Search PubMed⌕ Search

Biomedical subjects

L Garrido

Publications and source records attributed to L Garrido.

At least 37 records · Page 2Linked to original sources
PubMed1996

Analytical interpretation of feed-forward nets outputs after training.

The minimization quadratic error criterion which gives rise to the back-propagation algorithm is studied using functional analysis techniques. With them, we recover easily the well-known statistical result which states that the searched global minimum is a function which assigns, to each input pattern, the expected value of its corresponding output patterns. Its application to classification tasks shows that only certain output class representations can be used to obtain the optimal Bayesian decision rule. Finally, our method permits the study of other error criterions, finding out, for instance, that absolute value errors lead to medians instead of mean values.

Algorithms↗
PubMed1996

Silicone gel-filled breast implants in women: findings at H-1 MR spectroscopy.

PURPOSE: To evaluate, at hydrogen-1 magnetic resonance (MR) spectroscopy, the effect of implantation time, implant status, and implant removal on the amount of silicone in the liver in women with silicone gel-filled breast prostheses. MATERIALS AND METHODS: The study population included 55 women (39 patients with silicone gel-filled prostheses and seven from whom implants had been removed, and nine control subjects [eight with no implant and one with saline-filled implants]). Stimulated-echo acquisition mode, or STEAM, H-1 MR spectroscopy was performed to determine the concentration of silicone in the liver. Implant status at the time of spectroscopy was diagnosed at MR imaging. RESULTS: Twenty of 39 (51%) women with implants had ruptured prostheses. Resonances associated with the presence of silicone and partially hydrolyzed silicone (0.3 to -0.8 ppm with respect to water at 4.7 ppm) and other resonances that are not yet assigned (-2 to -5 ppm) were detected in 27 (69%) of the 39 women (17 with ruptured implants). Relative signal intensities of the silicone species detected in the liver in these women were found to vary substantially and were not correlated with the status of the implants (P > .70). Silicone resonances were not detected in the livers in the nine control subjects. After implant removal, no resonances between 0.3 and -0.8 ppm were observed in six of seven women, but silicone-related peaks were still detectable in the region of -2 to -5 ppm. CONCLUSION: Proton MR spectra obtained in the liver of women with silicone gel-filled breast implants helped measure silicone exposure.

Adult↗
PubMed1995

Study of aging of silicone rubber biomaterials with NMR.

Multinuclear nuclear magnetic resonance (NMR) spectroscopy (29Si, 13C, 1H) is used to characterize the aging process of silicone rubber-based biomaterials in a rat model. 1H NMR relaxation measurements (spin-lattice, T1, and spin-spin, T2, relaxation times) were performed to better understand the molecular dynamics of polysiloxane chains in implants. After 1 year of implantation in animals, changes in the 1H T2 relaxation times and the NMR spectra were observed in polydimethylsiloxane, Silastic sheets and chin implants, while these measurements remain unchanged in finger joints. Very small amounts of fat were detected in all types of silicone rubber implants at the end of the implantation period. This work shows that free silicone migrates from the implants to adjacent tissues and distant sites, such as spleen or liver, and is chemically modified.

Animals↗
PubMed1995

Migration and accumulation of silicone in the liver of women with silicone gel-filled breast implants.

1H NMR localized spectroscopy (STEAM), combined with echocardiography (ECG), respiratory gating, and water and fat suppression, was used to quantify silicone concentrations in the liver of women with silicone gel-filled breast implants. Localized spectroscopy was performed on 15 patients with silicone gel-filled breast prostheses and on eight volunteers with no implants. The 1H spectra in the liver of patients showed silicone resonances from 0.3 to -0.8 ppm, attributable to protons in the methyl groups of silicone. The presence of silicone in the liver could first be detected 3-4 years after breast prostheses implantation. No correlation between silicone concentrations and implantation times was observed. However, our results indicated that silicone concentrations may reflect implant integrity: detectable silicone concentrations in the liver appeared to be higher when the implants were ruptured than when the implants appeared intact. Moreover, new resonances in the range of -2.6 to -4 ppm were observed in most patients after long-term implantation. As these species increase with implantation time, the new resonances may reflect chemically changed silicone (paramagnetically shifted silicon complexes bound to iron) accumulated over time. The sensitivity of 1H NMR localized spectroscopy is sufficient to detect silicon concentrations as low as 0.20 mM. Results from one patient whose implants had been removed 14 months prior to the NMR examination showed no detectable silicone in the liver, indicating that it may have been excreted via bile or degraded to silica and high coordinated silicon complexes. Quantitative 1H localized spectroscopy of the liver in women with silicone gel-filled breast implants may provide valuable information concerning silicone accumulation and degradation in vivo, as well as about the kinetics of its elimination from the body after implant removal.

Breast Implants↗
PubMed1995

Solid state phosphorus-31 magnetic resonance imaging of bone mineral.

Chemically selective solid state phosphorus-31 nuclear magnetic resonance (NMR) imaging of the mineral phase of bone and synthetic calcium phosphate models for bone mineral is demonstrated with microscopy-scale (about 5 mm field of view) apparatus at 6.0 T magnetic field strength. Pixel-by-pixel linear combination of image data from multiple radio frequency (RF) pulse sequences, chosen to develop contrast between chemical constituents of interest in the mineral, generates derived images showing the distribution of individual constituents. The technique combines the noninvasive character of magnetic resonance imaging (MRI) with the ability of solid state NMR spectroscopy to characterize subtle chemical variations in bone mineral, as well as to measure the amount of mineral. These methods are, in principle, extensible to larger dimensional scales suitable for live animal subjects or human limbs.

Animals↗
PubMed1995

Use of multilayer feedforward neural nets as a display method for multidimensional distributions.

We present a new method based on multilayer feedforward neural nets for displaying an n-dimensional distribution in a projected space of 1, 2 or 3 dimensions. A fully nonlinear net with several hidden layers is used. Efficient learning is achieved using multi-seed backpropagation. As a principal component analysis (PCA), the proposed method is useful for extracting information on the structure of the data set, but unlike the PCA, the transformation between the original distribution and the projected one is not restricted to be linear. Artificial examples and a real application are presented in order to show the reliability and potential of the method.

Artificial Intelligence↗
PubMed1995

Role of vasogenic edema and tissue cavitation in ischemic evolution on diffusion-weighted imaging: comparison with multiparameter MR and immunohistochemistry.

PURPOSE: To examine the mechanisms of further evolution that occurs from the early to late phase after initial changes in diffusion-weighted imaging after cerebral ischemia. METHODS: Sprague-Dawley rats were subjected to middle cerebral artery occlusion. Diffusion-, proton density-, T1- and T2-weighted imaging were performed on days 0, 2, and 6. Histologic examination (IgG, glial fibrillary acidic protein, and cresyl violet staining) was done after scanning. RESULTS: Apparent diffusion coefficients (ADCs) in the ischemic hemisphere were significantly decreased on day 0. Thereafter, ADCs increased over time and became significantly higher than the contralateral side by day 6. Changes in basal ganglia occurred more rapidly than in cortex. Proton density-, T1-, and T2-weighted scans showed maximal changes on day 2. From day 0 to day 2, there are significant correlations between changes in ADC and changes in T1-weighted signals and T2-weighted signals. Histologic exam showed early neuronal injury on day 0, intense gliotic activity and protein leakage associated with infarction and edema on day 2, and cavitation in severely infarcted areas on day 6. CONCLUSION: After initial reduction of ADC, the subsequent increase in ADC values on day 2 may be associated with vasogenic edema and cell lysis. Later elevations in ADC may be related to cavitation of infarcted tissue.

Animals↗
PubMed1994

Pharmacologic reversal of acute changes in diffusion-weighted magnetic resonance imaging in focal cerebral ischemia.

Recently, diffusion-weighted magnetic resonance imaging (DWI) has been shown to visualize acute ischemic lesions in the brain before changes are observable with conventional magnetic resonance imaging. However, the underlying mechanisms of these acute DWI changes are unclear and may include both reversible and irreversible damage. In this study, we demonstrate that acute DWI lesions may be reversed with MK801 therapy postischemia. Sprague-Dawley rats (n = 12) were subjected to middle cerebral artery occlusion and DWI scans were obtained beginning 60 min postocclusion. Distinct regions of hyperintensity were observed in the basal ganglia and cortex, corresponding with the expected distribution of ischemia in this model. After the first scan, animals were treated with MK801 (0.5 mg/kg i.v.) or normal saline and subsequently scanned again 30 and 60 min after treatment. In the control group, the area of hyperintense lesions continued to increase, by 55% in the cortex and 57% in the basal ganglia. MK801 therapy significantly (p < 0.01) reduced the area of damage by the third DWI scan at 60 min posttreatment (-50% cortex, -22% basal ganglia, -41% total hemisphere) compared to pretreatment scans. Tetrazolium (TTC) stains at 24 h confirmed that MK801 significantly reduced the volumes of infarction (p < 0.05). These results demonstrate that significant portions of the acute ischemic lesion on DWI are reversible with pharmacologic intervention.

Animals↗
PubMed1994

Anisotropy of water diffusion in the myocardium of the rat.

Pulsed field gradient nuclear magnetic resonance methods combined with nuclear magnetic resonance imaging were used to determine the water diffusion anisotropy in perfused rat hearts at 37 degrees C. It was found that the observed diffusion coefficient D(app) (apparent diffusion coefficient) depends on the orientation of the applied gradient g. When g is parallel to the epicardial surface, the observed diffusivity is D(app) parallel = 1.8 +/- 0.4 x 10(-9) m2.s-1, whereas when g is perpendicular to it, diffusivity is D(app) perpendicular = 2.5 +/- 0.5 x 10(-9) m2.s-1. To better characterize this directional dependence, images of the second-order diffusion tensor D of the myocardium were obtained. These data demonstrate several essential features of cardiac myoarchitecture, including the helicity of fiber orientation with respect to the ventricular axis and the variation of fiber pitch angle with transmural depth. Diffusion anisotropy may be quantified in a coordinate-independent manner by the eigenvalues of the diffusion tensor. In the myocardial midwall, these eigenvalues were E1 = 3.29 +/- 0.57, E2 = 2.01 +/- 0.42, and E3 = 0.77 +/- 0.58 x 10(-9) m2.s-1 (mean +/- SD). These data suggest that myocardial water diffusion is essentially unrestricted parallel to the myofibers. They further show that failure to measure the complete diffusion tensor may lead to substantial underestimates of diffusion anisotropy in the myocardium.

Animals↗
PubMed1993

In vivo 1H chemical shift imaging of silicone implants.

In order to study the aging process (i.e., silicone migration, fat infiltration) of silicone (polydimethylsiloxane, PDMS) based biomaterials in living subjects by NMR imaging, a hybrid 1H selective excitation and saturation chemical shift imaging technique (IR/CHESS-CSSE) has been developed. This sequence allows selective mapping of the distribution of silicone protons in vivo, while suppressing the contributions of fat and water. Our results indicate that a combined inversion recovery and CHESS pulse, followed by a spoiler gradient, must be applied to suppress all contributions of fat protons to the NMR signal. The sensitivity of our experiments allows the detection of a chemically unchanged silicone concentration of 5% in a voxel of 0.9 mm3 at a signal/noise ratio of 2.

Adipose Tissue↗
PubMed1993

In vivo degradation of silicones.

29Si nuclear magnetic resonance (NMR) spectroscopy is applied to study the degradation of polysiloxanes (silicones) in vivo. Our results with animal models show that silicone migrates from the implant to the liver (29Si resonance at -20 ppm) and new silicon containing compounds form after the silicones are introduced into the rats. The new 29Si resonances in the chemical shift range of -40 to -85 ppm are related to hydrolyzed silicone, those at -90 to -115 ppm are indicative of the presence of silica (SiO2), and the peaks observed at -120 to -150 are related to high coordinated silicon complexes. These resonances are not present in the 29Si spectra of the silicones before implantation. Our findings demonstrate that silicones are not metabolically inert.

Animals↗
PubMed1993

In vivo localized proton NMR spectroscopy of silicone.

1H NMR localized spectroscopy (STEAM) can assess unambiguously the presence of free chemically unchanged silicone in animal tissue after injection of silicone oil. Although the signal-to-noise ratio obtained in 1H imaging is sufficient to detect the distribution of relatively large amounts of silicone in vivo, the specificity of silicone detection can be improved by using 1H localized spectroscopy techniques. The sensitivity of the STEAM experiments is sufficient to detect silicone at a concentration of 0.5% in a voxel of 27 mm3. Preliminary results from rats with silicone gel-filled implants show no detectable amounts of silicone in sites such as lymph nodes, the liver or the spleen, 3 or 6 months after implantation.

Animals↗
PubMed1993

Migration and biodegradation of free silicone from silicone gel-filled implants after long-term implantation.

In vivo 1H NMR chemical shift imaging (CSI), 1H NMR localized spectroscopy (STEAM) and multinuclear NMR spectroscopy (29Si, 13C, 1H) were used to characterize the aging process of silicone gel-filled implants in a rat model after long-term implantation. Although no significant changes could be observed in the implants or surrounding tissue by in vivo 1H chemical shift imaging, in vivo 1H localized spectroscopy of the livers from the longer term population revealed the presence of silicone. Ex vivo 29Si spectroscopy of the liver, spleen, and the capsule formed around the 9 and 12 month implants clearly demonstrated and confirmed for the first time that a significant amount of free silicone migrates from silicone gel-filled implants. Also, these results show that silicones are not metabolically inert, and their biodegradation in tissue and within the implant can be monitored after 9 and 12 months by changes in the 29Si chemical shifts seen in corresponding ex vivo spectra. The NMR findings are supported by those obtained by atomic absorption spectroscopy. Silicone aging changes not only the chemical composition of the gel, but also its proton T2 relaxation times, which increase with long implantation times. The three dimensional structure of the gel disintegrates (i.e., polymer chain rupture), increasing the molecular mobility of the polymer and, consequently, its protons T2 values. The relaxation data we obtained reflect this in vivo degradation, especially in the case of implant rupture. Additionally, small concentrations of fat in the silicone gel were found within the implants. The presence of these lipophilic substances also might increase the T2 values (plasticizer effect). These findings may assist in evaluating the implant integrity and disease symptoms related to their presence in humans.

Animals↗
PubMed1993

Echo-planar chemical shift imaging of silicone gel prostheses.

We have developed an echo-planar (EP) proton chemical shift imaging (CSI) MR technique that allow us to discriminate the polydimethylsiloxane (PDMS, silicone) proton MR signal from that of the fat and water protons found in tissues, in order to map the distribution of PDMA in humans who have silicone gel prostheses. Silicone gel-filled prosthetic implants induce histologic changes in the surrounding tissue which are attributed to the leakage of free PDMS from the prosthesis. The T2 relaxation measurements of three silicone gels show that there are two components in them, each with a different degree of molecular mobility. The presence of free silicone is confirmed by chloroform extraction, which removed 14-28% of the material. This free polymer present in the gel can pass through the intact or ruptured membrane of the implant into the surrounding tissue. Our preliminary imaging results indicate that EP-CSI MR might be useful as a diagnostic technique for implant malfunction.

Breast↗