Deep learning methods are the state-of-the-art for medical imaging segmentation tasks. Still, numerous segmentation algorithms based on heuristic-based methods have been proposed with exceptional results. To validate segmentation algorithms, manual annotations are typically considered as ground truth. However, manual annotations often suffer from inter/intra-operator variability and can also be occasionally inaccurate, especially when considering time-consuming and precise tasks. A sample case is the manual delineation of the lumen-intima (LI) and media-adventitia (MA) borders for intima-media thickness (IMT) measurement in B-mode ultrasound images. In this work, a novel hybrid learning paradigm which combines manual segmentations with the automatic segmentation of a dynamic programming technique for ground truth determination is presented. A profile consensus strategy is proposed to construct the hybrid ground truth. Two open-source datasets (n=2576) were employed for training four deep learning networks using the hybrid learning paradigm and three single source training targets as a comparison. The pipeline was fixed across the four tests and included a Faster R-CNN detection network to locate the carotid artery and then subsequent division into patches which were segmented using a UNet. The validation of the results was performed on an external test set comparing the predictions of the four different models to the annotations of three independent manual operators. The hybrid learning paradigm showed the best overall segmentation results (Dice=0.907±0.037, p<0.001) and demonstrated an exceptional correlation between the mean of three operators and the automatic measure (ICC(2,1)=0.958), demonstrating how the incorporation of heuristic-based segmentation methods within the learning paradigm of a deep neural network can enhance and improve final segmentation performance results.
Diabetic foot ulcer (DFU) is a diabetic complication due to peripheral vasculopathy and neuropathy. A promising technology for wound healing in DFU is low-level light therapy (LLLT). Despite several studies showing positive effects of LLLT on DFU, LLLT’s physiological effects have not yet been studied. The objective of this study was to investigate vascular and nervous systems modification in DFU after LLLT. Two samples of 45 DFU patients and 11 healthy controls (HCs) were recruited. The total hemoglobin (totHb) concentration change was monitored before and after LLLT by near-infrared spectroscopy and analyzed in time and frequency domains. The spectral power of the totHb changes in the very-low frequency (VLF, 20 to 60 mHz) and low frequency (LF, 60 to 140 mHz) bandwidths was calculated. Data analysis revealed a mean increase of totHb concentration after LLLT in DFU patients, but not in HC. VLF/LF ratio decreased significantly after the LLLT period in DFU patients (indicating an increased activity of the autonomic nervous system), but not in HC. Eventually, different treatment intensities in LLLT therapy showed a different response in DFU. Overall, our results demonstrate that LLLT improves blood flow and autonomic nervous system regulation in DFU and the importance of light intensity in therapeutic protocols.
Since 2005, our research team has been developing automated techniques for carotid artery (CA) wall segmentation and
intima-media thickness (IMT) measurement. We developed a snake-based technique (which we named CULEX1,2), a
method based on an integrated approach of feature extraction, fitting, and classification (which we named CALEX3), and
a watershed transform based algorithm4. Each of the previous methods substantially consisted in two distinct stages: Stage-I - Automatic carotid artery detection. In this step, intelligent procedures were adopted to automatically
locate the CA in the image frame. Stage-II - CA wall segmentation and IMT measurement. In this second step, the CA distal (or far) wall is segmented in order to trace the lumen-intima (LI) and media-adventitia (MA) boundaries. The distance between
the LI/MA borders is the IMT estimation.
The aim of this paper is the description of a novel and completely automated technique for carotid artery segmentation
and IMT measurement based on an innovative multi-resolution approach.
High-resolution ultrasonography (HRUS) has potentialities in differential diagnosis between malignant and benign
thyroid lesions, but interpretative pitfalls remain and accuracy is still poor.
We developed an image processing technique for characterizing the intra-nodular vascularization of thyroid lesions.
Twenty nodules (ten malignant) were analyzed by 3-D contrast-enhanced ultrasound imaging.
The 3-D volumes were preprocessed and skeletonized. Seven vascular parameters were computed on the skeletons:
number of vascular trees (NT); vascular density (VD); number of branching nodes (or branching points) (NB); mean
vessel radius (MR); 2-D (DM) and 3-D (SOAM) tortuosity; and inflection count metric (ICM). Results showed that the
malignant nodules had higher values of NT (83.1 vs. 18.1), VD (00.4 vs. 0.01), NB (1453 vs. 552), DM (51 vs. 18), ICM
(19.9 vs. 8.7), and SOAM (26 vs. 11).
Quantification of nodular vascularization based on 3-D contrast-enhanced ultrasound and skeletonization could help
differential diagnosis of thyroid lesions.
The carotid intima-media thickness (IMT) is the most used marker for the progression of atherosclerosis and onset of the
cardiovascular diseases. Computer-aided measurements improve accuracy, but usually require user interaction.
In this paper we characterized a new and completely automated technique for carotid segmentation and IMT
measurement based on the merits of two previously developed techniques. We used an integrated approach of intelligent
image feature extraction and line fitting for automatically locating the carotid artery in the image frame, followed by
wall interfaces extraction based on Gaussian edge operator. We called our system - CARES.
We validated the CARES on a multi-institutional database of 300 carotid ultrasound images. IMT measurement bias
was 0.032 ± 0.141 mm, better than other automated techniques and comparable to that of user-driven methodologies.
Our novel approach of CARES processed 96% of the images leading to the figure of merit to be 95.7%. CARES ensured
complete automation and high accuracy in IMT measurement; hence it could be a suitable clinical tool for processing of
large datasets in multicenter studies involving atherosclerosis.pre-
Most of the algorithms for the common carotid artery (CCA) segmentation require human interaction. The aim of this
study is to show a novel accurate algorithm for the computer-based automated tracing of CCA in longitudinal B-Mode
ultrasound images.
One hundred ultrasound B-Mode longitudinal images of the CCA were processed to delineate the region of interest
containing the artery. The algorithm is based on geometric feature extraction, line fitting, and classification. Output of
the algorithm is the tracings of the near and far adventitia layers. Performance of the algorithm was validated against
human tracings (ground truth) and benchmarked with a previously developed automated technique.
Ninety-eight images were correctly processed, resulting in an overall system error (with respect to ground truth) equal to
0.18 ± 0.17 mm (near adventitia) and 0.17 ± 0.24 mm (far adventitia). In far adventitia detection, our novel technique
outperformed the current standard method, which showed overall system errors equal to 0.07 ± 0.07 mm and 0.49 ± 0.27
mm for near and far adventitia, respectively. We also showed that our new technique is quite insensitive to noise and has
performance independent on the subset of images used for training the classifiers.
Superior architecture of this methodology could constitute a general basis for the development of completely automatic
CCA segmentation strategies.
The identification of structural systems using time-frequency analysis has been recently proposed to detect possible damages under normal serviceability conditions. Accelerometric signals are recorded at different points of the structure. They consist of the superposition of vibration modes (that identify the system) and residual components. Each vibration mode is represented by its frequency, its amplitude at different points of the structure, and the damping factor that determines its amplitude modulation. We have lately proposed a Cohen Class cross-time-frequency based technique to identify the vibration modes. In this paper we show how the technique may be developed in a fully automatic procedure and we discuss how the use of adaptive kernels may improve the reliability of the identification. The automatic procedure is based on two properties that characterize the vibration modes: (1) the ratio between the amplitude of the same modal component at different points of the structure is constant; and (2) the phase difference between the signals corresponding to the same modal component at different point of the structure is constant. These properties enable the vibration modes and residual components to be discriminated.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.