<p>The realm of the current study is ovarian magnetic resonance spectroscopy (MRS). Presented are the selected recent advances in the Padé-based signal processing by shape estimations alone. The goal is to substantially improve extraction of quantitative information by sole reliance upon non-parametric estimations of total shape spectra (envelopes) from encoded time signals. The task is to resolve the given envelope into its true partial spectra (components) without solving the quantification problem (i.e. no polynomial rooting, etc.). The rescue is in derivative quantitative shape estimations void of fitting. Splitting apart an envelope into the genuine components amounts to quantification. With any quadrature rule, integrations of the reconstructed well-isolated unstructured derivative lineshapes and their power spectra determine the peak areas and peak widths, respectively. Metabolite concentrations ensue thereby as a key diagnostic information for recognized and potential cancer biomarkers alike. Special attention is drawn to abundant non-derivative singlet-appearing resonances that can contain sub-peaks in derivative lineshapes. Failure to detect such occurrences compromises the critical decision-making (normal vs. diseased tissues or biofluids) in the clinic. The salient illustrations are reported for benign and malignant tumors from human ovarian cyst fluid samples.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Recognized and potential cancer biomarkers in ovarian MRS: Padé quantitative derivative shape estimation without fitting

  • Dževad Belkić,
  • Karen Belkić

摘要

The realm of the current study is ovarian magnetic resonance spectroscopy (MRS). Presented are the selected recent advances in the Padé-based signal processing by shape estimations alone. The goal is to substantially improve extraction of quantitative information by sole reliance upon non-parametric estimations of total shape spectra (envelopes) from encoded time signals. The task is to resolve the given envelope into its true partial spectra (components) without solving the quantification problem (i.e. no polynomial rooting, etc.). The rescue is in derivative quantitative shape estimations void of fitting. Splitting apart an envelope into the genuine components amounts to quantification. With any quadrature rule, integrations of the reconstructed well-isolated unstructured derivative lineshapes and their power spectra determine the peak areas and peak widths, respectively. Metabolite concentrations ensue thereby as a key diagnostic information for recognized and potential cancer biomarkers alike. Special attention is drawn to abundant non-derivative singlet-appearing resonances that can contain sub-peaks in derivative lineshapes. Failure to detect such occurrences compromises the critical decision-making (normal vs. diseased tissues or biofluids) in the clinic. The salient illustrations are reported for benign and malignant tumors from human ovarian cyst fluid samples.