David McG. Squire
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
David McG. Squire has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
Thierry Pun
12 shared publications- Henning Müller
Informatics Institute, University of Applied Sciences Western Switzerland (HES-SO Valais), Sierre, Switzerland; Medical Faculty, University of Geneva, Geneva, Switzerland; The Sense Innovation and Research Center, Sion, Switzerland, Switzerland
11 shared publications - Wolfgang Müller
Media Education and Visualisation Group (MEVIS), University of Education Weingarten, Kirchplatz 2, 88250 Weingarten, Germany
6 shared publications Laurence Bull
Monash University
4 shared publications- Nabeel Mohammed
North South University
4 shared publications Stephane Marchand-Maillet
3 shared publications
Latest Publications
- Open Access
- 20 April 2026
- Published by Public Library of Science (PLoS) in PLOS Genetics
- Vol. 22 (4) , e1012102
Examining the association between fetal HLA-C, maternal KIR haplotypes and birth weight
Human birth weight is under stabilizing selection, seeking balance between extremes of high and low, thereby reducing fetal and maternal perinatal mortality risk. Certain combinations of maternal killer immunoglobulin-like receptor ( KIR ) and paternally derived fetal human leuokocyte antigen-C ( HLA-C ) alleles were previously associated with higher risk of high and low birth weight in a study with limited sample size (n = 1,316). Using recently developed methods to impute HLA and KIR haplotypes using single nucleotide polymorphism (SNP) genotype data, we tested associations of fetal HLA and maternal KIR genotypes with offspring birth weight in a large sample. We imputed KIR haplotypes using the KIR*IMP imputation software in 10,602 mother-offspring pairs of European descent from singleton pregnancies from five studies. Using mixed linear regression models to account for mothers with multiple children, we tested associations between maternal KIR A vs B haplotypes ( AA, AB/BA, BB genotypes) as well as copy number of activating receptor gene KIR2DS1 (0, 1, 2 copies of the gene) in the presence of fetal HLA C1/C2 alleles, and offspring birth weight. Associations were analyzed in each cohort before performing a meta-analysis to estimate the interaction effects between maternal KIR and fetal HLA-C2 on birth weight across the entire sample. The KIR haplotypes achieved imputation accuracy estimated at >95% in most of the cohorts. No interaction effects were observed between either the maternal A vs. B haplotype or the maternal KIR2DS1 locus and fetal HLA-C . When specifically trying to replicate the previously associated combination of maternal KIR2DS1 and paternally inherited fetal HLA-C2 , there was a negligible change in offspring birth weight for each additional KIR2DS1 allele and HLA-C2 of paternal origin (7g lower birth weight per allele [95% CI: -54, 40], P = 0.78). We found little evidence of association between birth weight and maternal KIR haplotypes or fetal HLA-C2 and were unable to replicate previously reported findings. Our observations reinforce the importance of replication and the use of large sample sizes in the validation of genetic associations.
- 20 April 2026
- Published by Public Library of Science (PLoS) in PLOS Genetics
- Vol. 22 (4) , e1012102
Open AccessExamining the association between fetal HLA-C, maternal KIR haplotypes and birth weight
Human birth weight is under stabilizing selection, seeking balance between extremes of high and low, thereby reducing fetal and maternal perinatal mortality risk. Certain combinations of maternal killer immunoglobulin-like receptor ( KIR ) and paternally derived fetal human leuokocyte antigen-C ( HLA-C ) alleles were previously associated with higher risk of high and low birth weight in a study with limited sample size (n = 1,316). Using recently developed methods to impute HLA and KIR haplotypes using single nucleotide polymorphism (SNP) genotype data, we tested associations of fetal HLA and maternal KIR genotypes with offspring birth weight in a large sample. We imputed KIR haplotypes using the KIR*IMP imputation software in 10,602 mother-offspring pairs of European descent from singleton pregnancies from five studies. Using mixed linear regression models to account for mothers with multiple children, we tested associations between maternal KIR A vs B haplotypes ( AA, AB/BA, BB genotypes) as well as copy number of activating receptor gene KIR2DS1 (0, 1, 2 copies of the gene) in the presence of fetal HLA C1/C2 alleles, and offspring birth weight. Associations were analyzed in each cohort before performing a meta-analysis to estimate the interaction effects between maternal KIR and fetal HLA-C2 on birth weight across the entire sample. The KIR haplotypes achieved imputation accuracy estimated at >95% in most of the cohorts. No interaction effects were observed between either the maternal A vs. B haplotype or the maternal KIR2DS1 locus and fetal HLA-C . When specifically trying to replicate the previously associated combination of maternal KIR2DS1 and paternally inherited fetal HLA-C2 , there was a negligible change in offspring birth weight for each additional KIR2DS1 allele and HLA-C2 of paternal origin (7g lower birth weight per allele [95% CI: -54, 40], P = 0.78). We found little evidence of association between birth weight and maternal KIR haplotypes or fetal HLA-C2 and were unable to replicate previously reported findings. Our observations reinforce the importance of replication and the use of large sample sizes in the validation of genetic associations.
- Open Access
Examining the association between fetal HLA-C , maternal KIR haplotypes and birth weight
Human birth weight is under stabilizing selection, seeking balance between extremes of high and low, thereby reducing fetal and maternal perinatal mortality risk. Certain combinations of maternal killer immunoglobulin-like receptor ( KIR ) and paternally derived fetal human leuokocyte antigen-C ( HLA-C ) alleles were previously associated with higher risk of high and low birth weight in a study with limited sample size (n=1,316). Using recently developed methods to impute HLA and KIR haplotypes using single nucleotide polymorphism (SNP) genotype data, we tested associations of fetal HLA and maternal KIR genotypes with offspring birth weight in a large sample. We imputed KIR haplotypes using the KIR*IMP imputation software in 10,602 mother-offspring pairs of European descent from singleton pregnancies from five studies. Using mixed linear regression models to account for mothers with multiple children, we tested associations between maternal KIR A vs B haplotypes ( AA, AB/BA, BB genotypes) as well as copy number of activating receptor gene KIR2DS1 (0, 1, 2 copies of the gene) in the presence of fetal HLA C1/C2 alleles, and offspring birth weight. Associations were analyzed in each cohort before performing a meta-analysis to estimate the interaction effects between maternal KIR and fetal HLA-C2 on birth weight across the entire sample. The KIR haplotypes achieved imputation accuracy estimated at >95% in most of the cohorts. No interaction effects were observed between either the maternal A vs. B haplotype or the maternal KIR2DS1 locus and fetal HLA-C . When specifically trying to replicate the previously associated combination of maternal KIR2DS1 and paternally inherited fetal HLA-C2 , there was a negligible change in offspring birth weight for each additional KIR2DS1 allele and HLA-C2 of paternal origin (7g lower birth weight per allele [95% CI: -54, 40], P = 0.78). We found little evidence of association between birth weight and maternal KIR haplotypes or fetal HLA-C2 and were unable to replicate previously reported findings. Our observations reinforce the importance of replication and the use of large sample sizes in the validation of genetic associations. Author Summary: Babies born with very high or low birth weights and their mothers are at a higher risk of illness and death than babies with weights close to average. Genes involved in the maternal immune system, called “ KIR ”, and the fetal immune system, called “ HLA-C ”, are important for early development of the placenta. Previously published research using a small sample has provided evidence for the role of interacting combinations of these genes in driving the spectrum of birth weight and maintaining the balancing selection of mother-child physiology that results in healthy birth outcomes. Here we harness recently developed methods to impute these genetic data to test associations of maternal KIR and fetal HLA with child’s birth weight in a larger sample. By examining >10,000 European ancestry mother-child pairs, we found no relationship between child’s birth weight and any of the genetic combinations we tested of KIR in the mother and HLA-C in the fetus. We show that despite biological plausibility, it is important to validate genetic associations through replication and using the largest sample sizes possible. Future research could benefit from including birth weights in the true extremes of the spectrum, using methods such as high throughput genome sequencing technologies which could provide more accurate data for these gene regions on a larger scale, and investigation in ancestrally diverse populations.
Open AccessExamining the association between fetal HLA-C , maternal KIR haplotypes and birth weight
Human birth weight is under stabilizing selection, seeking balance between extremes of high and low, thereby reducing fetal and maternal perinatal mortality risk. Certain combinations of maternal killer immunoglobulin-like receptor ( KIR ) and paternally derived fetal human leuokocyte antigen-C ( HLA-C ) alleles were previously associated with higher risk of high and low birth weight in a study with limited sample size (n=1,316). Using recently developed methods to impute HLA and KIR haplotypes using single nucleotide polymorphism (SNP) genotype data, we tested associations of fetal HLA and maternal KIR genotypes with offspring birth weight in a large sample. We imputed KIR haplotypes using the KIR*IMP imputation software in 10,602 mother-offspring pairs of European descent from singleton pregnancies from five studies. Using mixed linear regression models to account for mothers with multiple children, we tested associations between maternal KIR A vs B haplotypes ( AA, AB/BA, BB genotypes) as well as copy number of activating receptor gene KIR2DS1 (0, 1, 2 copies of the gene) in the presence of fetal HLA C1/C2 alleles, and offspring birth weight. Associations were analyzed in each cohort before performing a meta-analysis to estimate the interaction effects between maternal KIR and fetal HLA-C2 on birth weight across the entire sample. The KIR haplotypes achieved imputation accuracy estimated at >95% in most of the cohorts. No interaction effects were observed between either the maternal A vs. B haplotype or the maternal KIR2DS1 locus and fetal HLA-C . When specifically trying to replicate the previously associated combination of maternal KIR2DS1 and paternally inherited fetal HLA-C2 , there was a negligible change in offspring birth weight for each additional KIR2DS1 allele and HLA-C2 of paternal origin (7g lower birth weight per allele [95% CI: -54, 40], P = 0.78). We found little evidence of association between birth weight and maternal KIR haplotypes or fetal HLA-C2 and were unable to replicate previously reported findings. Our observations reinforce the importance of replication and the use of large sample sizes in the validation of genetic associations. Author Summary: Babies born with very high or low birth weights and their mothers are at a higher risk of illness and death than babies with weights close to average. Genes involved in the maternal immune system, called “ KIR ”, and the fetal immune system, called “ HLA-C ”, are important for early development of the placenta. Previously published research using a small sample has provided evidence for the role of interacting combinations of these genes in driving the spectrum of birth weight and maintaining the balancing selection of mother-child physiology that results in healthy birth outcomes. Here we harness recently developed methods to impute these genetic data to test associations of maternal KIR and fetal HLA with child’s birth weight in a larger sample. By examining >10,000 European ancestry mother-child pairs, we found no relationship between child’s birth weight and any of the genetic combinations we tested of KIR in the mother and HLA-C in the fetus. We show that despite biological plausibility, it is important to validate genetic associations through replication and using the largest sample sizes possible. Future research could benefit from including birth weights in the true extremes of the spectrum, using methods such as high throughput genome sequencing technologies which could provide more accurate data for these gene regions on a larger scale, and investigation in ancestrally diverse populations.
- 1 November 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 8, 1-8
An Evaluation of Sparseness as a Criterion for Selecting Independent Component Filters, When Applied to Texture Retrieval
In this paper we evaluate the utility of sparseness as a criterion for selecting a sub-set of independent component filters (ICF). Four sparseness measures were presented more than a decade ago by Le Borgne et al., but have since been ignored for ICF selection. In this paper we present our evaluation in the context of texture retrieval. We compare the sparseness-based method with the dispersal-based method, also proposed by Le Borgne et al., and the clustering-based method previously proposed by us. We show that the sparse filters and highly dispersed filters are quite different. In fact we show that highly dispersed filters tend to have lower sparseness. We also show that the sparse filters give better results compared to the highly dispersed filters when applied to texture retrieval. However the sparseness measures are calculated over filter response energies, making this method susceptible to choosing a redundant filter set. This issue is demonstrated and we show that ICF selected using our clustering-based method, which chooses a filter set with much lower redundancy, outperforms the sparse filters.
- 1 November 2014
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 8, 1-8
An Evaluation of Sparseness as a Criterion for Selecting Independent Component Filters, When Applied to Texture Retrieval
In this paper we evaluate the utility of sparseness as a criterion for selecting a sub-set of independent component filters (ICF). Four sparseness measures were presented more than a decade ago by Le Borgne et al., but have since been ignored for ICF selection. In this paper we present our evaluation in the context of texture retrieval. We compare the sparseness-based method with the dispersal-based method, also proposed by Le Borgne et al., and the clustering-based method previously proposed by us. We show that the sparse filters and highly dispersed filters are quite different. In fact we show that highly dispersed filters tend to have lower sparseness. We also show that the sparse filters give better results compared to the highly dispersed filters when applied to texture retrieval. However the sparseness measures are calculated over filter response energies, making this method susceptible to choosing a redundant filter set. This issue is demonstrated and we show that ICF selected using our clustering-based method, which chooses a filter set with much lower redundancy, outperforms the sparse filters.
- 1 September 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15224880,p. 2887-2891
Efficient and accurate independent component filter-based features for texure similarity
This paper evaluates the accuracy and efficiency of collection-specific texture features for Content-Based Image Retrieval. Independent Component Analysis is used to extract Independent Component Filters (ICF) from an image set. As these ICF are learned from the image set, the hypothesis is that they should provide texture features that are more effective than those extracted using generic filter banks. We describe a method for extracting candidate ICF from an image set, and choosing a representative subset from them. These are then used to extract image features. A simple CBIR system has been developed to evaluate the performance of these features on two standard texture image collections, compared with features extracted using multiple banks of Gabor filters. The results indicate that ICF-based features perform better than Gabor-based features, even when a much smaller number of ICF features is used than Gabor features. The ICF features are thus more accurate, and more efficient.
- 1 September 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15224880,p. 2887-2891
Efficient and accurate independent component filter-based features for texure similarity
This paper evaluates the accuracy and efficiency of collection-specific texture features for Content-Based Image Retrieval. Independent Component Analysis is used to extract Independent Component Filters (ICF) from an image set. As these ICF are learned from the image set, the hypothesis is that they should provide texture features that are more effective than those extracted using generic filter banks. We describe a method for extracting candidate ICF from an image set, and choosing a representative subset from them. These are then used to extract image features. A simple CBIR system has been developed to evaluate the performance of these features on two standard texture image collections, compared with features extracted using multiple banks of Gabor filters. The results indicate that ICF-based features perform better than Gabor-based features, even when a much smaller number of ICF features is used than Gabor features. The ICF features are thus more accurate, and more efficient.
- 1 June 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 95 (19493983) , 143-148
Improved texture features for CBIR using response scaling and locally normalised convolution
Independent Component Filters have been shown to be more effective for collection-specific CBIR than generic texture features such as Gabor filter banks. This paper describes changes to a filter selection method and feature extraction process that significantly improve the performance of both ICF- and Gabor filter-based features. We describe, and correct a possible oversight in, a previously published variance-based filter selection method. We also propose the use of locally normalised convolution as a technique to better match texture patterns in images with local intensity differences. We evaluate these changes using a simple CBIR system and our Precision and Recall results are significantly better than those previously published.
- 1 June 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 95 (19493983) , 143-148
Improved texture features for CBIR using response scaling and locally normalised convolution
Independent Component Filters have been shown to be more effective for collection-specific CBIR than generic texture features such as Gabor filter banks. This paper describes changes to a filter selection method and feature extraction process that significantly improve the performance of both ICF- and Gabor filter-based features. We describe, and correct a possible oversight in, a previously published variance-based filter selection method. We also propose the use of locally normalised convolution as a technique to better match texture patterns in images with local intensity differences. We evaluate these changes using a simple CBIR system and our Precision and Recall results are significantly better than those previously published.
- 1 January 2013
- Published by Springer Nature
- p. 83-95
Effectiveness of ICF Features for Collection-Specific CBIR
- 1 January 2013
- Published by Springer Nature
- p. 83-95
Effectiveness of ICF Features for Collection-Specific CBIR
- 18 January 2011
- Published by IGI Global Scientific Publishing
Design and Evaluation of a Content-Based Image Retrieval System
Design and Evaluation of a Content-Based Image Retrieval System: 10.4018/978-1-931777-07-0.ch002: The growth in size and accessibility of multimedia databases has changed our approach to information retrieval. Classical text-based systems show their
- 18 January 2011
- Published by IGI Global Scientific Publishing
Design and Evaluation of a Content-Based Image Retrieval System
Design and Evaluation of a Content-Based Image Retrieval System: 10.4018/978-1-931777-07-0.ch002: The growth in size and accessibility of multimedia databases has changed our approach to information retrieval. Classical text-based systems show their
- 18 January 2011
- Published by IGI Global Scientific Publishing
Design and Evaluation of a Content-Based Image Retrieval System
Design and Evaluation of a Content-Based Image Retrieval System: 10.4018/978-1-930708-00-6.ch007: The growth in size and accessibility of multimedia databases has changed our approach to information retrieval. Classical text-based systems show their
- 18 January 2011
- Published by IGI Global Scientific Publishing
Design and Evaluation of a Content-Based Image Retrieval System
Design and Evaluation of a Content-Based Image Retrieval System: 10.4018/978-1-930708-00-6.ch007: The growth in size and accessibility of multimedia databases has changed our approach to information retrieval. Classical text-based systems show their
- 1 December 2010
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 196-201
Local Adaptive SVM for Object Recognition
The Support Vector Machine (SVM) is an effective classification tool. Though extremely effective, SVMs are not a panacea. SVM training and testing is computationally expensive. Also, tuning the kernel parameters is a complicated procedure. On the other hand, the Nearest Neighbor (KNN) classifier is computationally efficient. In order to achieve the classification efficiency of an SVM and the computational efficiency of a KNN classifier, it has been shown previously that, rather than training a single global SVM, a separate SVM can be trained for the neighbourhood of each query point. In this work, we have extended this Local SVM (LSVM) formulation. Our Local Adaptive SVM (LASVM) formulation trains a local SVM in a modified neighborhood space of a query point. The main contributions of the paper are twofold: First, we present a novel LASVM algorithm to train a local SVM. Second, we discuss in detail the motivations behind the LSVM and LASVM formulations and its possible impacts on tuning the kernel parameters of an SVM. We found that training an SVM in a local adaptive neighborhood can result in significant classification performance gain. Experiments have been conducted on a selection of the UCIML, face, object, and digit databases.
- 1 December 2010
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 196-201
Local Adaptive SVM for Object Recognition
The Support Vector Machine (SVM) is an effective classification tool. Though extremely effective, SVMs are not a panacea. SVM training and testing is computationally expensive. Also, tuning the kernel parameters is a complicated procedure. On the other hand, the Nearest Neighbor (KNN) classifier is computationally efficient. In order to achieve the classification efficiency of an SVM and the computational efficiency of a KNN classifier, it has been shown previously that, rather than training a single global SVM, a separate SVM can be trained for the neighbourhood of each query point. In this work, we have extended this Local SVM (LSVM) formulation. Our Local Adaptive SVM (LASVM) formulation trains a local SVM in a modified neighborhood space of a query point. The main contributions of the paper are twofold: First, we present a novel LASVM algorithm to train a local SVM. Second, we discuss in detail the motivations behind the LSVM and LASVM formulations and its possible impacts on tuning the kernel parameters of an SVM. We found that training an SVM in a local adaptive neighborhood can result in significant classification performance gain. Experiments have been conducted on a selection of the UCIML, face, object, and digit databases.