[Use of autofluorescence bronchoscopy and spectrometry in the diagnosis of central lung cancer].

The paper evaluates the available data as well as our own on use of autofluorescence bronchoscopy in conjunction with spectrometric examination. We used qualitative and quantitative assessment of images obtained by conventional and autofluorescence (ClearVu Elite) means in real time. Our double-stage study evaluated sensitivity and specificity of autofluorescence bronchoscopy in diagnosing lung cancer as well as constructed spectrometric curves (ROC) and areas under them (AUC). Endoscopy was used in 171 patients with central lung cancer. Autofluorescence bronchoscopy established high sensitivity--94.74% (95%CI: 80.9-99%) and sufficient specificity--79.95% (95%CI: 75.8-83.6%). Application of a wide range of spectrometric coefficients contributed to high specificity thus reducing the number of biopsies as well as the injury from the treatment. The AUC for a best predictive index was 0.89 (99%: 0.83-0.95).

[Application of automated quantitative cytometry in diagnosis of lung cancer].

The paper deals with evaluation of the literature data and our experience with automated quantitative cytometric examination of sputum for diagnosis of lung cancer and, in particular, early one. This novel procedure uses measurement of quantitative indices which characterise tumors-induced alterations. The LungSign computerized system was employed to scan cellular nuclei. The results were evaluated by linear discriminative analysis with the aid of ROC-curves and underlying areas. The procedures were run in 248 cases and its sensitivity was significantly higher that of a standard cytological one (36.6% and 13.3%, respectively; p = 0.033), albeit a slight decrease in specificity (93.7% and 100%, respectively; p = 0.003). Automated quantitative cytometric indices varied significantly in cohorts of patients with confirmed (-0.275871) and false (-1.24990) diagnosis of lung cancer (p = 0.0001).