оглавление              текст:   pdf  

УДК: 612.821:615.47:616.12-008 



А. А. Десова 1, А. А. Дорофеюк 1, 2, А. М. Анохин 1, Ю. А. Дорофеюк 1


1 Институт проблем управления им. В.А.Трапезникова РАН, 117997, Москва, Профсоюзная, 65

2 ФИЦ «Информатика и управление» РАН, 119333, Москва, Вавилова, 44-2


Статья поступила в редакцию 28 июня 2017 г.


Аннотация. Приведены результаты исследования характеристик пульсового сигнала лучевой артерии применительно к выявлению онкологических заболеваний легких. Сформировано исходное пространство признаков, определяемых вариабельностью амплитудных и временных базовых параметров единичных колебаний сигнала, их взаимными соотношениями, а также рядом статистических характеристик динамических рядов. Определены наиболее информативные признаки ритмической структуры пульсового сигнала. Показана возможность и оценено качество дифференциальной диагностики заболеваний легких.

Ключевые слова: Пульсовой сигнал, дыхательная волна, частота пульса, ритмическая структура, диагностика, информативный признак.

Abstract. The paper presents the radial artery pulse signal characteristics analysis results, as applied to the solution of the tumor (malignant and benign) lungs diseases differential diagnosis problem. Research conducted on the real pulse signals basis obtained during patients clinical examinations at the radiology Department of the Sechenov’s Moscow medical Academy. The research purpose was to evaluate the possibility of the lung tumors diseases differential diagnosis reliability increasing at the clinical and radiological examination stage by using additional information extracted from the radial artery pulse signal characteristics. It is known that for clinical - radiological examination (without histological studies) the accuracy of differential diagnosis between central lung cancer and limited pulmonary fibrosis, as well as between peripheral lung cancer and tuberculoma is not high enough, which requires the use of additional examination methods. Informative signs were formed on the basis of values of the oscillatory component describing the amplitude and time base variability of the pulse signal basic period parameters, their mutual relations, and also a number of their statistical characteristics. The features informative value evaluation and the decision rule construction for investigated diseases differential diagnosis was carried out using pattern recognition algorithms. The peripheral pulse signal was measured on the radial artery of the patient's right hand using an optoelectronic pulse signal sensor. The analyzed dataset classification (90 vectors on "learning" and 50 vectors on the "exam") quality recognition amounted to 89.5% of correct decisions for first class (malignant diseases) and 78% for second class (benign diseases). As a classification criterion the clinical diagnoses were used, in some cases confirmed by the results of histological studies.  The diagnostic consider method does not require instrumental intervention, the use of expensive reagents, has no contraindications and excludes the radiation loads. The method is relatively fast, not traumatic, and allows simple and efficient differential diagnosis, regardless of the location of the pathological formation and its size. The obtained results indicate the diagnostic efficiency of the radial artery pulse signal rhythmic structure using during clinical and pre-hospital examinations of pulmonological patients. The method can be used as a population screening test in order to identify a predisposition to malignant pathology. 

Key words: Pulse signal, respiratory wave, pulse rate, rhythm structure, diagnostics, informative sign.


Ссылка на статью:

А. А. Десова, А. А. Дорофеюк, А. М. Анохин, Ю. А. Дорофеюк. Использование характеристик пульсового сигнала лучевой артерии в задаче дифференциальной диагностики онкологических и доброкачественных заболеваний легких. Журнал радиоэлектроники [электронный журнал]. 2017. №7. Режим доступа: http://jre.cplire.ru/jre/jul17/9/text.pdf