Dataset Information

Evaluating a Fully Automated Pulmonary Nodule Detection Approach and Its Impact on Radiologist Performance.

ABSTRACT:

Purpose

To compare sensitivity in the detection of lung nodules between the deep learning (DL) model and radiologists using various patient population and scanning parameters and to assess whether the radiologists' detection performance could be enhanced when using the DL model for assistance.

Materials and methods

A total of 12 754 thin-section chest CT scans from January 2012 to June 2017 were retrospectively collected for DL model training, validation, and testing. Pulmonary nodules from these scans were categorized into four types: solid, subsolid, calcified, and pleural. The testing dataset was divided into three cohorts based on radiation dose, patient age, and CT manufacturer. Detection performance of the DL model was analyzed by using a free-response receiver operating characteristic curve. Sensitivities of the DL model and radiologists were compared by using exploratory data analysis. False-positive detection rates of the DL model were compared within each cohort. Detection performance of the same radiologist with and without the DL model were compared by using nodule-level sensitivity and patient-level localization receiver operating characteristic curves.

Results

The DL model showed elevated overall sensitivity compared with manual review of pulmonary nodules. No significant dependence regarding radiation dose, patient age range, or CT manufacturer was observed. The sensitivity of the junior radiologist was significantly dependent on patient age. When radiologists used the DL model for assistance, their performance improved and reading time was reduced.

Conclusion

DL shows promise to enhance the identification of pulmonary nodules and benefit nodule management.© RSNA, 2019Supplemental material is available for this article.

SUBMITTER: Liu K

PROVIDER: S-EPMC8017422 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:An ever-growing collection of commercial biostimulants is becoming available in a wide variety of forms and compositions to improve crop performance. Given the intricate nature of deciphering the underlying mechanisms of commercial products, which typically comprise various biological components, it is crucial for research in this area to have robust tools to demonstrate their effectiveness in field trials. Here, we took a multi-attribute approach to evaluating the impact of biostimulants on crop performance. First, we assessed the impact of a biostimulant on the soil and rhizosphere microbiomes associated to crops in eight reference farms, including corn (3 farms), soybean (2), cotton (2) and sugarcane (1), in different biomes and production contexts in Brazil and Paraguay. Second, we modeled a set of integrated indicators to measure crop responses to biostimulant application, including five analytical themes as follows: i) crop development and production (9 indicators), ii) soil chemistry (9), iii) soil physics (5), iv) soil biology (6) and v) plant health (10). Amplicon 16S rRNA and ITS sequencing revealed that the use of the biostimulant consistently changes the structure of bacterial and fungal communities associated with the production system for all evaluated crops. In the rhizosphere samples, the most responsive bacterial taxa to biostimulant application were Prevotella in cotton; Prauserella and Methylovirgula in corn; and Methylocapsa in sugar cane. The most responsive fungal taxa to biostimulant use were Arachnomyces in soybean and cotton; and Rhizophlyctis in corn. The proposed integrated indicators yielded highly favorable positive impact indices (averaging at 0.80), indicating that biostimulant-treated fields correlate with better plant development and crop performance. Prominent indices were observed for indicators in four themes: soil biology (average index 0.84), crop production (0.81), soil physics (compaction reduction 0.81), and chemical fertility (0.75). The multi-attribute approach employed in this study offers an effective strategy for assessing the efficacy of biostimulant products across a wide range of crops and production systems.