Implementable Deep Learning for Multi-sequence Proton MRI Lung Segmentation: A Multi-center, Multi-vendor, and Multi-disease Study

BACKGROUND

Recently, deep learning via convolutional neural networks (CNNs) has largely superseded conventional methods for proton (1 H)-MRI lung segmentation. However, previous deep learning studies have utilized single-center data and limited acquisition parameters.

PURPOSE

Develop a generalizable CNN for lung segmentation in 1 H-MRI, robust to pathology, acquisition protocol, vendor, and center.

STUDY TYPE

Retrospective.

POPULATION

A total of 809 1 H-MRI scans from 258 participants with various pulmonary pathologies (median age (range): 57 (6-85); 42% females) and 31 healthy participants (median age (range): 34 (23-76); 34% females) that were split into training (593 scans (74%); 157 participants (55%)), testing (50 scans (6%); 50 participants (17%)) and external validation (164 scans (20%); 82 participants (28%)) sets.

FIELD STRENGTH/SEQUENCE

1.5-T and 3-T/3D spoiled-gradient recalled and ultrashort echo-time 1 H-MRI.

ASSESSMENT

2D and 3D CNNs, trained on single-center, multi-sequence data, and the conventional spatial fuzzy c-means (SFCM) method were compared to manually delineated expert segmentations. Each method was validated on external data originating from several centers. Dice similarity coefficient (DSC), average boundary Hausdorff distance (Average HD), and relative error (XOR) metrics to assess segmentation performance.

STATISTICAL TESTS

Kruskal-Wallis tests assessed significances of differences between acquisitions in the testing set. Friedman tests with post hoc multiple comparisons assessed differences between the 2D CNN, 3D CNN, and SFCM. Bland-Altman analyses assessed agreement with manually derived lung volumes. A P value of <0.05 was considered statistically significant.

RESULTS

The 3D CNN significantly outperformed its 2D analog and SFCM, yielding a median (range) DSC of 0.961 (0.880-0.987), Average HD of 1.63 mm (0.65-5.45) and XOR of 0.079 (0.025-0.240) on the testing set and a DSC of 0.973 (0.866-0.987), Average HD of 1.11 mm (0.47-8.13) and XOR of 0.054 (0.026-0.255) on external validation data.

DATA CONCLUSION

The 3D CNN generated accurate 1 H-MRI lung segmentations on a heterogenous dataset, demonstrating robustness to disease pathology, sequence, vendor, and center.

EVIDENCE LEVEL

4.

TECHNICAL EFFICACY

Stage 1.

A hybrid model- and deep learning-based framework for functional lung image synthesis from multi-inflation CT and hyperpolarized gas MRI

BACKGROUND

Hyperpolarized gas MRI is a functional lung imaging modality capable of visualizing regional lung ventilation with exceptional detail within a single breath. However, this modality requires specialized equipment and exogenous contrast, which limits widespread clinical adoption. CT ventilation imaging employs various metrics to model regional ventilation from non-contrast CT scans acquired at multiple inflation levels and has demonstrated moderate spatial correlation with hyperpolarized gas MRI. Recently, deep learning (DL)-based methods, utilizing convolutional neural networks (CNNs), have been leveraged for image synthesis applications. Hybrid approaches integrating computational modeling and data-driven methods have been utilized in cases where datasets are limited with the added benefit of maintaining physiological plausibility.

PURPOSE

To develop and evaluate a multi-channel DL-based method that combines modeling and data-driven approaches to synthesize hyperpolarized gas MRI lung ventilation scans from multi-inflation, non-contrast CT and quantitatively compare these synthetic ventilation scans to conventional CT ventilation modeling.

METHODS

In this study, we propose a hybrid DL configuration that integrates model- and data-driven methods to synthesize hyperpolarized gas MRI lung ventilation scans from a combination of non-contrast, multi-inflation CT and CT ventilation modeling. We used a diverse dataset comprising paired inspiratory and expiratory CT and helium-3 hyperpolarized gas MRI for 47 participants with a range of pulmonary pathologies. We performed six-fold cross-validation on the dataset and evaluated the spatial correlation between the synthetic ventilation and real hyperpolarized gas MRI scans; the proposed hybrid framework was compared to conventional CT ventilation modeling and other non-hybrid DL configurations. Synthetic ventilation scans were evaluated using voxel-wise evaluation metrics such as Spearman's correlation and mean square error (MSE), in addition to clinical biomarkers of lung function such as the ventilated lung percentage (VLP). Furthermore, regional localization of ventilated and defect lung regions was assessed via the Dice similarity coefficient (DSC).

RESULTS

We showed that the proposed hybrid framework is capable of accurately replicating ventilation defects seen in the real hyperpolarized gas MRI scans, achieving a voxel-wise Spearman's correlation of 0.57 ± 0.17 and an MSE of 0.017 ± 0.01. The hybrid framework significantly outperformed CT ventilation modeling alone and all other DL configurations using Spearman's correlation. The proposed framework was capable of generating clinically relevant metrics such as the VLP without manual intervention, resulting in a Bland-Altman bias of 3.04%, significantly outperforming CT ventilation modeling. Relative to CT ventilation modeling, the hybrid framework yielded significantly more accurate delineations of ventilated and defect lung regions, achieving a DSC of 0.95 and 0.48 for ventilated and defect regions, respectively.

CONCLUSION

The ability to generate realistic synthetic ventilation scans from CT has implications for several clinical applications, including functional lung avoidance radiotherapy and treatment response mapping. CT is an integral part of almost every clinical lung imaging workflow and hence is readily available for most patients; therefore, synthetic ventilation from non-contrast CT can provide patients with wider access to ventilation imaging worldwide.

Longitudinal Lung Function Assessment of Patients Hospitalized With COVID-19 Using 1H and 129Xe Lung MRI

BACKGROUND

Microvascular abnormalities and impaired gas transfer have been observed in patients with COVID-19. The progression of pulmonary changes in these patients remains unclear.

RESEARCH QUESTION

Do patients hospitalized with COVID-19 without evidence of architectural distortion on structural imaging exhibit longitudinal improvements in lung function measured by using 1H and 129Xe MRI between 6 and 52 weeks following hospitalization?

STUDY DESIGN AND METHODS

Patients who were hospitalized with COVID-19 pneumonia underwent a pulmonary 1H and 129Xe MRI protocol at 6, 12, 25, and 51 weeks following hospital admission in a prospective cohort study between November 2020 and February 2022. The imaging protocol was as follows: 1H ultra-short echo time, contrast-enhanced lung perfusion, 129Xe ventilation, 129Xe diffusion-weighted, and 129Xe spectroscopic imaging of gas exchange.

RESULTS

Nine patients were recruited (age 57 ± 14 [median ± interquartile range] years; six of nine patients were male). Patients underwent MRI at 6 (n = 9), 12 (n = 9), 25 (n = 6), and 51 (n = 8) weeks following hospital admission. Patients with signs of interstitial lung damage were excluded. At 6 weeks, patients exhibited impaired 129Xe gas transfer (RBC to membrane fraction), but lung microstructure was not increased (apparent diffusion coefficient and mean acinar airway dimensions). Minor ventilation abnormalities present in four patients were largely resolved in the 6- to 25-week period. At 12 weeks, all patients with lung perfusion data (n = 6) showed an increase in both pulmonary blood volume and flow compared with 6 weeks, although this was not statistically significant. At 12 weeks, significant improvements in 129Xe gas transfer were observed compared with 6-week examinations; however, 129Xe gas transfer remained abnormally low at weeks 12, 25, and 51.

INTERPRETATION

129Xe gas transfer was impaired up to 1 year following hospitalization in patients who were hospitalized with COVID-19 pneumonia, without evidence of architectural distortion on structural imaging, whereas lung ventilation was normal at 52 weeks.

PhysVENeT: a physiologically-informed deep learning-based framework for the synthesis of 3D hyperpolarized gas MRI ventilation

Functional lung imaging modalities such as hyperpolarized gas MRI ventilation enable visualization and quantification of regional lung ventilation; however, these techniques require specialized equipment and exogenous contrast, limiting clinical adoption. Physiologically-informed techniques to map proton (¹H)-MRI ventilation have been proposed. These approaches have demonstrated moderate correlation with hyperpolarized gas MRI. Recently, deep learning (DL) has been used for image synthesis applications, including functional lung image synthesis. Here, we propose a 3D multi-channel convolutional neural network that employs physiologically-informed ventilation mapping and multi-inflation structural ¹H-MRI to synthesize 3D ventilation surrogates (PhysVENeT). The dataset comprised paired inspiratory and expiratory ¹H-MRI scans and corresponding hyperpolarized gas MRI scans from 170 participants with various pulmonary pathologies. We performed fivefold cross-validation on 150 of these participants and used 20 participants with a previously unseen pathology (post COVID-19) for external validation. Synthetic ventilation surrogates were evaluated using voxel-wise correlation and structural similarity metrics; the proposed PhysVENeT framework significantly outperformed conventional ¹H-MRI ventilation mapping and other DL approaches which did not utilize structural imaging and ventilation mapping. PhysVENeT can accurately reflect ventilation defects and exhibits minimal overfitting on external validation data compared to DL approaches that do not integrate physiologically-informed mapping.

A Dual-Channel Deep Learning Approach for Lung Cavity Estimation From Hyperpolarized Gas and Proton MRI

BACKGROUND

Hyperpolarized gas MRI can quantify regional lung ventilation via biomarkers, including the ventilation defect percentage (VDP). VDP is computed from segmentations derived from spatially co-registered functional hyperpolarized gas and structural proton (1 H)-MRI. Although acquired at similar lung inflation levels, they are frequently misaligned, requiring a lung cavity estimation (LCE). Recently, single-channel, mono-modal deep learning (DL)-based methods have shown promise for pulmonary image segmentation problems. Multichannel, multimodal approaches may outperform single-channel alternatives.

PURPOSE

We hypothesized that a DL-based dual-channel approach, leveraging both 1 H-MRI and Xenon-129-MRI (129 Xe-MRI), can generate LCEs more accurately than single-channel alternatives.

STUDY TYPE

Retrospective.

POPULATION

A total of 480 corresponding 1 H-MRI and 129 Xe-MRI scans from 26 healthy participants (median age [range]: 11 [8-71]; 50% females) and 289 patients with pulmonary pathologies (median age [range]: 47 [6-83]; 51% females) were split into training (422 scans [88%]; 257 participants [82%]) and testing (58 scans [12%]; 58 participants [18%]) sets.

FIELD STRENGTH/SEQUENCE

1.5-T, three-dimensional (3D) spoiled gradient-recalled 1 H-MRI and 3D steady-state free-precession 129 Xe-MRI.

ASSESSMENT

We developed a multimodal DL approach, integrating 129 Xe-MRI and 1 H-MRI, in a dual-channel convolutional neural network. We compared this approach to single-channel alternatives using manually edited LCEs as a benchmark. We further assessed a fully automatic DL-based framework to calculate VDPs and compared it to manually generated VDPs.

STATISTICAL TESTS

Friedman tests with post hoc Bonferroni correction for multiple comparisons compared single-channel and dual-channel DL approaches using Dice similarity coefficient (DSC), average boundary Hausdorff distance (average HD), and relative error (XOR) metrics. Bland-Altman analysis and paired t-tests compared manual and DL-generated VDPs. A P value < 0.05 was considered statistically significant.

RESULTS

The dual-channel approach significantly outperformed single-channel approaches, achieving a median (range) DSC, average HD, and XOR of 0.967 (0.867-0.978), 1.68 mm (37.0-0.778), and 0.066 (0.246-0.045), respectively. DL-generated VDPs were statistically indistinguishable from manually generated VDPs (P = 0.710).

DATA CONCLUSION

Our dual-channel approach generated LCEs, which could be integrated with ventilated lung segmentations to produce biomarkers such as the VDP without manual intervention.

EVIDENCE LEVEL

4.

TECHNICAL EFFICACY

Stage 1.

129 Xe and Free-Breathing 1 H Ventilation MRI in Patients With Cystic Fibrosis: A Dual-Center Study

BACKGROUND

Free-breathing 1 H ventilation MRI shows promise but only single-center validation has yet been performed against methods which directly image lung ventilation in patients with cystic fibrosis (CF).

PURPOSE

To investigate the relationship between 129 Xe and 1 H ventilation images using data acquired at two centers.

STUDY TYPE

Sequence comparison.

POPULATION

Center 1; 24 patients with CF (12 female) aged 9-47 years. Center 2; 7 patients with CF (6 female) aged 13-18 years, and 6 healthy controls (6 female) aged 21-31 years. Data were acquired in different patients at each center.

FIELD STRENGTH/SEQUENCE

1.5 T, 3D steady-state free precession and 2D spoiled gradient echo.

ASSESSMENT

Subjects were scanned with 129 Xe ventilation and 1 H free-breathing MRI and performed pulmonary function tests. Ventilation defect percent (VDP) was calculated using linear binning and images were visually assessed by H.M., L.J.S., and G.J.C. (10, 5, and 8 years' experience).

STATISTICAL TESTS

Correlations and linear regression analyses were performed between 129 Xe VDP, 1 H VDP, FEV1 , and LCI. Bland-Altman analysis of 129 Xe VDP and 1 H VDP was carried out. Differences in metrics were assessed using one-way ANOVA or Kruskal-Wallis tests.

RESULTS

129 Xe VDP and 1 H VDP correlated strongly with; each other (r = 0.84), FEV1 z-score (129 Xe VDP r = -0.83, 1 H VDP r = -0.80), and LCI (129 Xe VDP r = 0.91, 1 H VDP r = 0.82). Bland-Altman analysis of 129 Xe VDP and 1 H VDP from both centers had a bias of 0.07% and limits of agreement of -16.1% and 16.2%. Linear regression relationships of VDP with FEV1 were not significantly different between 129 Xe and 1 H VDP (P = 0.08), while 129 Xe VDP had a stronger relationship with LCI than 1 H VDP.

DATA CONCLUSION

1 H ventilation MRI shows large-scale agreement with 129 Xe ventilation MRI in CF patients with established lung disease but may be less sensitive to subtle ventilation changes in patients with early-stage lung disease.

EVIDENCE LEVEL

2 TECHNICAL EFFICACY: Stage 2.

Image Phenotyping of Preterm-Born Children Using Hyperpolarized 129Xe Lung Magnetic Resonance Imaging and Multiple-Breath Washout

Rationale: Preterm birth is associated with low lung function in childhood, but little is known about the lung microstructure in childhood. Objectives: We assessed the differential associations between the historical diagnosis of bronchopulmonary dysplasia (BPD) and current lung function phenotypes on lung ventilation and microstructure in preterm-born children using hyperpolarized 129Xe ventilation and diffusion-weighted magnetic resonance imaging (MRI) and multiple-breath washout (MBW). Methods: Data were available from 63 children (aged 9-13 yr), including 44 born preterm (⩽34 weeks' gestation) and 19 term-born control subjects (⩾37 weeks' gestation). Preterm-born children were classified, using spirometry, as prematurity-associated obstructive lung disease (POLD; FEV1 < lower limit of normal [LLN] and FEV1/FVC < LLN), prematurity-associated preserved ratio of impaired spirometry (FEV1 < LLN and FEV1/FVC ⩾ LLN), preterm-(FEV1 ⩾ LLN) and term-born control subjects, and those with and without BPD. Ventilation heterogeneity metrics were derived from 129Xe ventilation MRI and SF6 MBW. Alveolar microstructural dimensions were derived from 129Xe diffusion-weighted MRI. Measurements and Main Results: 129Xe ventilation defect percentage and ventilation heterogeneity index were significantly increased in preterm-born children with POLD. In contrast, mean 129Xe apparent diffusion coefficient, 129Xe apparent diffusion coefficient interquartile range, and 129Xe mean alveolar dimension interquartile range were significantly increased in preterm-born children with BPD, suggesting changes of alveolar dimensions. MBW metrics were all significantly increased in the POLD group compared with preterm- and term-born control subjects. Linear regression confirmed the differential effects of obstructive disease on ventilation defects and BPD on lung microstructure. Conclusion: We show that ventilation abnormalities are associated with POLD, and BPD in infancy is associated with abnormal lung microstructure.

Xenon ventilation MRI in difficult asthma: initial experience in a clinical setting

Background

Hyperpolarised gas magnetic resonance imaging (MRI) can be used to assess ventilation patterns. Previous studies have shown the image-derived metric of ventilation defect per cent (VDP) to correlate with forced expiratory volume in 1 s (FEV₁)/forced vital capacity (FVC) and FEV₁ in asthma.

Objectives

The aim of this study was to explore the utility of hyperpolarised xenon-129 (¹²⁹Xe) ventilation MRI in clinical care and examine its relationship with spirometry and other clinical metrics in people seen in a severe asthma service.

Methods

26 people referred from a severe asthma clinic for MRI scanning were assessed by contemporaneous ¹²⁹Xe MRI and spirometry. A subgroup of 18 patients also underwent reversibility testing with spirometry and MRI. Quantitative MRI measures of ventilation were calculated, VDP and the ventilation heterogeneity index (VH_I), and compared to spirometry, Asthma Control Questionnaire 7 (ACQ7) and blood eosinophil count. Images were reviewed by a multidisciplinary team.

Results

VDP and VH_I correlated with FEV₁, FEV₁/FVC and forced expiratory flow between 25% and 75% of FVC but not with ACQ7 or blood eosinophil count. Discordance of MRI imaging and symptoms and/or pulmonary function tests also occurred, prompting diagnostic re-evaluation in some cases.

Conclusion

Hyperpolarised gas MRI provides a complementary method of assessment in people with difficult to manage asthma in a clinical setting. When used as a tool supporting clinical care in a severe asthma service, occurrences of discordance between symptoms, spirometry and MRI scanning indicate how MRI scanning may add to a management pathway.

This article demonstrates the feasibility of using ¹²⁹Xe MRI in clinical practice. Discordance between symptoms, spirometry and MRI can support the use of further treatment or suggest coexisting breathing control issues or laryngeal disorders.https://bit.ly/3ky4oXP

Patterns of regional lung physiology in cystic fibrosis using ventilation magnetic resonance imaging and multiple-breath washout

Hyperpolarised helium-3 (3He) ventilation magnetic resonance imaging (MRI) and multiple-breath washout (MBW) are sensitive methods for detecting lung disease in cystic fibrosis (CF). We aimed to explore their relationship across a broad range of CF disease severity and patient age, as well as assess the effect of inhaled lung volume on ventilation distribution.

32 children and adults with CF underwent MBW and 3He-MRI at a lung volume of end-inspiratory tidal volume (EIVT). In addition, 28 patients performed 3He-MRI at total lung capacity. 3He-MRI scans were quantitatively analysed for ventilation defect percentage (VDP), ventilation heterogeneity index (VHI) and the number and size of individual contiguous ventilation defects. From MBW, the lung clearance index, convection-dependent ventilation heterogeneity (Scond) and convection–diffusion-dependent ventilation heterogeneity (Sacin) were calculated.

VDP and VHI at EIVT strongly correlated with lung clearance index (r=0.89 and r=0.88, respectively), Sacin (r=0.84 and r=0.82, respectively) and forced expiratory volume in 1 s (FEV1) (r=−0.79 and r=−0.78, respectively). Two distinct 3He-MRI patterns were highlighted: patients with abnormal FEV1 had significantly (p<0.001) larger, but fewer, contiguous defects than those with normal FEV1, who tended to have numerous small volume defects. These two MRI patterns were delineated by a VDP of ∼10%. At total lung capacity, when compared to EIVT, VDP and VHI reduced in all subjects (p<0.001), demonstrating improved ventilation distribution and regions of volume-reversible and nonreversible ventilation abnormalities.

Three-dimensional DNA image cytometry by optical projection tomographic microscopy for early cancer diagnosis

Aneuploidy is typically assessed by flow cytometry (FCM) and image cytometry (ICM). We used optical projection tomographic microscopy (OPTM) for assessing cellular DNA content using absorption and fluorescence stains. OPTM combines some of the attributes of both FCM and ICM and generates isometric high-resolution three-dimensional (3-D) images of single cells. Although the depth of field of the microscope objective was in the submicron range, it was extended by scanning the objective's focal plane. The extended depth of field image is similar to a projection in a conventional x-ray computed tomography. These projections were later reconstructed using computed tomography methods to form a 3-D image. We also present an automated method for 3-D nuclear segmentation. Nuclei of chicken, trout, and triploid trout erythrocyte were used to calibrate OPTM. Ratios of integrated optical densities extracted from 50 images of each standard were compared to ratios of DNA indices from FCM. A comparison of mean square errors with thionin, hematoxylin, Feulgen, and SYTOX green was done. Feulgen technique was preferred as it showed highest stoichiometry, least variance, and preserved nuclear morphology in 3-D. The addition of this quantitative biomarker could further strengthen existing classifiers and improve early diagnosis of cancer using 3-D microscopy.

Local noise estimation in low-dose chest CT images

PURPOSE

Image noise in computed tomography (CT) images may have significant local variation due to tissue properties, dose, and location of the X-ray source. We developed and tested an automated tissue-based estimator method for estimating local noise in CT images.

METHOD

An automated TBE method for estimating the local noise in CT image in 3 steps was developed: (1) Partition the image into homogeneous and transition regions, (2) For each pixel in the homogeneous regions, compute the standard deviation in a 15 x 15 x 1 voxel local region using only pixels from the same homogeneous region, and (3) Interpolate the noise estimate from the homogeneous regions in the transition regions. Noise-aware fat segmentation was implemented. Experiments were conducted on the anthropomorphic phantom and in vivo low-dose chest CT scans to validate the TBE, characterize the magnitude of local noise variation, and determine the sensitivity of noise estimates to the size of the region in which noise is computed. The TBE was tested on all scans from the Early Lung Cancer Action Program public database. The TBE was evaluated quantitatively on the phantom data and qualitatively on the in vivo data.

RESULTS

The results show that noise can vary locally by over 200 Hounsfield units on low-dose in vivo chest CT scans and that the TBE can characterize these noise variations within 5 %. The new fat segmentation algorithm successfully improved segmentation on all 50 scans tested.

CONCLUSION

The TBE provides a means to estimate noise for image quality monitoring, optimization of denoising algorithms, and improvement of segmentation algorithms. The TBE was shown to accurately characterize the large local noise variations that occur due to changes in material, dose, and X-ray source location.

The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): a completed reference database of lung nodules on CT scans

PURPOSE

The development of computer-aided diagnostic (CAD) methods for lung nodule detection, classification, and quantitative assessment can be facilitated through a well-characterized repository of computed tomography (CT) scans. The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI) completed such a database, establishing a publicly available reference for the medical imaging research community. Initiated by the National Cancer Institute (NCI), further advanced by the Foundation for the National Institutes of Health (FNIH), and accompanied by the Food and Drug Administration (FDA) through active participation, this public-private partnership demonstrates the success of a consortium founded on a consensus-based process.

METHODS

Seven academic centers and eight medical imaging companies collaborated to identify, address, and resolve challenging organizational, technical, and clinical issues to provide a solid foundation for a robust database. The LIDC/IDRI Database contains 1018 cases, each of which includes images from a clinical thoracic CT scan and an associated XML file that records the results of a two-phase image annotation process performed by four experienced thoracic radiologists. In the initial blinded-read phase, each radiologist independently reviewed each CT scan and marked lesions belonging to one of three categories ("nodule > or =3 mm," "nodule <3 mm," and "non-nodule > or =3 mm"). In the subsequent unblinded-read phase, each radiologist independently reviewed their own marks along with the anonymized marks of the three other radiologists to render a final opinion. The goal of this process was to identify as completely as possible all lung nodules in each CT scan without requiring forced consensus.

RESULTS

The Database contains 7371 lesions marked "nodule" by at least one radiologist. 2669 of these lesions were marked "nodule > or =3 mm" by at least one radiologist, of which 928 (34.7%) received such marks from all four radiologists. These 2669 lesions include nodule outlines and subjective nodule characteristic ratings.

CONCLUSIONS

The LIDC/IDRI Database is expected to provide an essential medical imaging research resource to spur CAD development, validation, and dissemination in clinical practice.

A public image database to support research in computer aided diagnosis

The Public Lung Database to address drug response (PLD) has been developed to support research in computer aided diagnosis (CAD). Originally established for applications involving the characterization of pulmonary nodules, the PLD has been augmented to provide initial datasets for CAD research of other diseases. In general, the best performance for a CAD system is achieved when it is trained with a large amount of well documented data. Such training databases are very expensive to create and their lack of general availability limits the targets that can be considered for CAD applications and hampers development of the CAD field. The approach taken with the PLD has been to make available small datasets together with both manual and automated documentation. Furthermore, datasets with special properties are provided either to span the range of task complexity or to provide small change repeat images for direct calibration and evaluation of CAD systems. This resource offers a starting point for other research groups wishing to pursue CAD research in new directions. It also provides an on-line reference for better defining the issues relating to specific CAD tasks.

Automated nodule location and size estimation using a multi-scale Laplacian of Gaussian filtering approach

Estimation of nodule location and size is an important pre-processing step in some nodule segmentation algorithms to determine the size and location of the region of interest. Ideally, such estimation methods will consistently find the same nodule location regardless of where the the seed point (provided either manually or by a nodule detection algorithm) is placed relative to the "true" center of the nodule, and the size should be a reasonable estimate of the true nodule size. We developed a method that estimates nodule location and size using multi-scale Laplacian of Gaussian (LoG) filtering. Nodule candidates near a given seed point are found by searching for blob-like regions with high filter response. The candidates are then pruned according to filter response and location, and the remaining candidates are sorted by size and the largest candidate selected. This method was compared to a previously published template-based method. The methods were evaluated on the basis of stability of the estimated nodule location to changes in the initial seed point and how well the size estimates agreed with volumes determined by a semi-automated nodule segmentation method. The LoG method exhibited better stability to changes in the seed point, with 93% of nodules having the same estimated location even when the seed point was altered, compared to only 52% of nodules for the template-based method. Both methods also showed good agreement with sizes determined by a nodule segmentation method, with an average relative size difference of 5% and -5% for the LoG and template-based methods respectively.

Estimation of anatomical locations using standard frame of reference in chest CT scans

We propose a method to establish the standard chest frame of reference (CFOR) using the rib cage in a lung CT scan. Such a reference frame is essential for referring to a certain location within a chest region and may facilitate the registration across multiple scans of a given subject as well as the comparative studies within a cohort of subjects. The robustness of the established CFOR was evaluated by estimating the anatomical locations within chest in the follow-up scan given the location in the first scan. Specifically, tracheal bifurcation point of airway tree and the center of pulmonary nodule were used as the anatomical points of interest. The CFOR was also used for exploring the spatial distribution of the anatomical location for a large number of individuals. The results show that on average the point of interest can be estimated accurately within 10.3 mm for the bifurcation point and within 12.5 mm for the pulmonary nodule's center point. Further analyzing the spatial distribution of the CFOR coordinates across 86 subjects shows that we can localize the bifurcation point to the small subregion within the CFOR.

Production and genetic characterization of near-isogenic lines in the bread-wheat cultivar Alpe

Two biotypes of the bread-wheat cultivar Alpe were shown to possess contrasting alleles at each of the glutenin (Glu-B1, Glu-D1, Glu-B3 and Glu-D3) and gliadin (Gli-B1 and Gli-D1) loci on chromosomes 1B and 1D. Fourteen near-isogenic lines (NILs) were produced by crossing these biotypes and used to determine the genetic control of both low-molecular-weight (LMW) glutenin subunits and gliadins by means of one-dimensional or two-dimensional electrophoresis. Genes coding for the B, C and D groups of EMW subunits were found to be inherited in clusters tightly linked with those controlling gliadins. Southern-blot analysis of total genomic DNAs hybridized to a γ-gliadin-specific cDNA clone revealed that seven NILs lack both the Gli-D1 and Glu-D3 loci on chromosome 1D. Segregation data indicated that these "null" alleles are normally inherited. Comparison of the "null" NILs with those possessing allele b at the Glu-D3 locus showed one B subunit, seven C subunits and two D subunits, as fractionated by two-dimensional A-PAGExSDS-PAGE, to be encoded by this allele. Alleles b and k at Glu-B3 were found to code for two C subunits plus eight and six B subunits respectively, whereas alleles b and k at Gli-B1 each controlled the synthesis of two β-gliadins, one γ and two ω-gliadins. The novel Gli-B5 locus coding for two ω-gliadins was shown to recombine with the Gli-B1 locus on chromosome 1B. The two-dimensional map of glutenin subunits showed α-gliadins encoded at the Gli-A2 locus on chromosome 6A. The use of Alpe NILs in the study of the individual and combined effects of glutenin subunits on dough properties is discussed.

Genetics of gliadins coded by the group 1 chromosomes in the high-quality bread wheat cultivar Neepawa

The inheritance and biochemical properties of gliadins controlled by the group 1 chromosomes of the high-quality bread wheat cultivar Neepawa were studied in the progeny of the cross Neepawa x Costantino by six different electrophoretic procedures. Chromosome 1B of Neepawa contains two gliadin loci, one (Gli-B1) coding for at least six ω- or γ-gliadins, the other (Gli-B3) controlling the synthesis of gliadin N6 only. The map distance between these loci was calculated as 22.1 cM. Amongst the chromosome 1A gliadins, three proteins are encoded at the Gli-A1 locus whereas polypeptides N14-N15-N16 are controlled by a remote locus which recombines with Gli-A1. Six other gliadins are controlled by a gene cluster at Gli-D1 on chromosome 1D. Canadian wheat cultivars sharing the Gli-B1 allele of Neepawa were found to differ in the presence or absence of gliadin N6. The electrophoretic mobilities of proteins N6 and N14-N15-N16 were unaffected by the addition of a reducing agent during two-dimensional sodium dodecyl sulphate polyacrylamid-gel electrophoresis, suggesting the absence of intra-chain disulphide bonds in their structure.