No abstract

³He MR imaging depicted significant improvements in the distribution of ³He gas after bronchodilator therapy in ex-smokers with COPD with and those without clinically important changes in FEV₁.

Concise and unambiguous assessment of a machine learning algorithm is key to classifier design and performance improvement. In the multi-class classification task, where each instance can only be labeled as one class, the confusion matrix is a powerful tool for performance assessment by quantifying the classification overlap. However, in the multi-label classification task, where each instance can be labeled with more than one class, the confusion matrix is undefined. Performance assessment of the multilabel classifier is currently based on calculating performance averages, such as hamming loss, precision, recall, and F-score. While the current assessment techniques present a reasonable representation of each class and overall performance, their aggregate nature results in ambiguity when identifying false negative (FN) and false positive (FP) results. To address this gap, we define a method of creating the multi-label confusion matrix (MLCM) based on three proposed categories of multi-label problems. Once establishing the shortcomings of current methods for identifying FN and FP, we demonstrate the usage of the MLCM with the classification of two publicly available multi-label data sets: i) a 12-lead ECG data set with nine classes, and ii) a movie poster data set with eighteen classes. A comparison of the MLCM results against statistics from the current techniques is presented to show the effectiveness in providing a concise and unambiguous understanding of a multi-label classifier behavior.

The objective of this study was to evaluate the regional effects of bronchodilator administration in chronic obstructive pulmonary disease (COPD) using hyperpolarized helium-3 ((3)He) MRI apparent diffusion coefficient (ADC). Ten COPD ex-smokers provided written, informed consent and underwent diffusion-weighted, hyperpolarized (3)He MRI, spirometry, and plethysmography before and 25 ± 2 min after bronchodilator administration. Pre- and postsalbutamol whole-lung (WL) ADC maps were generated and registered together to identify the lung regions containing the (3)He signal at both time points, and mean ADC within those regions of interest (ROI) was determined for a measurement of previously ventilated ROI ADC (ADC(P)). Lung ROI with (3)He signal at both time points was used as a binary mask on postsalbutamol WL ADC maps to obtain an ADC measurement for newly ventilated ROI (ADC(N)). Postsalbutamol, no significant differences were detected in WL ADC (P = 0.516). There were no significant differences between ADC(N) and ADC(P) postsalbutamol (P = 1.00), suggesting that the ADC(N) lung regions were not more emphysematous than the lung ROI participating in ventilation before bronchodilator administration. Postsalbutamol, a statistically significant decrease in ADC(P) (P = 0.01) was detected, and there were significant differences between ADC(P) in the most anterior and most posterior image slices (P = 0.02), suggesting a reduction in regional gas trapping following bronchodilator administration. Regional evaluation of tissue microstructure using hyperpolarized (3)He MRI ADC provides insights into lung alterations that accompany improvements in regional (3)He gas distribution after bronchodilator administration.

No abstract