Nonlinear Dynamic Projection for Noise Reduction of Dispersed Manifolds.

The search for a low-dimensional structure in high-dimensional data is one of the fundamental tasks in machine learning and pattern recognition. Manifold learning algorithms have recently emerged as alternatives to traditional linear dimension reduction techniques. In this paper, we propose a novel projection method that can be combined with any manifold learning methods to improve their dimension reduction performance when applied to high-dimensional data with a high level of noise. The method first builds a dispersion function that describes the distribution of dispersed manifold where the data lie. It then projects the noisy data onto a region wrapping the true manifold sufficiently close to it by applying a dynamical projection system associated with the constructed dispersion function. The effectiveness of the proposed projection method is validated by applying it to some real-world data sets with promising results.

A Study of Mixed Phenotype Acute Leukemia Based on the 2008 World Health Organization Classification / 대한진단검사의학회지

BACKGROUND: We evaluated the clinical significance of revised 2008 WHO classification needed to diagnose mixed phenotype acute leukemia (MPAL). METHODS: A total of 22 MPAL patients, previously diagnosed by applying the scoring system of the European Group for Immunological Classification of Acute Leukemias (EGIL) were reclassified based on the 2008 WHO classification. RESULTS: In 2008 WHO classification, the number of monoclonal antibodies (mAbs) required for assigning more than one lineage was markedly decreased, from 26 to 11, compared with that of EGIL. Seventeen of the 22 MPAL patients were reclassified as MPAL with following details: 6 MPAL with t(9;22)(q34;q11.2); BCR-ABL1, 1 MPAL with t(v;11q23); MLL rearranged, 7 MPAL, B/Myeloid, not otherwise specified (NOS) and 3 MPAL, T/Myeloid, NOS. Five patients were excluded from MPAL in the revised classification: 4 cytoplasmic myeloperoxidase (cMPO)-negative and 1 CD19-negative. The failure of complete remission achievement and occurrence of relapse were associated with poor prognosis (P=0.0002 and P=0.009, respectively). But the presence of Philadelphia chromsome was not significantly related with patient outcome (P=0.082). One patient with cCD79a, CD20, CD38, cMPO and CD15, whose diagnosis was reclassified from MPAL to AML has survived during the study period. CONCLUSIONS: Because of decreased number of mAbs needed, it is possible that acute leukemia panel is designed to include all mAbs required to diagnose MPAL according to 2008 WHO classification. When diagnosing MPAL, it is critical to figure out positivity in either cMPO or CD19, and AML expressing more than 2 lymphoid antigens are considered as MPAL.

Diamond-Blackfan Anemia Confirmed by RPS19 Gene Mutation Analysis: A Case Study and Literature Review of Korean Patients / 대한진단검사의학회지

Diamond-Blackfan anemia (DBA) is a rare congenital erythroid hypoplastic anemia that usually presents early in infancy and is inherited in up to 45% of cases. It is characterized by red cell aplasia, congenital anomalies, and a predisposition to cancer. Corticosteroids and red blood cell transfusions are the mainstays of therapy. We describe a case of 3-month-old infant who presented with severe anemia, elevated levels of HbF and adenosine deaminase and bilateral hydronephrosis, who was later confirmed as DBA by mutation analysis using the direct sequencing method. Direct sequencing analysis of RPS19 gene was performed with both cDNA and genomic DNA extracted from peripheral blood and a c.3G>A point mutation of exon 2 resulting in p.Met1Ile was identified in this patient. The patient showed an inadequate response to steroid therapy and a partial response to RBC transfusion with a follow-up Hb level of 8.3 g/dL on her last visit to the outpatient clinic. DBA is a genetically and phenotypically heterogeneous disease, and we have reviewed the clinical characteristics of 25 Korean patients thus far reported in the literature. To our knowledge, this is the first case of DBA confirmed by mutation analysis in Korea, and mutation identification using molecular method is recommended for confirmation of this genetically and phenotypically heterogeneous disease.

An improved cluster labeling method for support vector clustering.

The support vector clustering (SVC) algorithm is a recently emerged unsupervised learning method inspired by support vector machines. One key step involved in the SVC algorithm is the cluster assignment of each data point. A new cluster labeling method for SVC is developed based on some invariant topological properties of a trained kernel radius function. Benchmark results show that the proposed method outperforms previously reported labeling techniques.

Listerial peumonia and bacteremia in pregnant woman / 대한내과학회지

No abstract available.