Early detection increases overall survival among patients with lung cancer. This study formulated a machine learning method that processes chest X-rays (CXRs) to detect lung cancer early. After we preprocessed our dataset using monochrome and brightness correction, we used different kinds of preprocessing methods to enhance image contrast and then used U-net to perform lung segmentation. We used 559 CXRs with a single lung nodule labeled by experts to train a You Only Look Once version 4 (YOLOv4) deep-learning architecture to detect lung nodules. In a testing dataset of 100 CXRs from patients at Taipei Veterans General Hospital and 154 CXRs from the Japanese Society of Radiological Technology dataset, the sensitivity of the AI model using a combination of different preprocessing methods performed the best at 79%, with 3.04 false positives per image. We then tested the AI by using 383 sets of CXRs obtained in the past 5 years prior to lung cancer diagnoses. The median time from detection to diagnosis for radiologists assisted with AI was 46 (3–523) days, longer than that for radiologists (8 (0–263) days). The AI model can assist radiologists in the early detection of lung nodules.

The tori around active galactic nuclei (AGN) are potential formation sites for large amounts of dust, and they may help resolve the so-called dust budget crisis at high redshift. We investigate the dust composition in 53 of the 87 Palomar Green (PG) quasars showing the 9.7 micron silicate feature in emission. By simultaneously fitting the mid-infrared spectroscopic features and the underlying continuum, we estimate the mass fraction in various amorphous and crystalline dust species. We find that the dust consists predominantly of alumina and amorphous silicates, with a small fraction in crystalline form. The mean crystallinity is 8 +/- 6%, with more than half of the crystallinities greater than 5%, well above the upper limit determined for the Galaxy. Higher values of crystallinity are found for higher oxide fractions and for more luminous sources.