Objective and subjective evaluation of adaptive speech enhancement methods for functional MRI

Purpose:

To recover speech corrupted by functional magnetic resonance imaging (fMRI) acoustic noise using two‐channel adaptive speech enhancement techniques.

Materials and Methods:

Speech corrupted by noise generated from a 3 T MRI scanner was recorded using diffuse‐field microphones and a data acquisition board. Multiband and subband adaptive speech enhancement methods are used to recover the speech signal from the recordings. Normalized least mean squares (NLMS) algorithm was used for updating the filter coefficients in each band.

Results:

The methods are successful in enhancing the speech quality. They are successful in improving the convergence rate of the adaptive filter. Multiband and subband methods have a similar performance in terms of noise reduction and in the subjective tests. The subband method introduces less speech distortion compared to the multiband method. The subband method requires a lower number of computations per sample.

Conclusion:

Adaptive speech enhancement techniques are effective in reducing fMRI background noise in the recordings. Based on the analysis, we conclude that subband‐based methods are more suited for enhancing speech corrupted by fMRI noise. J. Magn. Reson. Imaging 2010;31:46–55. © 2009 Wiley‐Liss, Inc.     

9The TANTALUSTM System for obesity: effect on gastric emptying of solids

Background: Gastric electrical stimulation (GES) is currently investigated for the treatment of obesity. The TANTALUS System delivers gastric contractility modulation (GCM) signals in synchrony with gastric slow waves, resulting in significant augmentation of gastric contractions during food intake. We hypothesized that such modulation of contractile activity may affect gastric emptying and plasma ghrelin levels. Aim: To test the effect of GCM of the gastric antrum on gastric emptying of solids and ghrelin levels. Methods: 12 obese subjects were implanted with 2 pairs of antral electrodes and an implantable pulse generator (IPG, TANTALUS ^TM) Gastric emptying test (GE) for solids was performed twice, on separate days, in each subject, starting few weeks after implantation: 1) control, before the start of stimulation, and 2) with stimulation, after device was turned on. Blood samples for ghrelin, were taken at baseline, and at 15, 30, 60 and 120 min after the test meal. Results as mean + SD, analysis by t‐test and p < 0.05. Results: 11 females, 1 male, age: 39.1 ± 8.9 years, BMI: 41.6 ± 3.4, 3 subjects with type 2 diabetes. One diabetic patient did not complete GE test because of technical issues. GCM significantly accelerated gastric emptying: retention at 2 hours 18.7 ± 12.2% vs. 31.9 ± 16.4%, stimulation vs. control respectively, p = 0.008. T ^1/2 78.3 ± 23.5 vs. 95 ± 31.7 min, stimulation vs. control respectively, p = 0.04. Mean results for gastric emptying were within normal at both baseline and stimulation. Meal ingestion induced only minimal, insignificant reduction in ghrelin levels. There was no significant difference in AUC of ghrelin between control and stimulation. Conclusions: After GCM stimulation, there is significant acceleration of gastric emptying of solids in obese patients, without affect on ghrelin levels. The obese subjects did not exhibit the significant, meal‐induced reduction in ghrelin.      

Evaluation of postpartum blood loss after misoprostol-induced labour

Objective  To objectively evaluate postpartum blood loss after successful misoprostol induction and compare it with blood loss after oxytocin induction of labour.
Design  Prospective randomised study.
Setting  Labour ward in university maternity hospital.
Population  A total of 150 women up to third parity with completed 40 weeks of singleton normal pregnancy, average size cephalic fetus.
Methods  Cases were randomised between oxytocin induction and misoprostol induction. Blood was collected in suction set and measured in the delivery room starting after delivery of the fetus and was evaluated by pad weighing in the following 6 hours. Pre- and postdelivery haematocrit were measured and difference between the two values was assessed and analysed.
Main outcome measures  Success of induction, induction delivery interval, postpartum blood loss, and difference between pre- and postdelivery haematocrit.
Results  Induced labour was significantly faster with misoprostol induction ( P < 0.001). Blood loss and haematocrit difference was significantly greater in the misoprostol group than in oxytocin group ( P < 0.02 and 0.001, respectively). Blood loss in both groups was significantly correlated with higher initial Bishop score ( P < 0.001 and 0.024, respectively) and short labour duration ( P < 0.0002 and 0.0001, respectively).
Conclusions  Misoprostol induction is associated with increased blood loss especially when used in women with high Bishop score; therefore, it is better reserved for cases requiring cervical ripening.     

Is MSH2 a breast cancer susceptibility gene?

Mutations in the DNA mismatch repair gene MSH2 lead to increased replication error and microsatellite instability and account for a substantial proportion of hereditary non-polyposis colorectal cancer (Lynch syndrome). A recent international collaborative genome-wide linkage scan (GWS) for breast cancer susceptibility loci found some evidence for there being a breast cancer susceptibility gene in a genomic region on chromosome 2p close to MSH2. We sought to investigate the possibility that mutations in MSH2 might explain the multiple cases of breast cancer in some families that were included in the international GWS. DNA samples from the affected probands of 59 multiple-case breast cancer families, many of whom gave LOD scores >0.5 in the MSH2 region, were screened for large genomic alterations in MSH2 via the Multiplex Ligation-dependant Probe Amplification (MLPA) assay and for coding region mutations via exonic sequencing. Several of the families also contained cases of colorectal cancer in addition to breast cancer and had been included in the GWS that had identified a positive LOD score on chromosome 2p. Using MLPA, c.1236C > T was identified in one proband but this variant was not predicted to create an alternate acceptor/donor site within exon 7 MSH2 using in silico analyses. A c.1734T > C was identified in a second proband via exonic sequencing but testing of the variant in other family members did not support segregation of this variant with disease. Extensive screening of 59 multiple-case breast cancer families did not identify any coding region mutations or larger genomic alterations in MSH2 that might implicate MSH2 as a breast cancer susceptibility gene.     

Regulation of intestinal apolipoprotein A-IV synthesis

Abstract Apolipoprotein (apo) A-IV is a protein synthesized, in humans, only by the small intestine. It has a molecular weight of 46 000 Da. This paper summarizes the evidence supporting its role as a satiety factor following the ingestion of fat. This function of apo A-IV is unique and not shared by other apolipoproteins, including apo A-I. The satiety effect of apo A-IV is centrally mediated. The mechanism of how apo A-IV inhibits food intake is not clear but it probably acts by inhibiting both gastric acid secretion as well as gastric motility. Lipid absorption stimulates apo A-IV synthesis and secretion by the jejunum. In addition to lipid feeding, there is evidence that a factor which is released as a result of lipid absorption in the distal small intestine also stimulates the synthesis and release of apo A-IV by the jejunum. This factor is probably PYY.     

Effect of delayed blood processing on the yield of factor VIII in cryoprecipitate and factor VIII concentrate

Current standards for the preparation of factor VIII (FVIII) concentrates from human plasma recommend separation of plasma from red cells (RBCs) within 6 hours of blood donation, thereby reducing the volume of plasma from donated whole blood available for processing to FVIII concentrate. The decay of FVIII clotting activity (FVIII:C) in whole blood and plasma stored at 22 and 4 degrees C and the recovery of FVIII:C in cryoprecipitate and FVIII concentrate prepared from plasma separated from whole blood stored overnight at 4 degrees C were investigated. In whole blood stored at 22 degrees C and plasma stored at either 4 or 22 degrees C, 90 percent of the original FVIII:C was present at 6 hours, 80 percent at 12 hours, and 65 to 70 percent at 18 hours. At these times lower levels of FVIII:C were recovered from whole blood stored at 4 degrees C, that is, 84, 68, and 56 percent, respectively. In cryoprecipitates prepared from plasma separated from RBCs after 18 hours' storage at 4 degrees C (18-hour plasma), 43 percent of FVIII:C activity was recovered, as compared with 61 percent recovered from standard plasma separated within 6 hours of donation (6-hour plasma), p less than 0.05. With large-scale preparation of FVIII concentrates, however, the yield of FVIII:C was similar whether 18- or 6-hour plasma was used. Thus FVIII concentrates--but not cryoprecipitates--can be prepared from plasma separated from whole blood stored at 4 degrees C for up to 18 hours without undue loss of potency.