Effect of the surface water layer on the optical signal in apertureless scanning near field optical microscopy

Optical signals measured in apertureless scanning near field optical microscopy (ASNOM) under ambient conditions are found to be affected significantly by the thin water layer absorbed on the surface under investigation, the presence of which is detected through measurements of the shear force experienced by the tip. This water layer also results in a large hysteresis between optical signals measured during approach and withdrawal of the tip to the sample surface. The role of this effect in ASNOM is anticipated to be significant, with the possibility of resultant topographically induced artefacts for ASNOM involving intermittent contact of tip and sample, but also providing a potential mechanism for nanoscale optical resolution.

High-resolution reconstruction of atmospheric deposition of trace metals and metalloids since AD 1400 recorded by ombrotrophic peat cores in Hautes-Fagnes, Belgium

The objective of our study was to determine the trace metal accumulation rates in the Misten bog, Hautes-Fagnes, Belgium, and assess these in relation to established histories of atmospheric emissions from anthropogenic sources. To address these aims we analyzed trace metals and metalloids (Pb, Cu, Ni, As, Sb, Cr, Co, V, Cd and Zn), as well as Pb isotopes, using XRF, Q-ICP-MS and MC-ICP-MS, respectively in two 40-cm peat sections, spanning the last 600 yr. The temporal increase of metal fluxes from the inception of the Industrial Revolution to the present varies by a factor of 5–50, with peak values found between AD 1930 and 1990. A cluster analysis combined with Pb isotopic composition allows the identification of the main sources of Pb and by inference of the other metals, which indicates that coal consumption and metallurgical activities were the predominant sources of pollution during the last 600 years.

Amniotic membrane transplantation for ocular surface reconstruction

Aims - To evaluate the efficacy of amniotic membrane transplantation (AMT) for ocular surface reconstruction. Methods - 10 consecutive patients who underwent AMT were included. The indications were: group A, cases with persistent epithelial defect after corneal abscess (n = 1), radiation (n = 1), or chemical burn (n = 3); group B, cases with epithelial defect and severe stromal thinning and impending or recent perforation, due to chemical burn (two patients, three eyes) or corneal abscess (n = 2); group C, to promote corneal epithelium healing and prevent scarring after symblepharon surgery with extensive corneo-conjunctival adhesion (n = 1). Under sterile conditions amniotic membrane was prepared from a fresh placenta of a seronegative pregnant woman and stored at -70°C. This technique involved the use of amniotic membrane to cover the entire cornea and perilimbal area in groups A and B, and the epithelial defect only in group C. Results - The cornea healed satisfactorily in four of five patients in group A, but the epithelial defect recurred in one of these patients. After AMT three patients underwent limbal transplantation and one penetrating keratoplasty and cataract extraction. In group B amniotic membrane transplantation was not helpful, and all cases underwent an urgent tectonic corneal graft. Surgery successfully released the symblepharon, promoted epithelialisation and prevented adhesions in the case of group C. Conclusion - AMT was effective to promote corneal healing in patients with persistent epithelial defect, and appeared to be helpful after surgery to release corneo-conjunctival adhesion. Most surgery for further surface rehabilitation. Amniotic membrane used as a patch was not effective to prevent tectonic corneal graft in cases with severe stromal thinning and impending or recent perforation.

Combining missing-feature theory, speech enhancement, and speaker-dependent/-independent modeling for speech separation

This paper considers the separation and recognition of overlapped speech sentences assuming single-channel observation. A system based on a combination of several different techniques is proposed. The system uses a missing-feature approach for improving crosstalk/noise robustness, a Wiener filter for speech enhancement, hidden Markov models for speech reconstruction, and speaker-dependent/-independent modeling for speaker and speech recognition. We develop the system on the Speech Separation Challenge database, involving a task of separating and recognizing two mixing sentences without assuming advanced knowledge about the identity of the speakers nor about the signal-to-noise ratio. The paper is an extended version of a previous conference paper submitted for the challenge.