Automated detection and segmentation of pulmonary embolisms on computed tomography pulmonary angiography (CTPA) using deep learning but without manual outlining

We present a novel computer algorithm to automatically detect and segment pulmonary embolisms (PEs) on computed tomography pulmonary angiography (CTPA). This algorithm is based on deep learning but does not require manual outlines of the PE regions. Given a CTPA scan, both intra- and extra-pulmonary arteries were firstly segmented. The arteries were then partitioned into several parts based on size (radius). Adaptive thresholding and constrained morphological operations were used to identify suspicious PE regions within each part. The confidence of a suspicious region to be PE was scored based on its contrast in the arteries. This approach was applied to the publicly available RSNA Pulmonary Embolism CT Dataset (RSNA-PE) to identify three-dimensional (3-D) PE negative and positive image patches, which were used to train a 3-D Recurrent Residual U-Net (R2-Unet) to automatically segment PE. The feasibility of this computer algorithm was validated on an independent test set consisting of 91 CTPA scans acquired from a different medical institute, where the PE regions were manually located and outlined by a thoracic radiologist (>18 years' experience). An R2-Unet model was also trained and validated on the manual outlines using a 5-fold cross-validation method. The CNN model trained on the high-confident PE regions showed a Dice coefficient of 0.676±0.168 and a false positive rate of 1.86 per CT scan, while the CNN model trained on the manual outlines demonstrated a Dice coefficient of 0.647±0.192 and a false positive rate of 4.20 per CT scan. The former model performed significantly better than the latter model (p<0.01). The promising performance of the developed PE detection and segmentation algorithm suggests the feasibility of training a deep learning network without dedicating significant efforts to manual annotations of the PE regions on CTPA scans.

Pelvis magnetic resonance imaging to diagnose familial partial lipodystrophy

CONTEXT

The diagnosis of familial partial lipodystrophy (FPLD) is currently made based on clinical judgement.

OBJECTIVE

There is a need for objective diagnostic tools that can diagnose FPLD accurately.

METHODS

We have developed a new method that used measurements from pelvic magnetic resonance imaging (MRI) at the pubis level. We evaluated measurements from a lipodystrophy cohort (n = 59; median age [25-75 percentiles]: 32 [24-44]; 48 females and 11 males) and age- and gender-matched controls (n = 29). Another dataset included MRIs from 289 consecutive patients.

RESULTS

Receiver operating characteristic curve (ROC) analysis revealed a potential cut-point of ≤ 13 mm gluteal fat thickness for the diagnosis of FPLD. A combination of gluteal fat thickness ≤ 13 mm and pubic/gluteal fat ratio ≥ 2.5 (based on a ROC) provided 96.67% (95% Confidence Interval [CI]: 82.78-99.92%) sensitivity and 91.38% (95% CI: 81.02-97.14%) specificity in the overall cohort and 100.00% (95% CI: 87.23-100.00%) sensitivity and 90.00% (95% CI: 76.34-97.21%) specificity in females for the diagnosis of FPLD. When this approach was tested in a larger dataset of random patients, FPLD was differentiated from subjects without lipodystrophy with 96.67% (95% CI: 82.78-99.92%) sensitivity and 100.00% (95% CI: 98.73-100.00%) specificity. When only women were analyzed, the sensitivity and the specificity was 100.00% (95%CI: 87.23-100.00% and 97.95-100.00%, respectively). The performance of gluteal fat thickness and pubic/gluteal fat thickness ratio was comparable to readouts performed by radiologists with expertise in lipodystrophy.

CONCLUSION

The combined use of gluteal fat thickness and pubic/gluteal fat ratio from pelvic MRI is a promising method to diagnose FPLD that can reliably identify FPLD in women. Our findings need to be tested in larger populations and prospectively.