Spatially offset Raman scattering line-mapping as an adaptive tool ensuring accuracy for determination of component concentrations in tablets with different particle sizes.

Spatially offset Raman scattering (SORS) line-mapping was explored as a versatile tool to examine accuracy variations in compositional analyses of tablets with different particle sizes. SORS spectra collected near the laser irradiation were less representative of tablet composition due to the limited spectroscopic sampling volume, while the signal-to-noise (S/N) ratios of corresponding spectra were higher. On the other hand, SORS spectra at longer offset distances were better representative of tablet composition, while their S/N ratios were decreased considerably. Therefore, the use of only a certain portion of sliced (line-mapped) spectra balanced with the sample representation and S/N ratio could be advantageous to enhance accuracy. Moreover, a group of optimal slice spectra is expected to vary when the particle size of the tablet changes since the characteristics of internal photon propagation also would change. For the overall examination, SORS spectra of 30 Anaprox tablets (composed of 4 constituents including naproxen sodium) with 2 particle sizes (88.4 ± 11.8 µm and 118.9 ± 38.8 µm) were analyzed, and the concentrations of three components in these tablets were determined. A total of 6 cases (3 components and 2 particle sizes) were examined. When the average optimal slice spectra were employed in each case, the errors were lower compared to those using the average of all slice spectra. The demonstrated scheme was versatile to study the offset distance-dependent accuracy variations according to particle size and target component.

Asbestosis diagnosis algorithm combining the lung segmentation method and deep learning model in computed tomography image.

BACKGROUND: Early detection of asbestosis is important; hence, quick and accurate diagnostic tools are essential. This study aimed to develop an algorithm that combines lung segmentation and deep learning models that can be utilized as a clinical decision support system (CDSS) for diagnosing patients with asbestosis in segmented computed tomography (CT) images. METHODS: We accurately segmented the lungs in CT images of patients examined at Seoul St. Mary's Hospital using a threshold-based method. Lungs with asbestosis and normal lungs were classified by applying the segmented image to the long-term recurrent convolutional network deep learning model. Performance was evaluated using the area under the receiver operating characteristic curve (AUROC) and F1 score from the test data. RESULTS: The algorithm developed using the DenseNet201pre-trained model showed excellent performance, with a sensitivity of 0.962, specificity of 0.975, accuracy of 0.970, AUROC of 0.968, and F1 score of 0.961. CONCLUSIONS: We developed an algorithm with significantly better diagnostic accuracy than a radiologist (0.970 vs. 0.73-0.79). Our developed algorithm is expected to be an excellent support tool if used as a CDSS to diagnose asbestosis using CT images.