Purpose:
M5L, a fully automated computer‐aided detection (CAD) system for the detection and segmentation of lung nodules in thoracic computed tomography (CT), is presented and validated on several ...image datasets.
Methods:
M5L is the combination of two independent subsystems, based on the Channeler Ant Model as a segmentation tool lung channeler ant model (lungCAM) and on the voxel‐based neural approach. The lungCAM was upgraded with a scan equalization module and a new procedure to recover the nodules connected to other lung structures; its classification module, which makes use of a feed‐forward neural network, is based of a small number of features (13), so as to minimize the risk of lacking generalization, which could be possible given the large difference between the size of the training and testing datasets, which contain 94 and 1019 CTs, respectively. The lungCAM (standalone) and M5L (combined) performance was extensively tested on 1043 CT scans from three independent datasets, including a detailed analysis of the full Lung Image Database Consortium/Image Database Resource Initiative database, which is not yet found in literature.
Results:
The lungCAM and M5L performance is consistent across the databases, with a sensitivity of about 70% and 80%, respectively, at eight false positive findings per scan, despite the variable annotation criteria and acquisition and reconstruction conditions. A reduced sensitivity is found for subtle nodules and ground glass opacities (GGO) structures. A comparison with other CAD systems is also presented.
Conclusions:
The M5L performance on a large and heterogeneous dataset is stable and satisfactory, although the development of a dedicated module for GGOs detection could further improve it, as well as an iterative optimization of the training procedure. The main aim of the present study was accomplished: M5L results do not deteriorate when increasing the dataset size, making it a candidate for supporting radiologists on large scale screenings and clinical programs.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK
•We evaluated the potential of radiomics and machine leaning on small data samples.•Training predictive models is challenging when small data samples are available.•Small proprietary samples may be ...integrated with larger publicly available cases.•Predictive performances were evaluated on each sample and on the merged one.•Inter-sample cross validation is feasible when samples have similar composition.
Predictive models based on radiomics and machine-learning (ML) need large and annotated datasets for training, often difficult to collect. We designed an operative pipeline for model training to exploit data already available to the scientific community. The aim of this work was to explore the capability of radiomic features in predicting tumor histology and stage in patients with non-small cell lung cancer (NSCLC).
We analyzed the radiotherapy planning thoracic CT scans of a proprietary sample of 47 subjects (L-RT) and integrated this dataset with a publicly available set of 130 patients from the MAASTRO NSCLC collection (Lung1). We implemented intra- and inter-sample cross-validation strategies (CV) for evaluating the ML predictive model performances with not so large datasets.
We carried out two classification tasks: histology classification (3 classes) and overall stage classification (two classes: stage I and II). In the first task, the best performance was obtained by a Random Forest classifier, once the analysis has been restricted to stage I and II tumors of the Lung1 and L-RT merged dataset (AUC = 0.72 ± 0.11). For the overall stage classification, the best results were obtained when training on Lung1 and testing of L-RT dataset (AUC = 0.72 ± 0.04 for Random Forest and AUC = 0.84 ± 0.03 for linear-kernel Support Vector Machine).
According to the classification task to be accomplished and to the heterogeneity of the available dataset(s), different CV strategies have to be explored and compared to make a robust assessment of the potential of a predictive model based on radiomics and ML.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
•Different segmentation pipelines may provide inconsistent quantification of brain structures.•The intra- and inter-method agreement between two popular segmentation software packages SPM12 and ...FreeSurfer v6.0.•SPM provides more consistent results both in the intra- and the inter-method agreement evaluation.•There are consistent biases in the estimates of gray matter and white matter between SPM and FreeSurfer.•To cross-validate the findings of each study against different segmentation methods before interpreting of the results.
The lack of inter-method agreement can produce inconsistent results in neuroimaging studies. We evaluated the intra-method repeatability and the inter-method reproducibility of two widely-used automatic segmentation methods for brain MRI: the FreeSurfer (FS) and the Statistical Parametric Mapping (SPM) software packages.
We segmented the gray matter (GM), the white matter (WM) and subcortical structures in test-retest MRI data of healthy volunteers from Kirby-21 and OASIS datasets. We used Pearson’s correlation (r), Bland-Altman plot and Dice index to study intra-method repeatability and inter-method reproducibility. In order to test whether different processing methods affect the results of a neuroimaging-based group study, we carried out a statistical comparison between male and female volume measures.
A high correlation was found between test-retest volume measures for both SPM (r in the 0.98–0.99 range) and FS (r in the 0.95–0.99 range). A non-null bias between test-retest FS volumes was detected for GM and WM in the OASIS dataset. The inter-method reproducibility analysis measured volume correlation values in the 0.72–0.98 range and the overlap between the segmented structures assessed by the Dice index was in the 0.76–0.83 range. SPM systematically provided significantly greater GM volumes and lower WM and subcortical volumes with respect to FS. In the male vs. female brain volume comparisons, inconsistencies arose for the OASIS dataset, where the gender-related differences appear subtler with respect to the Kirby dataset.
The inter-method reproducibility should be evaluated before interpreting the results of neuroimaging studies.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
Abstract A computer-aided detection (CAD) system for the identification of small pulmonary nodules in low-dose and thin-slice CT scans has been developed. The automated procedure for selecting the ...nodule candidates is mainly based on a filter enhancing spherical-shaped objects. A neural approach based on the classification of each single voxel of a nodule candidate has been purposely developed and implemented to reduce the amount of false-positive findings per scan. The CAD system has been trained to be sensitive to small internal and sub-pleural pulmonary nodules collected in a database of low-dose and thin-slice CT scans. The system performance has been evaluated on a data set of 39 CT containing 75 internal and 27 sub-pleural nodules. The FROC curve obtained on this data set shows high values of sensitivity to lung nodules (80–85% range) at an acceptable level of false positive findings per patient (10–13 FP/scan).
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK
Abstract A completely automated system for the identification of pleural nodules in low-dose and thin-slice computed tomography (CT) of the lung has been developed. The directional-gradient ...concentration method has been applied to the pleura surface and combined with a morphological opening -based procedure to generate a list of nodule candidates. Each nodule candidate is characterized by 12 morphological and textural features, which are analyzed by a rule-based filter and a neural classifier. This detection system has been developed and validated on a dataset of 42 annotated CT scans. The k -fold cross validation has been used to evaluate the neural classifier performance. The system performance variability due to different ground truth agreement levels is discussed. In particular, the poor 44% sensitivity obtained on the ground truth with agreement level 1 (nodules annotated by only one radiologist) with six FP per scan grows up to the 72% if the underlying ground truth is changed to the agreement level 2 (nodules annotated by two radiologists).
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK
In-silico clinical trials with digital patient models and simulated devices are an alternative to expensive and long clinical trials on patient population for testing X-ray breast imaging ...apparatuses. In this work, we simulated a linear-response a-Se detector as an X-ray absorber, neglecting some physical processes, such as electro-hole tracking and thermal noise. In order to tune characteristics of the simulated images toward those of the clinical scanners, the detector response curve, modulation transfer function (MTF), and normalized noise power spectrum (NNPS) were measured on a clinical mammographic unit. The same tests were replicated in-silico via a custom-made Monte Carlo code in order to define a suitable model to modify simulated images and to have realistic pixel values, noise, and spatial resolution. The proposed approach resulted to restore the slope and the magnitude of the NNPS in simulated images toward curves evaluated on a clinical scanner. Similarly, the proposed strategy for tuning noise and spatial resolution in simulated images led to a contrast-to-noise ratio (CNR) evaluated on a custom-made phantom which differed from those in measured images less than 16% in absolute value.
The mammography is the most effective procedure for an early diagnosis of the breast cancer. In this paper, an algorithm for detecting masses in mammographic images will be presented. The database ...consists of 3762 digital images acquired in several hospitals belonging to the MAGIC-5 collaboration (Medical Applications on a Grid Infrastructure Connection). A reduction of the whole image's area under investigation is achieved through a segmentation process, by means of a ROI Hunter algorithm, without loss of meaningful information. In the following classification step, feature extraction plays a fundamental role: some features give geometrical information, other ones provide shape parameters. Once the features are computed for each ROI, they are used as inputs to a supervised neural network with momentum. The output neuron provides the probability that the ROI is pathological or not. Results are provided in terms of ROC and FROC curves: the area under the ROC curve was found to be A Z =0.862plusmn0.007, and we get a 2.8 FP/Image at a sensitivity of 82%. This software is included in the CAD station actually working in the hospitals belonging to the MAGIC-5 Collaboration
Lung cancer is one of the most lethal types of cancer, because its early diagnosis is not good enough. In fact, the detection of pulmonary nodule, potential lung cancers, in Computed Tomography scans ...is a very challenging and time-consuming task for radiologists. To support radiologists, researchers have developed Computer-Aided Diagnosis (CAD) systems for the automated detection of pulmonary nodules in chest Computed Tomography scans. Despite the high level of technological developments and the proved benefits on the overall detection performance, the usage of Computer-Aided Diagnosis in clinical practice is far from being a common procedure. In this paper we investigate the causes underlying this discrepancy and present a solution to tackle it: the M5L WEB- and Cloud-based on-demand Computer-Aided Diagnosis. In addition, we prove how the combination of traditional imaging processing techniques with state-of-art advanced classification algorithms allows to build a system whose performance could be much larger than any Computer-Aided Diagnosis developed so far. This outcome opens the possibility to use the CAD as clinical decision support for radiologists.
Mass localization plays a crucial role in computer-aided detection (CAD) systems for the classification of suspicious regions in mammograms. In this article we present a completely automated ...classification system for the detection of masses in digitized mammographic images. The tool system we discuss consists in three processing levels: (a) Image segmentation for the localization of regions of interest (ROIs). This step relies on an iterative dynamical threshold algorithm able to select iso-intensity closed contours around gray level maxima of the mammogram. (b) ROI characterization by means of textural features computed from the gray tone spatial dependence matrix (GTSDM), containing second-order spatial statistics information on the pixel gray level intensity. As the images under study were recorded in different centers and with different machine settings, eight GTSDM features were selected so as to be invariant under monotonic transformation. In this way, the images do not need to be normalized, as the adopted features depend on the texture only, rather than on the gray tone levels, too. (c) ROI classification by means of a neural network, with supervision provided by the radiologist’s diagnosis. The CAD system was evaluated on a large database of 3369 mammographic images 2307 negative, 1062 pathological (or positive), containing at least one confirmed mass, as diagnosed by an expert radiologist. To assess the performance of the system, receiver operating characteristic (ROC) and free-response ROC analysis were employed. The area under the ROC curve was found to be
A
z
=
0.783
±
0.008
for the ROI-based classification. When evaluating the accuracy of the CAD against the radiologist-drawn boundaries, 4.23 false positives per image are found at 80% of mass sensitivity.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK