Time Series Input Selection using Multiple Kernel Learning

In this paper we study the relevance of multiple kernel learning (MKL) for the automatic selection of time series inputs. Recently, MKL has gained great attention in the machine learning community due to its flexibility in modelling complex patterns and performing feature selection. In general, MKL constructs the kernel as a weighted linear combination of basis kernels, exploiting different sources of information. An efficient algorithm wrapping a Support Vector Regression model for optimizing the MKL weights, named SimpleMKL, is used for the analysis. In this sense, MKL performs feature selection by discarding inputs/kernels with low or null weights. The approach proposed is tested with simulated linear and nonlinear time series (AutoRegressive, Henon and Lorenz series).

Strategies for sequential prediction of stationary time series

We present simple procedures for the prediction of a real valued sequence. The algorithms are based on a combinationof several simple predictors. We show that if the sequence is a realization of a bounded stationary and ergodic random process then the average of squared errors converges, almost surely, to that of the optimum, given by the Bayes predictor. We offer an analog result for the prediction of stationary gaussian processes.

On a series of Goldbach and Euler

Theorem 1 of Euler s paper of 1737 'Variae Observationes Circa Series Infinitas', states the astonishing result that the series of all unit fractions whose denominators are perfect powers of integers minus unity has sum one. Euler attributes the Theorem to Goldbach. The proof is one of those examples of misuse of divergent series to obtain correct results so frequent during the seventeenth and eighteenth centuries. We examine this proof closelyand, with the help of some insight provided by a modern (and completely dierent) proof of the Goldbach-Euler Theorem, we present a rational reconstruction in terms which could be considered rigorous by modern Weierstrassian standards. At the same time, with a few ideas borrowed from nonstandard analysis we see how the same reconstruction can be also be considered rigorous by modern Robinsonian standards. This last approach, though, is completely in tune with Goldbach and Euler s proof. We hope to convince the reader then how, a few simple ideas from nonstandard analysis, vindicate Euler's work.

Improved nonparametric confidence intervals in time series regressions

Condence intervals in econometric time series regressions suffer fromnotorious coverage problems. This is especially true when the dependencein the data is noticeable and sample sizes are small to moderate, as isoften the case in empirical studies. This paper suggests using thestudentized block bootstrap and discusses practical issues, such as thechoice of the block size. A particular data-dependent method is proposedto automate the method. As a side note, it is pointed out that symmetricconfidence intervals are preferred over equal-tailed ones, since theyexhibit improved coverage accuracy. The improvements in small sampleperformance are supported by a simulation study.

Human large granular lymphocytes contain an esterase activity usually considered as specific for the myeloid series.

A cytochemical marker such as alpha-naphthyl acetate esterase (ANAE) has been found useful for the morphological identification of the subset of T lymphocytes having receptors for Fcμ (TM cells). ANAE reaction on TM cells gives a typical pattern of one to four positive spots, whereas this pattern is not found on T cells with receptors for Fcγ (TG cells). ANAE is abundant in monocytes but not detectable in granulocytes. Herein another type of esterase activity, naphthol-AS-D chloroacetate esterase (NCAE), is described; it is known to be abundant in granulocytes and was found to give a specific pattern of reactivity with the subpopulation of large granular lymphocytes (LGL). This pattern of fine granular staining was observed not only on LGL present in the TG cell subpopulation but also in LGL present in the non-T, non-B cells. Fractions of peripheral blood mononuclear cells which were ènriched up to 80% in LGL by Percoll discontinuous density gradient gave a similar percentage of specific NCAE pattern. In addition, among the different fractions from Percoll gradient, there was a good correlation (r = 0.94) between the number of NCAE-positive cells and the natural killer activity against the natural killer susceptible K562 target cells. It will be important to determine whether or not this enzymatic activity plays a role in the cytotoxic activities of LGL.

Clinical characteristics and outcome of isolated extracerebral relapses of primary central nervous system lymphoma: a case series.

There is very limited data on isolated systemic relapses of primary central nervous system lymphomas (PCNSL). We retrospectively reviewed the clinical characteristics and outcome of 10 patients with isolated systemic disease among 209 patients with PCNSL mainly treated with methotrexate-based chemotherapy (CT) with or without radiation therapy (RT). Isolated systemic relapse remained rare (4.8%, 10/209 patients). Median time from initial diagnosis to relapse was 33 months (range, 3-94). Sites of relapse were mostly extranodal. Three patients presented with early extra-cerebral (EC) relapse 3, 5 and 8 months from the beginning of initial treatment, respectively, and 7 patients had later relapses (range, 17-94 months). Treatment at relapse included surgery alone, RT alone, CT with or without radiotherapy, or CT with autologous stem cell transplantation (ASCT). Median overall survival (OS) after relapse was 15.5 months (range, 5.8-24.5) compared to 4.6 months (range, 3.6-6.5) for patients with central nervous system (CNS) relapse (p = 0.35). In conclusion, isolated systemic relapses exist but are infrequent. Early EC relapse suggests the presence of systemic disease undetectable by conventional evaluation at initial diagnosis. Patient follow-up must be prolonged because systemic relapse can occur as late as 10 years after initial diagnosis. Whether EC relapses of PCNSL have a better prognosis than CNS relapses needs to be assessed in a larger cohort. Copyright © 2010 John Wiley & Sons, Ltd.

Combined approach for the management of large vestibular schwannomas: Planned subtotal resection followed by gamma knife surgery in a series of 20 consecutive cases.

INTRODUCTION: The management of large lesions of the skull base, such as vestibular schwannomas (VS) is challenging. Microsurgery remains the main treatment option. Combined approaches (planned subtotal resection followed by gamma knife surgery (GKS) for residual tumor long-term control) are being increasingly considered to reduce the risk of neurological deficits following complete resection. The current study aims to prospectively evaluate the safety-efficacy of combined approach in patients with large VS. MATERIALS AND METHODS: We present our experience with planned subtotal resection followed by gamma knife surgery (GKS) in a consecutive a series of 20 patients with large vestibular schwannomas, treated between 2009 and 2014 in Lausanne University Hospital, Switzerland. Clinical and radiological data and audiograms were prospectively collected for all patients, before and after surgery, before and after GKS, at regular intervals, in dedicated case-report forms. Additionally, for GKS, dose-planning parameters were registered. RESULTS: Twenty patients (6 males and 14 females) with large VS had been treated by this approach. The mean age at the time of surgery was 51.6years (range 34.4-73.4). The mean presurgical diameter was 36.7 (range 26.1-45). The mean presurgical tumor volume was 15.9cm(3) (range 534.9). Three patients (15%) needed a second surgical intervention because of high volume of the tumor remnant considered too large for a safe GKS. The mean follow-up after surgery was 27.2months (range 6-61.3). The timing of GKS was decided on the basis of the residual tumor shape and size following surgery. The mean duration between surgery and GKS was 7.6months (range 413.9, median 6months). The mean tumor volume at the time of GKS was 4.1cm(3) (range 0.5-12.8). The mean prescription isodose volume was 6.3cm(3) (range 0.8-15.5). The mean number of isocenters was 20.4 (range 11-31) and the mean marginal prescription dose was 11.7Gy (range 11-12). We did not have any major complications in our series. Postoperative status showed normal facial nerve function (House-Brackmann grade I) in all patients. Six patients with useful pre-operative hearing (GR class 1) underwent surgery with the aim to preserve cochlear nerve function; of these patients, 5 (83.3%) of them remained in GR class 1 and one (16.7%) lost hearing (GR class 5). Two patients having GR class 3 at baseline remained in the same GR class, but the tonal audiometry improved in one of them during follow-up. Eleven patients (57.8%) were in GR class 5 preoperatively; one patient improved hearing after surgery, passing to GR class 3 postoperatively. Following GKS, there were no new neurological deficits, with facial and hearing function remaining identical to that after surgery. CONCLUSION: Our data suggest that planned subtotal resection followed by GKS has an excellent clinical outcome with respect to retaining facial and cochlear nerve function. This represents a paradigm shift of the treatment goals from a complete tumor excision perspective to that of a surgery designed to preserve neural functions. As long-term results emerge, this approach of a combined treatment (microsurgery and GKS) will most probably become the standard of care in the management of large vestibular schwanomma.

Primary cutaneous posttransplant lymphoproliferative disorders in solid organ transplant recipients: a multicenter European case series.

Primary cutaneous posttransplant lymphoproliferative disorders (PTLD) are rare. This retrospective, multicenter study of 35 cases aimed to better describe this entity. Cases were (re)-classified according to the WHO-EORTC or the WHO 2008 classifications of lymphomas. Median interval between first transplantation and diagnosis was 85 months. Fifty-seven percent of patients had a kidney transplant. Twenty-four cases (68.6%) were classified as primary cutaneous T cell lymphoma (CTCL) and 11 (31.4%) as primary cutaneous B cell PTLD. Mycosis fungoides (MF) was the most common (50%) CTCL subtype. Ten (90.9%) cutaneous B cell PTLD cases were classified as EBV-associated B cell lymphoproliferations (including one plasmablastic lymphoma and one lymphomatoid granulomatosis) and one as diffuse large B cell lymphoma, other, that was EBV-negative. Sixteen (45.7%) patients died after a median follow-up of 19.5 months (11 [68.8%] with CTCL [6 of whom had CD30(+) lymphoproliferative disorders (LPD)] and 5 [31.2%] with cutaneous B cell PTLD. Median survival times for all patients, CTCL and cutaneous B cell PTLD subgroups were 93, 93, and 112 months, respectively. Survival rates for MF were higher than those for CD30(+) LPD. The spectrum of primary CTCL in organ transplant recipients (OTR) is similar to that in the general population. The prognosis of posttransplant primary cutaneous CD30(+) LPD is worse than posttransplant MF and than its counterpart in the immunocompetent population. EBV-associated cutaneous B cell LPD predominates in OTR.

Posterior quadrantic epilepsy surgery: technical variants, surgical anatomy, and case series.

OBJECTIVE: Patients with intractable epilepsy due to extensive lesions involving the posterior quadrant (temporal, parietal, and occipital lobes) form a small subset of epilepsy surgery. This study was done with a view to analyze our experience with this group of patients and to define the changes in the surgical technique over the last 15 years. We also describe the microsurgical technique of the different surgical variants used, along with their functional neuroanatomy. METHODS: In this series there were 13 patients with a median age of 17 years. All patients had extensive presurgical evaluation that provided concordant evidence localizing the lesion and seizure focus to the posterior quadrant. The objective of the surgery was to eliminate the effect of the epileptogenic tissue and preserve motor and sensory functions. RESULTS: During the course of this study period of 15 years, the surgical procedure performed evolved toward incorporating more techniques of disconnection and minimizing resection. Three technical variants were thus utilized in this series, namely, (i) anatomical posterior quadrantectomy (APQ), (ii) functional posterior quadrantectomy (FPQ), and (iii) periinsular posterior quadrantectomy (PIPQ). After a median follow-up period of 6 years, 12/13 patients had Engel's Class I seizure outcome. CONCLUSION: The results of surgery for posterior quadrantic epilepsy have yielded excellent seizure outcomes in 92% of the patients in the series with no mortality or major morbidity. The incorporation of disconnective techniques in multilobar surgery has maintained the excellent results obtained earlier with resective surgery.

Severe primary cytomegalovirus infection in the immunocompetent adult patient: A case series.

Abstract Primary cytomegalovirus (CMV) infection is rare in immunocompetent adults, even rarer in elderly patients. Little is known about the severity of symptoms and the clinical course in this patient group. In children and younger adults, CMV mostly presents as an asymptomatic disease or a self-limiting mild mononucleosis-like syndrome. We describe the clinical course of an unusually severe primary CMV infection in a 69-y-old otherwise healthy man, as well as 6 other severe cases in immunocompetent adults at our institution, and compare them to adult cases from the literature. CMV primary infection and antiviral treatment should be considered in immunocompetent elderly persons presenting with a severe mononucleosis-like syndrome.

Multiple sclerosis-related trigeminal neuralgia: a prospective series of 43 patients treated with gamma knife surgery with more than one year of follow-up.

BACKGROUND: Trigeminal neuralgia (TN) related to multiple sclerosis (MS) is more difficult to manage pharmacologically and surgically. OBJECTIVE: This article aims to evaluate the safety and efficacy of Gamma Knife surgery (GKS) in this special group of patients. METHODS: Between July 1992 and November 2010, 43 cases with more than 1 year of follow-up were operated with GKS for TN related to MS and prospectively evaluated in the Timone University Hospital, Marseille, France. Radiosurgery using the Gamma Knife (model B or C or Perfexion) was performed. A single 4-mm isocenter was positioned at a median distance of 8 mm (range 5.7-14.7) anterior to the emergence of the nerve. A median maximum dose of 85 Gy (range 75-90) was delivered. RESULTS: The median follow-up period was 53.8 months (12-157.1). Thirty-nine patients (90.7%) were initially pain free. Their actuarial probability of remaining pain free without medication at 6 months, 1, 3, 5 and 10 years was 87.2, 71.8, 43.1, 38.3 and 20.5%, respectively, and remained stable till 12 years. The hypoesthesia actuarial rate at 6 months, 1 and 2 years was 11.5, 11.5 and 16%, and remained stable till 12 years. CONCLUSIONS: GKS proved safe and effective in this special group of patients.

A simple randomized algorithm for consistent sequential prediction of ergodic time series

We present a simple randomized procedure for the prediction of a binary sequence. The algorithm uses ideas from recent developments of the theory of the prediction of individual sequences. We show that if thesequence is a realization of a stationary and ergodic random process then the average number of mistakes converges, almost surely, to that of the optimum, given by the Bayes predictor.

Interstitial pregnancies' diagnosis and management: an eleven cases series.

PRINCIPLES: Interstitial pregnancy represents 2% of ectopic pregnancies, but it is a highly morbid condition with a 2.5% of maternal mortality. Its diagnostic and therapeutic management remains controversial. The aim of this review is to describe the management of interstitial pregnancy in our institution between 2001 and 2011 and to define some general rules for the clinical practice. METHODS: Single institution retrospective study. RESULTS: Eleven women were treated for interstitial pregnancy. The median age was 33 years and the median gestity was 4. Seven patients had a history of gynaecological surgery and four interstitial pregnancies followed in vitro fertilisation. The diagnosis was made at a median gestational age of seven weeks with a median beta-HCG level of 5,838 U/l. Six of the eleven patients received an initial treatment with intracornual methotrexate, three with intramuscular methotrexate and two with surgery. The median time to beta-HCG resolution was 58 days. Three of the eleven patients needed a second line treatment: two after intramuscular methotrexate and one after intracornual methotrexate. Six patients had further pregnancies and delivered by caesarean section. CONCLUSIONS: A high prevalence of previous ectopic pregnancies, gynaecological surgery and of pregnancies resulting from in vitro fertilisation was observed. The earliness of the diagnosis was the factor that allowed a conservative treatment in most cases. Beta-HCG level follow up was fundamental in allowing a second line therapy but beta-HCG can persist over a long period of time and this must be taken into account due to its possible psychological impact. Intracornual methotrexate seems to be more efficacious than intramuscular methotrexate in our series.