Background. - Underprivileged immigrants from endemic areas cumulate risk factors for infections by HIV-AIDS and hepatitis B and C. Free primary care consultations are available to them in the four health centers of the city of Paris. The objective of our study was to identify socio-demographic and medical factors related to the lack of screening proposition for HIV-AIDS and hepatitis B and C to new immigrant patients in these centers in 2003.Methods. - For each disease, the absence of screening proposition was analyzed according to geographical origin, length of stay in France, type of accommodation, type of health insurance and symptom motivating the encounter in logistic mixed models adjusted on sex and age.Results. - About 500 patients were included in the analysis. Three-quarters of them were male and from Sub-Saharan Africa. They were 36 years old on average. Half of them lived in shelters for homeless or immigrants. Their median stay lasted two years. They rarely came for screening (1%), sometimes for asthenia (6%) and two-thirds of them for uro-genito-digestive signs. The results were similar for the three screenings. The lack of screening proposition was about 45% and varied significantly between physicians. Factors significantly associated with the lack of screening proposition were: coming from non-Sub-Saharan Africa (especially from North Africa and Middle East; OR = 1.7 to 3.6) and having a health insurance (OR = 2.4 to 2.6) regardless of the disease; being a female (OR = 2.0 to 2.3) in the case of hepatitis; and having a length of stay in France greater than or equal to five years (OR = 1.9) for hepatitis B.Conclusions. - Our results should encourage practitioners to provide more screening to underprivileged immigrants and draws attention to immigrants from non-Sub-Saharan origin and those with health insurance. Factors that might explain doctor and gender-related variability observed in hepatitis are highlighted. (C) 2011 Elsevier Masson SAS. All rights reserved.

Our results should encourage practitioners to provide more screening to underprivileged immigrants and draws attention to immigrants from non-Sub-Saharan origin and those with health insurance. Factors that might explain doctor and gender-related variability observed in hepatitis are highlighted.

The study of small and/or slow reactivations of landslides requires describing their displacements over decades, which may be done with accurate multi-temporal digital terrain models (DTMs). We applied aerial stereophotogrammetry to build the historical topographies of old deep-seated landslides close to Oudenaarde in the Flemish Ardennes (West Belgium) at different dates. Three precise aerotriangulations (1996,1973,1952) were carried out. After capturing the ground data manually from the stereomodels, 2 m-resolution DTMs were interpolated by kriging, with a final accuracy ranging between similar to 45 cm and similar to 65 cm. Another DTM was interpolated with an accuracy of similar to 30 cm from airborne LIDAR data acquired in 2002. Differential DTMs were produced to identify vertical and horizontal ground displacements over the 1952-2002 period. We describe here the kinematics of a particularly active landslide with a well-documented recent activity. Until the first half of the 90 s, little activity of the landslide was detected. In February 1995 a reactivation event caused vertical displacements of up to -7 m along the main scarp and up to +4 m in the accumulation zone. Horizontal movements of 4 to 10 m are also inferred. These topographic changes correspond to reactivated slip along the rotational basal shear surface. In the same time, the main scarp retreated by up to 20 m. The reactivation, favoured by several anthropogenic factors (e.g. loading, impeded drainage), was triggered by intense rainfall. Between 1996 and 2002, the observed displacements correspond to limited scarp retreat (<= 4 m) and compaction of the slipped mass, partly enhanced by artificial drainage. (c) 2008 Elsevier B.V. All rights reserved.

[1] Using yearly leveling surveys performed from 1993 to 1998 in the Ardenne, as well as historic leveling results from 1948 and 1974, Demoulin and Collignon [2000] (hereinafter referred to as D&C) observed that the total vertical ground movement (less than 1 cm) over a 20-to 30-year period is barely higher than the yearly displacements measured to a few millimeters per year. They eliminated various sources of errors from their measurements and subsequently interpreted the resulting short-term oscillating crustal displacements as true tectonic motions within the upper crust accompanying the long-term deformation of uplifting areas in intraplate settings. Their justification is that the leveling discontinuities coincide with faults they assume to be active. In their conclusions, D&C suggested that such motions and their variations could be used to monitor potential seismogenic faults and even to foretell earthquakes. [2] Unfortunately, it is our opinion that the data analysis of D&C is inadequate and does not support the subsequent interpretation. Specifically, we interpret the apparent oscillations in the leveling data as the expected expression of the noisy character of leveling differences with amplitudes ranges of 0.5 to 1 cm. If this premise is accepted then there is also no evidence to suggest that the boundary faults in the study area are active. We thus feel that the conclusions of D&C are premature until their observations can be corroborated by improved observations or by other independent measurements. In this note, we present a careful statistical analysis to support our position. [3] The quality of the experimental work of D&C is not the concern of this discussion. D&C conducted their yearly leveling surveys following the requirement of Belgian first-order leveling and took sufficient precautions to minimize potential leveling errors. Our concern instead focuses on the minimum level of vertical deformation detectable using differences of high-precision leveling data. [7] In their paper, D&C considered confidence limits of ±1σ from the mean which corresponds only to a confidence interval of 68%. The commonly employed confidence interval is the 95% interval, which is approximately ±2σ from the mean [Bannister et al., 1998]. This error bar takes into account random errors, but there are also many different sources of possible error in the leveling as D&C themselves indicate (tidal correction, differential refraction, settlement), which, unfortunately, are then not considered in their analysis. Even if these errors seem small when considered individually, they could be significant at the millimeter scale when combined. [8] We reanalyzed the D&C data (their Figures 5, 6, and 7) in order to test the reliability of the conclusions with different confidence intervals. As we did not have access to the measured values of the different leveling surveys conducted by D&C, we scanned and digitized the yearly differences of their measurements from Figures 5 and 6. The resulting values do not exactly correspond to the original data set, but they are accurate enough to support our discussion. [9] In Tables 1 and 2, we present the relative height difference, ΔHi, and ΔHi/σi for each section in routes A and B, respectively. [10] In route A, of the 164 sections measured, ΔHi is less than 1σi for 93 sections, less than 2σi for 130 sections, and less than 3σi for 150 sections. If we adopt the 95% confidence limits, a significant height difference of the bench marks is then measured for only 34 sections (20.7% of the whole set) (see Figure 1). In their paper, D&C do state that the ΔHi along route A do not significantly exceed the noise. However, they (p. 697) go on to say that “valuable tilt information can be obtained as very clear trends of motion develop along some parts of the leveling route”. [11] Considering the profile of their Figure 5, yearly profile 1994–1995, they determined a trend from 20 leveling sections. In contrast, ΔHi/σi values (see Table 1) are >2 for only six sections and are >3 for only three sections (respective values of 3.2, 3.0, and 5.2). If we focus on the detailed part of the segment presented on the upper part of D&C's Figure 4 (see Figure 2), we are able to observe the lack of reliability of the apparent trend in the difference of these two leveling surveys. The diagram only shows 4 of the 20 sections with their five bench marks. In this case the error bar of each section represents a 2σi error. Two of these sections are among the three sections for which ΔHi/σi is greater than 3. Given that this diagram incorporates the most optimal data supporting the tilt statement, it is misleading for D&C to then state that the indicated ΔHi are significant and thus that the apparent trend on the cumulative ΔHi along the 20 sections (bottom of their Figure 4) is an actual real ground movement. [12] The reversal images obtained for this segment of route A (D&C's Figure 5) during the two leveling periods 1993–1994 and 1994–1995 and the absence of such a significant trend in all the other sets of leveling differences (1974–1993, 1993–1998, 1995–1996, 1996–1997, and 1997–1998) may suggest the existence of a systematic error during the 1994 leveling survey. D&C noted that they did not detect any height-dependent systematic errors in their leveling. However, this observation does not in itself preclude the existence of other systematic errors. Our analysis clearly demonstrates the difficulty in extracting an unambiguous geophysical signal from the yearly differences along route A. [13] In route B, of the 63 measured sections, 36 (57%) show significant height differences (ΔHi/σi is greater than 2), which could correspond to true displacements of the bench marks. On the other hand, some bench marks exhibit an apparent isolated and significant vertical displacement. This is the case for bench mark 5 for the 1993–1994 leveling difference and for bench mark 9 for the 1996–1997 difference. Such outliers should be ignored in the interpretation. However, even if these measurements are discarded from the analysis, some determined motions are significant. [14] The most spectacular movements occurred between 1993 and 1994. In their Table 2, D&C provide little information regarding the 1993 survey. However, previously, Demoulin and Collignon [1997] indicated that the loop closures never exceed 6.4 mm for the 24.1 km loop and 6.2 mm for the 17.8 km loop. These values correspond to standard deviations of 1.30 mm/km0.5 and 1.47 mm/km0.5, respectively. The D&C publications are not very clear on this point, but these values are most likely the standard deviations of the 1993 survey. They are significantly higher than those, 0.30 to 0.49 mm/km0.5 (Table 2 of D&C), obtained for the section closure. This indicates [Vanicek and Krakiwsky, 1986] that the height differences are statistically dependant and that there exists a cumulative error along the routes, which is independent of the section closures. [15] Using the loop standard deviation values of 1.47 mm/km0.5 for the 1993 survey and 0.80 mm/km0.5 for the 1994 survey (corrected values from D&C's Table 2), the height change standard deviation along loop 1 for this yearly difference is 1.67 mm/km0.5. Using this value, in Figure 3 we show the corresponding 1σ, 2σ, and 3σ parabolas. From Figure 3, it appears that the observed relative vertical movements are not as significant as the section error bars would indicate (Figure 3). [16] It is interesting to note that D&C considered that NIG-IGN surveys, conducted in 1948 and 1974, were of lower accuracy (p. 700), i.e., “some centimeter-level displayed by these profiles may not be considered significant.” Their loop misclosures indicate that their surveys are neither better nor worse than those of the NIG-IGN. [17] Figure 3 of D&C shows the accumulation of the section closures for loop 1 of the 1994 survey. It is interesting to note that the clear trend from 10 to 17.8 km corresponds to the Vesdre valley bottom (Quaternary deposits) section of route B and where the movements are significant. The value of the trend is ∼0.35 mm/km0.5. Also most of the total loop closure appeared along this section of the route. This result indicates that the measurements in this region are less reliable than in the first 10 km of the route. For the data presented in Figure 3, it is incorrect to say (in contrast to the statement made by D&C) that there is no spatial correlation between the amplitude of the trends and the local geology. [18] Route B follows the valley bottom of the Vesdre river where the sediments are recent (Pleistocene) alluvial deposits whereas only a small part of route A follows the Vesdre valley, along three sections that are common to route B. A majority of the sections for which ΔHi/σi is >3 (σi derived from the section closures) are located on these Quaternary alluvial sediments. Route A is also parallel to route B (with a maximum separation of 3 km), and both cross faults assumed to be active by D&C. It is noteworthy to remark that there is no significant vertical movement on route A, which is situated on the Paleozoic basement. This discrepancy further supports our suggestion that there is a correlation between bench mark movement and the material in which they are built rather than a correlation with tectonics. D&C discussed in detail leveling errors and nontectonic causes generating bench mark instabilities. Unfortunately, they did very little to assess the impact of such errors on their data. They did, however, conduct an experiment in which they performed weekly releveling of a part of a section over a time span of 3 months. They observed a random motion possibly associated with monument instability of a few tenths of a millimeter and considered this negligible for their surveys. [19] To be able to interpret their data in terms of crustal deformations, it is necessary that one studies the bench mark vertical motion in more detail, not only along route B, but also along route A. Indeed, they affirmed (p. 701) that “a large majority of these buildings have vaults carved in fresh Paleozoïc rocks to a depth of 2.5 m.” However, there are no data to confirm that the bedrock is located so near to the surface. On the contrary, geophysical prospecting performed in the same area (Hautes-Fagnes) showed that bedrock (Vs > 800 m/s) would be located at a depth between 4 to 11 m [Horrent, 1992]. In addition, the bedrock, which is mostly pelitic, is covered with a weathering layer of clay, which is very sensitive to water content variations. Wyatt [1989] has demonstrated that quantifying monument stability is essential to performing detailed studies of crustal deformation. [20] From our analysis of the measurements along route B, we conclude that the observed vertical movement, significant in terms of the sections closures, are, in fact, not as significant in terms of the loop closures. We conclude that the observed motions are most likely due to the local geological conditions (recent alluvial deposits) rather than to true crustal movements. The data suggest that the influence of these geological conditions is twofold. First, it seems to decrease the quality of the measurements along the sections. Second, it could induce true movement of the bench mark themselves. We would suggest a more thorough analysis of the data in light of these considerations before claiming that the observed movements are significant and represent tectonic deformation. [21] To support their interpretation that their leveling results can be explained as tectonic motions of crustal blocks, D&C assume that the faults marking the limits of the blocks are active. This assumption is fundamental to their conclusions, because it is the only other source of evidence that is used to corroborate their hypothesis regarding the tectonic movement (“reversible fault creep”) observed on the leveling data. (see, for example, p 702: “Moreover, the leveling discontinuities along route B strikingly coincide with active faults evidenced by deformed upper Quaternary terraces…” and on p. 694, “The analysis of Quaternary terraces has also demonstrated vertical displacements of up to 7 m on several faults of … in the last 100 kyr”). [23] Hence we do not know if these faults are active or not, and we feel that D&C should not have presented this as an accepted fact. [24] Our rudimentary analysis of the yearly leveling surveys conducted by D&C in the Belgian Ardenne suggests that the measured relative height differences (0.5 to 1 cm) are not significant enough to infer crustal deformations and those that are significant most likely result from non-tectonic causes. It is our point of view that leveling surveys dedicated to study vertical tectonic movements in intraplate areas should be accompanied at least by gravimetric measurements as a complimentary technique capable of observing very small crustal motions [van Dam et al., 2000]. The present-day resolution of these gravimeters is 1 μGal ∼3 mm of uplift [Francis et al., 1998]. And so gravimetry could be very complimentary to the technique of the topographic leveling where small vertical crustal motions are suspected. Further, any analysis of leveling data with crustal deformation of less than 1 cm over baseline length of order 10 km or less should be accompanied by an analysis of monument stability [see Wyatt, 1989]. [25] We also feel that the assumption by D&C that the boundary faults in the region are active has not been substantiated and has been in fact refuted. Given these factors, we feel that the hypothesis of D&C of the existence of the “reversible fault creep” remains to be demonstrated in the Ardenne.

Yearly high‐precision leveling surveys have been performed from 1993 to 1998 along a 48‐km‐long local network centered on Gileppe Lake, at the foot of the Ardenne massif, Belgium. Institut Géographique Militaire and Institut Géographique National (IGM‐IGN) levelings of 1948 and 1974 yielded additional information. The data were corrected for collimation and thermal errors. Other sources of error (Earth tides, atmospheric refraction, rod miscalibration) were estimated, and corrections were made when necessary. Random errors of individual surveys range from 0.15 to 0.74 mm/√km. The recorded elevation changes indicate the existence of tilted segments some 5 to 10 km in length, with yearly tilt values amounting up to 2 μrads. Fault movements of a few mm/yr are also demonstrated. The segment limits are spatially stable from one year to the next, but very frequent sense reversals of the vertical motions are observed. This explains why the total movement over a 20‐ to 30‐year period is barely higher than the yearly displacements. Various nontectonic causes of vertical ground movement (e.g., monument instability, varying water loading of the lake, sediment compaction, etc.) have been considered, yielding little satisfactory explanation of the patterns of observed elevation changes. We propose that the recorded motions, which we call reversible fault creep and reversible tilt, respectively, are a true tectonic signal, reflecting the “tectonic noise” which accompanies the long‐term deformation of uplifting areas in intraplate settings. Such movements are characteristic of fault segments which may occasionally generate earthquakes.

Clostridium difficile is responsible for a syndrome ranging from mild or moderate diarrhea to severe pseudomembranous colitis. These pathologies occur most often after antibiotic treatment inducing a disruption of the intestinal barrier and allowing the emergence of C. difficile of endogenous or exogenous origin[1]. The treatment of these infections consists of stopping preliminary antibiotherapy and the introduction of specific antibiotherapy directed against C. difficile. The most frequently used therapeutic substances are glycopeptides (vancomycin, teicoplanin) and imidazoles (metronidazole). Other therapeutic agents have been used in combination, such as rifampin[2]. The antibiotic susceptibility of C. difficile has not been systematically determined[3]. In the current circumstances of evolution of resistance, especially to the glycopeptides, with the appearance of glycopeptide-resistant Enterococcus, and metronidazole-resistant C. difficile isolates, it seems interesting from an epidemiologic point of view to be able to evaluate antibiotic susceptibility of C. difficile[4–5]. In parallel, knowledge of the susceptibility profile may allow us to provide optimal therapy in difficult cases. The reference method for susceptibility testing of anaerobic bacteria is the agar dilution method as recommended by the National Committee for Clinical Laboratory Standards (NCCLS)[6]. This method is appropriate for reference laboratories; it is time-consuming and poorly adapted to the analysis of a few isolates. The E test has been evaluated for susceptibility testing of anaerobes, and it has been shown that the results obtained with this method correlate well with the reference method[7–11]. Few isolates of C. difficile were included in these studies. The purpose of this study was to compare the E test with the reference agar dilution method to determine the susceptibility of C. difficile to antibiotics used in therapy, in order to have a rapid, reliable and easy-to-use method. Fifty isolates of C. difficile were studied. Bacteroides fragilis ATCC 25285, Bacteroides thetaiotaomicron ATCC 29741 and Clostridium perfringens ATCC 13124 were used as controls. The following antimicrobial agents were tested: vancomycin (Lilly, Saint Cloud, France), teicoplanin, rifampin (Hoechst Marion Roussel, Paris, France), and metronidazole (Specia, Montrouge, France). MICs were determined by the reference agar dilution method, according to recommendations of the NCCLS, on Wilkins–Chalgren agar (Oxoid, Dardilly, France) supplemented with 5% horse blood and containing serial dilutions of the antimicrobial agents ranging from 0.062 to 32 mg/L. The inoculum size was approximately 105 CFU. Incubation was performed in an anaerobic chamber at 37 °C for 48 h. The breakpoints used for interpretation as susceptible were as follows: vancomycin ≤4 mg/L, teicoplanin ≤4 mg/L, rifampin ≤4 mg/L, and metronidazole ≤4 mg/L, according to the Comité de l'Antibiogramme de la Société Française de Microbiologie (CA-SFM). The E test (AB Biodisk, Solna, Sweden) was performed according to the recommendations of the manufacturer on Brucella agar (Fisher Scientific, Elancourt, France) supplemented with 5% horse blood. The inoculation was performed by swabbing the plates with an inoculum equivalent to 1 McFarland turbidity standard. The plates were dried for 20 min in an anaerobic atmosphere before applying E test strips. Incubation was performed for 48 h in an anaerobic chamber. The MIC was read where the inhibition ellipse intersects the scale. The E test MIC values that were between standard two-fold dilutions were rounded to the next higher value for comparison with agar dilution MIC values. A categorical discrepancy was defined as a ≥2-dilution difference that resulted in a change of interpretation from susceptible to resistant or vice versa. The results are summarized in Tables 1 and 2. MIC values obtained by the reference method for vancomycin ranged from 0.5 to 2 mg/L; those obtained by the E test ranged from 0.25 to 1 mg/L. Forty-nine of 50 (98%) of the E test MICs were within one dilution of the agar dilution MICs. All isolates were classified as susceptible by the two methods. MIC values obtained by the reference method for teicoplanin ranged from 0.125 to 1 mg/L; those obtained by the E test ranged from 0.09 to 0.5 mg/L. E test MICs were within one doubling dilution of agar dilution MICs in 76% of cases. As a whole, E test MICs were lower; however, all were within two dilutions of agar dilution MICs. MICs obtained by the two methods allowed the classification of all isolates in the susceptible category. For rifampin, there was no discrepancy in the classification into susceptible or resistant categories. Six of 50 (12%) isolates were resistant to rifampin, with MICs >32 mg/L by the two methods. Similarly, MIC values obtained by the two methods were concordant for susceptible isolates, MICs ranging from 0.5 to >0.06 mg/L. MIC values obtained by the reference method for metronidazole ranged from 0.125 to 8 mg/L. Those obtained by the E test ranged from 0.09 to 1.5 mg/L. The E test MICs were within one dilution of the agar dilution MICs in 80% of cases and within two dilutions in 90% of cases. For four isolates, MICs obtained by the reference method were 6 and 8 mg/L, whereas the E test MICs ranged from 0.5 to 1.5 mg/L, leading to discrepancies in categorization. Moreover, a lack of reproducibility of the E test was noted for the MICs clustering around the breakpoint. Preliminary results have shown that the E test performed on Wilkins–Chalgren agar gave values lower than those obtained on Brucella agar[12]. The reference agar dilution method is reserved for reference laboratories that test many isolates. Indeed, it is relatively difficult to carry out for a small number of isolates. The E test, which is easy to perform, appears to be a suitable method to test susceptibilities of anaerobic bacteria[7–11]. For C. difficile, the E test, as compared to the reference method, yielded good results and was easy to read. For vancomycin, teicoplanin and rifampin, no discrepancies in categorization were observed. However, as previously shown[8,9–11], the E test values were slightly lower, especially for teicoplanin, despite the use of a 1 McFarland inoculum and an incubation period of 48 h. As for metronidazole, for the low MICs, the E test results correlated well with those obtained by the reference method. For MICs clustering around the breakpoint, the E test MICs were lower (more than two two-fold dilution) and led to discrepancies in categorization. These results (E test and reference method) have been verified under strictly controlled conditions of anaerobiosis, since strict anaerobiosis is critical to accurate metronidazole susceptibility testing[13–14]. Variability of the results for metronidazole concerning anaerobic bacteria when using the E test has been described: several factors could intervene, such as test conditions, and type and origin of the medium[9,12–14]. In conclusion, the E test appears to be a reliable and easy-to-perform method for C. difficile antibiotic susceptibility testing in clinical laboratories. For metronidazole, the test conditions have to be well controlled and the results have to be interpreted with caution.

PURPOSE The primary objective of this study is to perform a blinded evaluation of a group of retrospective image registration techniques using as a gold standard a prospective, marker-based registration method. To ensure blindedness, all retrospective registrations were performed by participants who had no knowledge of the gold standard results until after their results had been submitted. A secondary goal of the project is to evaluate the importance of correcting geometrical distortion in MR images by comparing the retrospective registration error in the rectified images, i.e., those that have had the distortion correction applied, with that of the same images before rectification. METHOD Image volumes of three modalities (CT, MR, and PET) were obtained from patients undergoing neurosurgery at Vanderbilt University Medical Center on whom bone-implanted fiducial markers were mounted. These volumes had all traces of the markers removed and were provided via the Internet to project collaborators outside Vanderbilt, who then performed retrospective registrations on the volumes, calculating transformations from CT to MR and/ or from PET to MR. These investigators communicated their transformations again via the Internet to Vanderbilt, where the accuracy of each registration was evaluated. In this evaluation, the accuracy is measured at multiple volumes of interest (VOIs), i.e., areas in the brain that would commonly be areas of neurological interest. A VOI is defined in the MR image and its centroid c is determined. Then, the prospective registration is used to obtain the corresponding point c' in CT or PET. To this point, the retrospective registration is then applied, producing c" in MR. Statistics are gathered on the target registration error (TRE), which is the distance between the original point c and its corresponding point c". RESULTS This article presents statistics on the TRE calculated for each registration technique in this study and provides a brief description of each technique and an estimate of both preparation and execution time needed to perform the registration. CONCLUSION Our results indicate that retrospective techniques have the potential to produce satisfactory results much of the time, but that visual inspection is necessary to guard against large errors.

All retrospective image registration methods have attached to them some intrinsic estimate of registration error. However, this estimate of accuracy may not always be a good indicator of the distance between actual and estimated positions of targets within the cranial cavity. This paper describes a project whose principal goal is to use a prospective method based on fiducial markers as a 'gold standard' to perform an objective, blinded evaluation of the accuracy of several retrospective image-to-image registration techniques. Image volumes of three modalities -- CT, MR, and PET -- were taken of patients undergoing neurosurgery at Vanderbilt University Medical Center. These volumes had all traces of the fiducial markers removed, and were provided to project collaborators outside Vanderbilt, who then performed retrospective registrations on the volumes, calculating transformations from CT to MR and/or from PET to MR, and communicated their transformations to Vanderbilt where the accuracy of each registration was evaluated. In this evaluation the accuracy is measured at multiple 'regions of interest,' i.e. areas in the brain which would commonly be areas of neurological interest. A region is defined in the MR image and its centroid C is determined. Then the prospective registration is used to obtain the corresponding point C' in CT or PET. To this point the retrospective registration is then applied, producing C' in MR. Statistics are gathered on the target registration error (TRE), which is the disparity between the original point C and its corresponding point C'. A second goal of the project is to evaluate the importance of correcting geometrical distortion in MR images, by comparing the retrospective TRE in the rectified images, i.e., those which have had the distortion correction applied, with that of the same images before rectification. This paper presents preliminary results of this study along with a brief description of each registration technique and an estimate of both preparation and execution time needed to perform the registration .

A new approach to the problem of multimodality medical image registration is proposed, using a basic concept from information theory, mutual information (MI), or relative entropy, as a new matching criterion. The method presented in this paper applies MI to measure the statistical dependence or information redundancy between the image intensities of corresponding voxels in both images, which is assumed to be maximal if the images are geometrically aligned. Maximization of MI is a very general and powerful criterion, because no assumptions are made regarding the nature of this dependence and no limiting constraints are imposed on the image content of the modalities involved. The accuracy of the MI criterion is validated for rigid body registration of computed tomography (CT), magnetic resonance (MR), and photon emission tomography (PET) images by comparison with the stereotactic registration solution, while robustness is evaluated with respect to implementation issues, such as interpolation and optimization, and image content, including partial overlap and image degradation. Our results demonstrate that subvoxel accuracy with respect to the stereotactic reference solution can be achieved completely automatically and without any prior segmentation, feature extraction, or other preprocessing steps which makes this method very well suited for clinical applications.

C. difficile est maintenant largement reconnu comme le principal agent de diarrhées nosocomiales ou associées à l'antibiotiques, en particulier chez les malades à risque qu'ils soint âgés ou que leurs défenses immunitaires soient amoindries.

Septic osteomyelitis of the hip in a previously healthy child is described. A weakly toxigenicCorynebacterium diphtheriae strain was isolated from the bone aspirate. The results of the treatment were rapidly satisfactory, after surgical drainage and antibiotic therapy with pristinamycin.

An algorithm is presented for the geometric registration of 3D multi-modality medical images using a weighted combination of point landmarks and anatomical surfaces. The algorithm includes a new fully automatic outlier treatment method.

Multimodal fuzzy voxel labeling is presented as the basis for a new image registration criterion. The corresponding registration system's architecture performs an iterative calculation of the labeling and the registration process simultaneously, while most other registration systems perform segmentation and iterative estimation of registration parameters sequentially. It will be argued that its application leads to more automated and more accurate registration solutions than does e.g. the use of typical surface based registration systems. In order to support the arguments raised we have performed a case study using both 2D MR software phantom image, and 2D and 3D MR/CT image data. In this case study we looked at the behaviour of maximum likelihood voxel labeling as the simplest instantiation of a fuzzy voxel labeling algorithm. However, the architecture is open to integration of more general multimodal fuzzy voxel labeling algorithms.

This paper gives an overview of existing surface based registration methods. Properties of commonly used classification schemes for registration algorithms are divided into two groups: external attributes are used to distinguish surface based from non-surface based registration algorithms, and internal attributes serve to distinguish surface based registration algorithms from each other. An overall comparison is performed based on a quality constrained cost analysis and the limits of the algorithms' applicability for neurosurgical therapy planning systems are investigated.

In this presentation a new search method is proposed to improve the speed and accuracy of surface based registration algorithms. Furthermore, a parallel point projection based, multicomponent distance evaluation method is presented. This method offers an elegant solution to the problem of partially overlapping data sets. An adaptive outlier treatment method is also presented. Combination of all these new techniques results in a faster surface based algorithm with better accuracy, but above all with better reliability then existing surface based 3D registration algorithms. In the context of the surface correspondence problem, surface based registration algorithms are compared to feature matching methods.

The incidence rate of juvenile Type 1 (insulin-dependent) diabetes in France was reported as the lowest in Europe 13 years ago, but during the recent years increasing rates have been observed in different European countries. A prospective programme has been designed to study the incidence rate of Type 1 diabetes in patients up to 20 years of age in four regions located in the north and south of France (population <20 years = 2.31 million inhabitants; 15% of the French population). All cases were independently identified by four specially trained research assistants through hospital admission files, paediatricians, diabetologists and general practitioners. A specific questionnaire was filled out for each newly diagnosed case. Degree of ascertainment was 96% with the data from Sécurité Sociale, the French National Health Insurance. In 1988, 166 cases of juvenile Type 1 diabetes were identified. The incidence rate was 7.17 cases per 105 children (95% confidence interval = 6.1–8.2/105). The values were not statistically different among the four regions. Age specific incidence rates were as follows: 0–4 years = 3.8; 5–9 years = 8.0; 10–14 years = 9.7 and 15–19 years = 7.3/105. Sex ratio was 1.2 (male/female). These data indicate that incidence of juvenile Type 1 diabetes in France was higher in 1988 than previously reported but remains lower than in Northern Europe. This is consistent with the concept of a north to south gradient of the disease.