Recurrent neural networks and robust time series prediction

We propose a robust learning algorithm and apply it to recurrent neural networks. This algorithm is based on filtering outliers from the data and then estimating parameters from the filtered data. The filtering removes outliers from both the target function and the inputs of the neural network. The filtering is soft in that some outliers are neither completely rejected nor accepted. To show the need for robust recurrent networks, we compare the predictive ability of least squares estimated recurrent networks on synthetic data and on the Puget Power Electric Demand time series. These investigations result in a class of recurrent neural networks, NARMA(p,q), which show advantages over feedforward neural networks for time series with a moving average component. Conventional least squares methods of fitting NARMA(p,q) neural network models are shown to suffer a lack of robustness towards outliers. This sensitivity to outliers is demonstrated on both the synthetic and real data sets. Filtering the Puget Power Electric Demand time series is shown to automatically remove the outliers due to holidays. Neural networks trained on filtered data are then shown to give better predictions than neural networks trained on unfiltered time series.     

A unified approach to the surgical exposure of pancreatic and duodenal injuries

OBJECTIVE: It is necessary to have thorough knowledge of the survival of extreme low birth weight infants (ELBWI) in order to make it easier for obstetricians, neonatologists and the family to make a decision. PATIENTS AND METHODS: A revision of the 100 ELBWI in our service between 1988 and 1995, considering live births, those deceased in the same birthing room and those followed until their discharge from the hospital, was performed. The differences between the periods before and after the introduction of pulmonary surfactant in 1992 were analyzed. RESULTS: The total survival was 37% for those with a birth weight superior to 750 g or 26 weeks gestation. There were 44.2% males and 28.9% females. The total survival improved from 26.1% during 1988-1991 to 46.3% during the period of 1992-1995. During this period (1992-1995), the newborns weighing more than 750 g had a survival rate of 72.4% and for those of 26 weeks gestation it was 73.3%. Those born at 28 weeks gestation and those with 25 weeks of gestation and weighing more than 750 g, the total survival was 63% and the survival rate in the last four years was 75.9%. CONCLUSIONS: The mortality of the ELBWI descends in similar proportion to the remainder fo the ELBWI. In order to predict the prognosis, it would be necessary to carry out a correct ultrasound estimation of the gestational age and weight. It is necessary to offer a mother in the process of childbirth with a fetus of 28 weeks gestation or with 25 weeks gestation and a fetus with an ultrasound weight greater than 750 g, intrapartum fetal monitoring and to finish by Cesarean section in case of acute fetal distress, as well as intense and immediate neonatal attention as indicated by the index of survival reached in the group mentioned during the later years.     

Clarification of the relationship between free radical spin trapping and carbon tetrachloride metabolism in microsomal systems

It has been proposed that the C-phenyl-N-tert-butylnitrone/trichloromethyl radical adduct (PBN/.CCl3) is metabolized to either the C-phenyl-N-tert-butylnitrone/carbon dioxide anion radical adduct (PBN/.CO2-) or the glutathione (GSH) and CCl4-dependent PBN radical adduct (PBN/[GSH-.CCl3]). Inclusion of PBN/.CCl3 in microsomal incubations containing GSH, nicotinamide adenine dinucleotide phosphate (NADPH), or GSH plus NADPH produced no electron spin resonance (ESR) spectral data indicative of the formation of either the PBN/[GSH-.CCl3] or PBN/.CO2- radical adducts. Microsomes alone or with GSH had no effect on the PBN/.CCl3 radical adduct. Addition of NADPH to a microsomal system containing PBN/.CCl3 presumably reduced the radical adduct to its ESR-silent hydroxylamine because no ESR signal was observed. The Folch extract of this system produced an ESR spectrum that was a composite of two radicals, one of which had hyperfine coupling constants identical to those of PBN/.CCl3. We conclude that PBN/.CCl3 is not metabolized into either PBN/[GSH-.CCl3] or PBN/.CO2- in microsomal systems.     

The effect of pre-polymer/cross-linker storage on the elasticity and reliability of PDMS microfluidic devices

Polydimethylsiloxane (PDMS) is a commonly used material in biomedical engineering (Sollier et al. in Lab Chip 11(22):3752–3765, 2011; Palchesko et al. in PLoS ONE 7(12):e51499, 2012; Berthier et al. in Lab Chip 12(7):1224–1237, 2012). Its elastic nature makes PDMS especially attractive for microfluidic large-scale integration (mLSI) technology where micromechanical valves are actuated by deflecting a PDMS membrane under pressure. Therefore, understanding and control of PDMS elastic properties have commercial and scientific significance. In this study, we have investigated the effects of pre-polymer/cross-linker storage conditions on the mechanical properties of cured PDMS films as well as on microfluidic devices. We have showed that when the uncured components of PDMS are exposed to different humidity conditions, the elasticity of the PDMS changes and this is revealed as a change in the Young’s modulus of the cured PDMS. The high humidity (~85%) exposure for 24 h causes PDMS to become softer as confirmed by a significant decrease in the Young’s modulus values from 1.2 to 0.9 MPa. Furthermore, as the PDMS is exposed to high humidity conditions for longer periods (72 h), the Young’s modulus decreases down to 0.7 MPa. We found that exposing only the pre-polymer PDMS (Part A) to humid air does not alter the cured PDMS properties significantly, whereas exposure of the cross-linker (Part B) is responsible for the elasticity change. We have strictly controlled the storage humidity to build more reliable microfluidic chips using mLSI. As a result, actuation pressure of valves (10 psi) and defects of devices (in <30% of chips) are significantly reduced. These results suggest that to improve the manufacturing yield and reliability of PDMS devices, storage humidity should be controlled immediately after the material synthesis.     

Stage IA1 cervical adenocarcinoma: definition and treatment

OBJECTIVE: To propose a definition for stage IA1 cervical adenocarcinoma, based on the International Federation of Gynecology and Obstetrics (FIGO) staging system, and to determine if patients meeting criteria might be candidates for conservative surgery. METHODS: Two hundred women were diagnosed with early-stage cervical adenocarcinoma from 1982 to 1996. Histopathologic sections were reviewed by a gynecologic pathologist. Medical records were reviewed, and patients included in this study had microscopically identifiable lesions, up to 3 mm invasive depth, up to 7 mm tumor width, and negative margins if cone biopsy was performed. RESULTS: Twenty-one patients with microinvasive adenocarcinoma met criteria for FIGO stage IA1 carcinoma of the cervix. The median (range) follow-up was 76 (30-172) months and median (range) patient age was 38 (24-75) years. Definitive treatment included type II or III radical hysterectomy in 16 cases, simple abdominal or vaginal hysterectomy in four cases, and loop electrosurgical excision procedure in one case; one patient received adjuvant pelvic radiation. The histologic subtypes were endocervical adenocarcinoma in 18 cases, adenosquamous carcinoma in two cases, and clear-cell adenocarcinoma in one case. There was no evidence of parametrial invasion or lymph node metastases in any patient who had radical surgery, and there were no disease recurrences. CONCLUSION: Patients with microinvasive adenocarcinoma who met criteria for FIGO stage IA1 cervical carcinoma had disease limited to the cervix, and conservative surgery, such as cone biopsy or simple hysterectomy, might offer them definitive treatment.     

Clinical utility of human polymerized hemoglobin as a blood substitute after acute trauma and urgent surgery

We have previously documented the safety of 1 unit (50 gram) of human polymerized hemoglobin (Poly SFH-P) in healthy volunteers. This report describes the first patient trial to assess the therapeutic benefit of Poly SFH-P in acute blood loss. Thirty-nine patients received 1 (n = 14), 2 (n = 2), 3 (n = 15), or 6 (n = 8) units of Poly SFH-P instead of red cells as part of their blood replacement after trauma and urgent surgery. There were no safety issues related to the infusion of Poly SFH-P. The plasma hemoglobin concentration ([Hb]) after the infusion of 6 units (300 gram) of Poly SFH-P was 4.8 +/- 0.8 g/dL (mean +/- SD). Although the red cell [Hb] fell to 2.9 +/- 1.2 g/dL, the total [Hb] was maintained at 7.5 +/- 1.2 g/dL. Poly SFH-P maintained total [Hb], despite the marked fall in red cell [Hb] due to blood loss. The utilization of O2 (extraction ratio) was 27 +/- 16% from the red cells and 37 +/- 13% from the Poly SFH-P. Twenty-three patients (59%) avoided allogeneic transfusions during the first 24 hours after blood loss. Poly SFH-P effectively loads and unloads O2 and maintains total hemoglobin in lieu of red cells after acute blood loss, thereby reducing allogeneic transfusions. Poly SFH-P seems to be a clinically useful blood substitute.     

Driveway crush injuries in young children: a highly lethal, devastating, and potentially preventable event

BACKGROUND/PURPOSE: The aim of this study was to investigate driveway-related injuries in children, identify associated risk factors, and evaluate outcome compared with other mechanisms of blunt trauma. METHODS: A 6-year review (1991 to 1996) of pediatric (age less than 18 years) pedestrian injuries treated at two urban trauma centers was conducted: one regional pediatric trauma center and one level I trauma center with pediatric commitment. Five hundred twenty-seven children injured in pedestrian accidents were identified from the trauma registry; 51 children (10%) sustained traumatic injuries as a result of being struck in their driveway. Data are reported as mean +/- SEM. RESULTS: Children less than 5 years of age (n = 41) had an injury severity score (ISS) of 12.3+/-2.3, 15 (37%) sustained closed head injury, 13 (37%) had torso trauma, 19 (46%) skeletal trauma, and eight (20%) died. Children > or = 5 years old (n = 10) had an ISS of 10.7+/-2.4, three (30%) sustained closed head injury, four (40%) torso trauma, six (60%) skeletal trauma, and none died. In contrast, all other pediatric pedestrian accidents analyzed over the same time period had a mortality rate of only 2% (11 of 476). CONCLUSIONS: Pediatric driveway trauma carries a significant risk of head injury and a 10-fold increase in mortality in children under 5 years of age when compared with all other pediatric pedestrian accidents. More emphasis must be placed on injury prevention and public education to prevent this devastating mechanism of injury in these young, vulnerable children.