Efficiency of a computer-aided diagnosis (CAD) system with deep learning in detection of pulmonary nodules on 1-mm-thick images of computed tomography.

PURPOSE: To evaluate the performance of a deep learning-based computer-aided diagnosis (CAD) system at detecting pulmonary nodules on CT by comparing radiologists' readings with and without CAD. MATERIALS AND METHODS: A total of 120 chest CT images were randomly selected from patients with suspected lung cancer. The gold standard of nodules ≥ 3 mm was established by a panel of three expert radiologists. Two less experienced radiologists read the images without and afterward with CAD system. Their reading times were recorded. RESULTS: The radiologists' sensitivity increased from 20.9% to 38.0% with the introduction of CAD. The positive predictive value (PPV) decreased from 70.5% to 61.8%, and the F1-score increased from 32.2% to 47.0%. The sensitivity significantly increased from 13.7% to 32.4% for small nodules (3-6 mm) and from 33.3% to 47.6% for medium nodules (6-10 mm). CAD alone showed a sensitivity of 70.3%, a PPV of 57.9%, and an F1-score of 63.5%. Reading time decreased by 11.3% with the use of CAD. CONCLUSION: CAD improved the less experienced radiologists' sensitivity in detecting pulmonary nodules of all sizes, especially including a significant improvement in the detection of clinically important-sized medium nodules (6-10 mm) as well as small nodules (3-6 mm) and reduced their reading time.

Navigator-triggered and breath-hold 3D MRCP using compressed sensing: image quality and method selection factor assessment.

PURPOSE: To examine whether MRCP using a combination of compressed sensing and sensitivity encoding with navigator-triggered and breath-hold techniques (NT C-SENSE and BH C-SENSE, respectively) have comparable image quality to that of navigator-triggered MRCP using only sensitivity encoding (NT SENSE) at 1.5-T. METHODS: Fifty-one participants were enrolled in this prospective study between July and October 2018 and underwent the three 3D MRCP sequences each. The acquisition time and relative duct-to-periductal contrast ratios (RC values) of each bile duct segment were obtained. Visualization of the bile and main pancreatic ducts, background suppression, artifacts, and overall image quality were scored on 5-point scales. Mean and median differences in RC values and qualitative scores of NT C-SENSE and BH C-SENSE relative to NT SENSE were calculated with 95% confidence intervals (CIs). RESULTS: Acquisition time of NT SENSE, NT C-SENSE, and BH C-SENSE were 348, 143 (mean for both), and 18 s (for all participants), respectively. The RC value of each bile duct segment was inferior, but the lower limits of the 95% CIs of the mean differences were ≥ - 0.10, for both NT C-SENSE and BH C-SENSE. The visualization score of the intrahepatic duct in BH C-SENSE was inferior to that in NT SENSE (lower 95% CI limit, - 1.5). In both NT C-SENSE and BH C-SENSE, the 95% CIs of the median differences in the other qualitative scores were from - 1.0 to 0.0. CONCLUSION: NT C-SENSE and BH C-SENSE have comparable image quality to NT SENSE at 1.5-T.