查詢結果分析
來源資料
頁籤選單縮合
題 名 | 慢性阻塞性肺部疾病患者肺功能狀態與中醫證型及舌診影像科學化研究之關係=The Relationship between Lung Function Test of Chronic Obstructive Pulmonary Disease Patients and Disease Patterns of Traditional Chinese Medicine and Scientific Study of Tongue Image Analysis |
---|---|
作 者 | 陳建仲; | 書刊名 | 中醫藥年報 |
卷 期 | 17:2 1999.05[民88.05] |
頁 次 | 頁107-152 |
分類號 | 413.241、413.241 |
關鍵詞 | 舌診; 影像處理; 中醫診斷; 慢性阻塞性肺部疾病; 肺功能; 特徵擷取; RGB彩色分量; Tongue; Tongue diagnosis; Image analysis; Lung function test; Chronic obstructive pulmonary disease; COPD; Feature extraction; RGB color components; |
語 文 | 中文(Chinese) |
中文摘要 | 本研究在探討肺功能與舌診間之關係,並建立中醫診斷學中舌診的客觀指標。其蒐集51名肺部疾病患者,在控制的環境下擷取舌診影像,除了人為的觀察記錄外,並利用電腦以RBG和HSL模式分析舌診影像。研究結果發現:淡自舌、淡紅舌、紅舌、紫紅舌、淡紫舌、青紫舌等不同舌色,可反映出肺功能的不同嚴重程度。若以紅舌系列與紫舌系列比較,則顯現出紅舌系列病況較輕,紫舌病況較重。舌苔方面,若以苔色與肺功能作比較,白苔肺功能嚴重程度較輕,黃苔肺功能嚴重程度較重。本研究樣木中,臟俯證型以肺居多,腎居次,其次為心,肝、脾最少,而其分佈與肺功能狀態無關。虛實辨證和寒熱辨證與肺功能嚴重程度無關,但是有實證、熱證肺功能較差的區勢。此外,白苔、黃苔的自動判試,可用彩度、亮度和紅色、藍色、綠色等五個變數,或用藍色、綠色二個變數,經方程式計算後之數值作診斷依據。此外,自動化舌診首先必須從患者嘴部影像中擷取舌頭部分,本文藉由分析不同部位RGB彩色分量,利用其彩色分量間相對值差異。 強化舌頭邊緣和嘴唇、皮膚之間對比,再以邊界檢測法則擷取舌頭影像,利用色調與亮度特性將舌診辨証上兩大重要特徵-- 舌苔與舌質--分離,於舌頭區分成舌質部分與舌苔部分後,導出苔色、厚薄、偏全變化、剝脫、舌色、朱點、裂舌等特徵,以作為下一階段運用模糊理論斷症之依據。可見由舌頭上的表現,與肺功能狀態的好壞有關,而舌診自動判讀的目標可藉由電腦的運用來達成。 |
英文摘要 | The purpose of this study was to realize the relationship between pulmonary function and tongue diagnosis, and to create the objective index of tongue diagnosis of Chinese medicine. 51 cases of COPD patients were taken the tongue pictures under well-controlled environment. Beside observation, computerized analysis of tongue images with RGB and HSL models was performed. It revealed that the kinds of color, which were divided into pale, light red, red, purple-red, light purple, and dark purple, would agree with the severity of lung function. As to the condition of lung function, the purple series of tongue were more severe than red series. As to the condition of lung function, the yellow series of tongue-covering was more severe than white series. That automatic reading of tongue-covering white or yellow could be performed by the formula with five variables as lightness, saturation, red, blue, and green, or with two variables as blue and green. The extraction of tongue portion from the surrounding mouth image is the first step in the automatic tongue diagnosis. However, the separation of tongue from its surrounding lips and facial portions can not be based on simple grey-level differences due to the- close resemblance in terms of lighting condition. In this paper, an automatic feature extraction of a patient's tongue based on the RGB color components is proposed. The color differences in tongue and its surrounding areas are enhanced and discriminated by a contour extraction procedure. Once the boundary of the tongue is located, two primal characteristics utilized in the next stage fuzzy classification, tongue granularity and smoothness, are extracted. The color, thickness, area of distribution, peeling conditions, spottiness and cracks are then derived and employed as features to diagnose diseases. It revealed that the appearance of the tongue would correspond with the state of pulmonary function, and the goal of automatic diagnosis of the tongue pictures would be achieved by the use of computer. |
本系統中英文摘要資訊取自各篇刊載內容。