negativity for all basal cytokeratins, (2) CK5+/CK17-, (3) CK5-/CK17+ and (4) CK5+/CK17+. These patterns of basal cytokeratin expression associated with differences in patient outcome, clusters 1 and 3 showed better outcomes than cluster 4 and 2, with cluster 2 having the poorest prognosis. In conclusion, four basal cytokeratin expression patterns were identified in human breast cancer using unsupervised clustering analysis and these patterns are associated with differences in patient outcome. © 2013 Springer Science+Business Media New York.

Many algorithms have been proposed for the machine learning task of classification. One of the simplest methods, the naive Bayes classifier, has often been found to give good performance despite the fact that its underlying assumptions (of independence and a normal distribution of the variables) are perhaps violated. In previous work, we applied naive Bayes and other standard algorithms to a breast cancer database from Nottingham City Hospital in which the variables are highly non-normal and found that the algorithm performed well when predicting a class that had been derived from the same data. However, when we then applied naive Bayes to predict an alternative clinical variable, it performed much worse than other techniques. This motivated us to propose an alternative method, based on naive Bayes, which removes the requirement for the variables to be normally distributed, but retains the essential structure and other underlying assumptions of the method. We tested our novel algorithm on our breast cancer data and on three UCI datasets which also exhibited strong violations of normality. We found our algorithm outperformed naive Bayes in all four cases and outperformed multinomial logistic regression (MLR) in two cases. We conclude that our method offers a competitive alternative to MLR and naive Bayes when dealing with data sets in which non-normal distributions are observed. © 2011 Elsevier B.V. All rights reserved.

Objective: Despite the clinical similarities triple-negative and basal-like breast cancer are not synonymous. Indeed, not all basal-like cancers are negative for estrogen receptor, progesterone receptor and HER2 expression while triple-negative also encompasses other cancer types. P53 protein appears heterogeneously expressed in triple-negative breast cancers, suggesting that it may be associated with specific biological subgroups with a different outcome. Methods: We comparatively analyzed p53 expression in triple-negative tumors from two independent breast cancer case series (633 cases from the University of Ferrara and 1076 cases from the University of Nottingham). Results: In both case series, p53 protein expression was able to subdivide the triple-negative cases into two distinct subsets consistent with a different outcome. In fact, triple-negative patients with a p53 expressing tumor showed worse overall and event-free survival. Conclusions: The immunohistochemical evaluation of p53 expression may help in taming the currently stormy relationship between pathological (triple-negative tumors) and biological (basal breast cancers) classifications and in selecting patient subgroups with different biological features providing a potentially powerful prognostic contribution in triple-negative breast cancers. © The Author (2011). Published by Oxford University Press. All rights reserved.

Single clustering methods have often been used to elucidate clusters in high dimensional medical data, even though reliance on a single algorithm is known to be problematic. In this paper, we present a methodology to determine a set of 'core classes' by using a range of techniques to reach consensus across several different clustering algorithms, and to ascertain the key characteristics of these classes. We apply the methodology to immunohistochemical data from breast cancer patients. In doing so, we identify six core classes, of which several may be novel sub-groups not previously emphasised in literature. © 2010 Elsevier Ltd. All rights reserved.

The affinity propagation algorithm is applied to various problems of breast and cutaneous tumours subtyping using traditional biologic markers. The algorithm provides a procedure to determine the number of profiles to be considered. Well-known breast cancer case series and cutaneous melanoma were used to compare the results of the affinity propagation with the results obtained with standard algorithms and indexes for the optimal choice of the number of clusters. Results from affinity propagation are consistent with the results already obtained having the advantage of providing an indication about the number of clusters.

Post-translational histone modifications are known to be altered in cancer cells, and loss of selected histone acetylation and methylation marks has recently been shown to predict patient outcome in human carcinoma. Immunohistochemistry was used to detect a series of histone lysine acetylation (H3K9ac, H3K18ac, H4K12ac, and H4K16ac), lysine methylation (H3K4me2 and H4K20me3), and arginine methylation (H4R3me2) marks in a well-characterized series of human breast carcinomas (n = 880). Tissue staining intensities were assessed using blinded semiquantitative scoring. Validation studies were done using immunofluorescence staining and Western blotting. Our analyses revealed low or absent H4K16ac in the majority of breast cancer cases (78.9%), suggesting that this alteration may represent an early sign of breast cancer. There was a highly significant correlation between histone modifications status, tumor biomarker phenotype, and clinical outcome, where high relative levels of global histone acetylation and methylation were associated with a favorable prognosis and detected almost exclusively in luminal-like breast tumors (93%). Moderate to low levels of lysine acetylation (H3K9ac, H3K18ac, and H4K12ac), lysine (H3K4me2 and H4K20me3), and arginine methylation (H4R3me2) were observed in carcinomas of poorer prognostic subtypes, including basal carcinomas and HER-2-positive tumors. Clustering analysis identified three groups of histone displaying distinct pattern in breast cancer, which have distinct relationships to known prognostic factors and clinical outcome. This study identifies the presence of variations in global levels of histone marks in different grades, morphologic types, and phenotype classes of invasive breast cancer and shows that these differences have clinical significance.

The Fuzzy Integral (FI) is a non-linear aggregation operator which enables the fusion of information from multiple sources in respect to a Fuzzy Measure (FM) which captures the worth of both the individual sources and all their possible combinations. Based on the expected potential of non-linear aggregation offered by the FI, its application to decision-level fusion in ensemble classifiers, i.e. to fuse multiple classifiers outputs towards one superior decision level output, has recently been explored. A key example of such a FI-FM ensemble classification method is the Decision-level Fuzzy Integral Multiple Kernel Learning (DeFIMKL) algorithm, which aggregates the outputs of kernel based classifiers through the use of the Choquet FI with respect to a FM learned through a regularised quadratic programming approach. While the approach has been validated against a number of classifiers based on multiple kernel learning, it has thus far not been compared to the state-of-the-art in ensemble classification. Thus, this paper puts forward a detailed comparison of FI-FM based ensemble methods, specifically the DeFIMKL algorithm, with state-of-the art ensemble methods including Adaboost, Bagging, Random Forest and Majority Voting over 20 public datasets from the UCI machine learning repository. The results on the selected datasets suggest that the FI based ensemble classifier performs both well and efficiently, indicating that it is a viable alternative when selecting ensemble classifiers and indicating that the non-linear fusion of decision level outputs offered by the FI provides expected potential and warrants further study.

Food variety influences appetitive behaviour, motivation to eat and energy intake. Research found that repeated exposure to varied food images increases the motivation towards food in adults and children. This study investigates the effects of repetition on the modulation of early and late components of event-related potentials (ERPs) when participants passively viewed the same food and non-food images repeatedly. The motivational attention to food and non-food images were assessed in frontal, centroparietal, parietooccipital and occipitotemporal areas of the brain. Participants showed increased late positive potential (late ERP component) to high caloric image in the occipitotemporal region compared to low caloric and nonfood images. Similar effects could be seen in the early ERP component in the frontal region, but with reversed polarity. Data suggest that both the early and late ERP components show greater ERP amplitude when viewing high caloric images than low caloric and non-food images. Despite repeated exposure to same image, high caloric food continued to show sustained attention compared to low caloric and non-food image.

Side effects of prescription drugs present a serious issue. Existing algorithms that detect side effects generally require further analysis to confirm causality. In this paper we investigate attributes based on the Bradford-Hill causality criteria that could be used by a classifying algorithm to definitively identify side effects directly. We found that it would be advantageous to use attributes based on the association strength, temporality and specificity criteria. © 2013 IEEE.

In this paper, we describe a dataset relating to cellular and physical conditions of patients who are operated upon to remove colorectal tumours. This data provides a unique insight into immunological status at the point of tumour removal, tumour classification and post-operative survival. Attempts are made to cluster this dataset and important subsets of it in an effort to characterize the data and validate existing standards for tumour classification. It is apparent from optimal clustering that existing tumour classification is largely unrelated to immunological factors within a patient and that there may be scope for re-evaluating treatment options and survival estimates based on a combination of tumour physiology and patient histochemistry. © 2012 Polish Info Processing Socit.

Longitudinal observational databases have become a recent interest in the post marketing drug surveillance community due to their ability of presenting a new perspective for detecting negative side effects. Algorithms mining longitudinal observation databases are not restricted by many of the limitations associated with the more conventional methods that have been developed for spontaneous reporting system databases. In this paper we investigate the robustness of four recently developed algorithms that mine longitudinal observational databases by applying them to The Health Improvement Network (THIN) for six drugs with well document known negative side effects. Our results show that none of the existing algorithms was able to consistently identify known adverse drug reactions above events related to the cause of the drug and no algorithm was superior. © 2012 IEEE.

The wealth of computerised medical information becoming readily available presents the opportunity to examine patterns of illnesses, therapies and responses. These patterns may be able to predict illnesses that a patient is likely to develop, allowing the implementation of preventative actions. In this paper sequential rule mining is applied to a General Practice database to find rules involving a patients age, gender and medical history. By incorporating these rules into current health-care a patient can be highlighted as susceptible to a future illness based on past or current illnesses, gender and year of birth. This knowledge has the ability to greatly improve health-care and reduce health-care costs. © 2012 IEEE.

In this paper we present an original framework to extract representative groups from a dataset, and we validate it over a novel case study. The framework specifies the application of different clustering algorithms, then several statistical and visualisation techniques are used to characterise the results, and core classes are defined by consensus clustering. Classes may be verified using supervised classification algorithms to obtain a set of rules which may be useful for new data points in the future. This framework is validated over a novel set of histone markers for breast cancer patients. From a technical perspective, the resultant classes are well separated and characterised by low, medium and high levels of biological markers. Clinically, the groups appear to distinguish patients with poor overall survival from those with low grading score and better survival. Overall, this framework offers a promising methodology for elucidating core consensus groups from data. © 2010 IEEE.

Extracting usable and useful knowledge from large and complex data sets is a difficult and challenging problem. In this paper, we show how two complementary techniques have been used to tackle this problem in the context of breast cancer. Diagnosis concerns the identification of cancer within a patient; in contrast, prognosis concerns the prediction of the ongoing course of the disease, including issues such as the choice of potential treatments such as chemotherapy or drug therapy, in combination with estimation of chances (or length) of survival. Reliable prognosis depends on many factors, including the identification of the type of this heterogeneous disease. We first use a consensus clustering methodology to identify core, well-characterised sub-groups (or classes) of the disease based on a large database of protein biomarkers from over a thousand patients. We then use fuzzy rule induction and simplification algorithms to generate a simple, comprehensible set of rules for use in future model-based classification. The methods are described and their use is illustrated on real-world data. © ECMS.

The affinity propagation algorithm is applied to a problem of breast cancer subtyping using traditional biologic markers. The algorithm provides a procedure to determine the number of profiles to be considered. A well know breast cancer case series was used to compare the results of the affinity propagation with the results obtained with standard algorithms and indexes for the optimal choice of the number of clusters. Results from affinity propagation are consistent with the results already obtained having the advantage of providing an indication about the number of clusters. © 2008 IEEE.

The classification of breast cancer patients is of great importance in cancer diagnosis. During the last few years, many algorithms have been proposed for this task. In this paper, we review different supervised machine learning techniques for classification of a novel dataset and perform a methodological comparison of these. We used the C4.5 tree classifier, a Multilayer Perceptron and a naive Bayes classifier over a large set of tumour markers. We found good performance of the Multilayer Perceptron even when we reduced the number of features to be classified. We found naive Bayes achieved a competitive performance even though the assumption of normality of the data is strongly violated. © 2008 IEEE.

Clustering algorithms will, in general, either partition a given data set into a pre-specified number of clusters or will produce a hierarchy of clusters. In this paper we analyse several different clustering techniques and apply them to a particular data set of breast cancer data. When we do not know a priori which is the best number of groups, we use a range of different validity indices to test the quality of clustering results and to determine the best number of clusters. While for the K-means method there is not absolute agreement among the indices as to which is the best number of clusters, for the PAM algorithm all the indices indicate 4 as the best cluster number.