Artificial intelligence software in pulmonary nodule assessment

Background:

This study tests the impact of the addition of autonomous computed tomography (CT) interpreting software to radiologist assessment of pulmonary nodules.

Methods:

Computed tomography scans for nodule assessment were identified retrospectively. Lung cancer risk factors, initial radiologist (RAD) report, Philips Lung Nodule software report (computer-aided nodule (CAD)) and radiologist report following the review of CT images and CAD (RAD + CAD) were collected. Follow-up recommendations based on current guidelines were derived from each report.

Results:

In all, 100 patients were studied. Median maximal diameter of the largest nodule reported by RAD and RAD + CAD were similar at 10.0 and 9.0 mm, respectively ( p = 0.06) but were reported as larger by CAD at 11.8 mm ( p < 0.001). Follow-up recommendations derived from RAD + CAD were less intensive in 23 (23%) and more intensive in 34 (34%) than that of RAD.

Discussion:

This study suggests that autonomous software use can alter radiologist assessment of pulmonary nodules such that suggested follow-up is altered.

The role of 18F-fluoro-2-deoxyglucose positron emission tomography/computed tomography in the management of patients with carcinoma of unknown primary

Purpose

Carcinoma of unknown primary is one of the ten most frequent cancers worldwide. Its median survival time is less than 10 months. Detecting primary tumour locations and/or occult metastatic lesions may inform definitive treatment and improve patients’ prognosis. We aimed to determine: (1) the sensitivity, specificity and accuracy of 18F-fluoro-2-deoxyglucose positron emission tomography/computed tomography; (2) its detection rate of primary tumour locations and occult metastases and (3) factors associated with improved survival times.

Methods

We retrospectively reviewed all cases in the West of Scotland for the period 1 December 2007 to 31 May 2011 that met all our selection criteria: (1) diagnosis of carcinoma of unknown primary; (2) a thorough but negative ‘work-up’ and (3) 18F-fluoro-2-deoxyglucose positron emission tomography/computed tomography report. Statistical methods included frequencies, Kaplan-Meier graphs and log-rank tests to compare survival times.

Results

18F-fluoro-2-deoxyglucose positron emission tomography/computed tomography detected primary tumour sites in 19/51 (37.3%) and occult metastases in 28/51 (54.9%) of eligible patients. Its sensitivity, specificity and accuracy were 79.2%, 70.4% and 74.5%, respectively; 20/51 (39.2%) patients died during the study period with a median survival of 8.4 months (range 21.4, SD ± 6.2). The number of metastatic locations was strongly associated with survival ( p = 0.002), but detection of a primary tumour site ( p = 0.174) or histopathology ( p = 0.301) was not.

Conclusion

18F-fluoro-2-deoxyglucose positron emission tomography/computed tomography detected occult metastatic sites in the majority and a primary cancer location in a substantial minority of patients. Our results were comparable with international literature and may indicate that 18F-fluoro-2-deoxyglucose positron emission tomography/computed tomography have an early role to improve the accuracy of cancer staging and to optimise carcinoma of unknown primary management.