A concurrent, deep learning-based computer-aided detection system for prostate multiparametric MRI: a performance study involving experienced and less-experienced radiologists.

OBJECTIVES: To evaluate the effect of a deep learning-based computer-aided diagnosis (DL-CAD) system on experienced and less-experienced radiologists in reading prostate mpMRI. METHODS: In this retrospective, multi-reader multi-case study, a consecutive set of 184 patients examined between 01/2018 and 08/2019 were enrolled. Ground truth was combined targeted and 12-core systematic transrectal ultrasound-guided biopsy. Four radiologists, two experienced and two less-experienced, evaluated each case twice, once without (DL-CAD-) and once assisted by DL-CAD (DL-CAD+). ROC analysis, sensitivities, specificities, PPV and NPV were calculated to compare the diagnostic accuracy for the diagnosis of prostate cancer (PCa) between the two groups (DL-CAD- vs. DL-CAD+). Spearman's correlation coefficients were evaluated to assess the relationship between PI-RADS category and Gleason score (GS). Also, the median reading times were compared for the two reading groups. RESULTS: In total, 172 patients were included in the final analysis. With DL-CAD assistance, the overall AUC of the less-experienced radiologists increased significantly from 0.66 to 0.80 (p = 0.001; cutoff ISUP GG ≥ 1) and from 0.68 to 0.80 (p = 0.002; cutoff ISUP GG ≥ 2). Experienced radiologists showed an AUC increase from 0.81 to 0.86 (p = 0.146; cutoff ISUP GG ≥ 1) and from 0.81 to 0.84 (p = 0.433; cutoff ISUP GG ≥ 2). Furthermore, the correlation between PI-RADS category and GS improved significantly in the DL-CAD + group (0.45 vs. 0.57; p = 0.03), while the median reading time was reduced from 157 to 150 s (p = 0.023). CONCLUSIONS: DL-CAD assistance increased the mean detection performance, with the most significant benefit for the less-experienced radiologist; with the help of DL-CAD less-experienced radiologists reached performances comparable to that of experienced radiologists. KEY POINTS: • DL-CAD used as a concurrent reading aid helps radiologists to distinguish between benign and cancerous lesions in prostate MRI. • With the help of DL-CAD, less-experienced radiologists may achieve detection performances comparable to that of experienced radiologists. • DL-CAD assistance increases the correlation between PI-RADS category and cancer grade.

Incidence of symptomatic vertebral fractures in patients after percutaneous vertebroplasty.

The aim of this study was to evaluate the incidence of secondary symptomatic vertebral compression fractures (VCFs) in patients previously treated by percutaneous vertebroplasty (VTP). Three hundred sixteen patients with 486 treated VCFs were included in the study according to the inclusion criteria. Patients were kept in regular follow-up using a standardized questionairre before, 1 day, 7 days, 6 months, and 1 year after, and, further on, on a yearly basis after VTP. The incidence of secondary symptomatic VCF was calculated, and anatomical distribution with respect to previous fractures characterized. Mean follow-up was 8 months (6-56 months) after VTP. Fifty-two of 316 (16.4 %) patients (45 female, 7 male) returned for treatment of 69 secondary VCFs adjacent to (35/69; 51%) or distant from (34/69; 49%) previously treated levels. Adjacent secondary VCF occurred significantly more often compared to distant secondary VCF. Of the total 69 secondary VCFs, 35 of 69 occurred below and 27 of 69 above pretreated VCFs. Of the 65 sandwich levels generated, in 7 of 65 (11%) secondary VCFs were observed. Secondary VCF below pretreated VCF occurred significantly earlier in time compared to VCF above and compared to sandwich body fractures. No major complication occurred during initial or follow-up intervention. We conclude that secondary VCFs do occur in individuals after VTP but the rate found in our study remains below the level expected from epidemiologic studies. Adjacent fractures occur more often and follow the cluster distribution of VCF as expected from the natural history of the underlying osteoporosis. No increased rate of secondary VCF after VTP was observed in this retrospective analysis. In accordance with the pertinent literature, short-term and also midterm clinical results are encouraging and provide further support for the usefulness and the low complication rate of this procedure as an adjunct to the spectrum of pain management in patients with severe midline back pain due to osteoporotic spine fractures.

Detection of cement leakage after vertebroplasty with a non-flat-panel angio unit compared to multidetector computed tomography--an ex vivo study.

The purpose of this study was to investigate the detection of cement leakages after vertebroplasty using angiographic computed tomography (ACT) in a non-flat-panel angio unit compared to multidetector computed tomography (MDCT). Vertebroplasty was performed in 19 of 33 cadaver vertebrae (23 thoracic and 10 lumbar segments). In the angio suite, ACT (190 degrees ; 1.5 degrees per image) was performed to obtain volumetric data. Another volumetric data set of the specimen was obtained by MDCT using a standard algorithm. Nine multiplanar reconstructions in standardized axial, coronal, and sagittal planes of every vertebra were generated from both data sets. Images were evaluated on the basis of a nominal scale with 18 criteria, comprising osseous properties (e.g., integrity of the end plate) and cement distribution (e.g., presence of intraspinal cement). MDCT images were regarded as gold standard and analyzed by two readers in a consensus mode. Rotational acquisitions were analyzed by six blinded readers. Results were correlated with the gold standard using Cohen's j-coefficient analysis. Furthermore, interobserver variability was calculated. Correlation with the gold standard ranged from no correlation (osseous margins of the neuroforamen, j = 0.008) to intermediate (trace of vertebroplasty canula; j = 0.615) for criteria referring to osseous morphology. However, there was an excellent correlation for those criteria referring to cement distribution, with kappa values ranging from 0.948 (paravertebral cement distribution) to 0.972 (intraspinal cement distribution). With a minimum of j = 0.768 ("good correlation") and a maximum of j = 0.91 ("excellent"), interobserver variability was low. In conclusion, ACT in an angio suite without a flat-panel detector depicts a cement leakage after vertebroplasty as well as MDCT. However, the method does not provide sufficient depiction of osseous morphology.

Interventional management of renal bleeding after partial nephrectomy.

OBJECTIVE: Partial nephrectomy (PN) has emerged as a serious alternative to nephrectomy in oncologic therapy of renal tumours. While complications are rare in general, renal hemorrhage may occur und necessitate angiographic embolization. In this retrospective study, we evaluate the clinical, imaging and procedural findings of seven interventions in five patients with renal hemorrhage after PN. In four out of five patients (80%) the bleeding could be treated successfully by embolotherapy. CONCLUSION: Angiographic embolization in patients with renal hemorrhage after PN is feasible and has a high success rate. The procedure might facilitate avoidance of nephrectomy.

Complications in percutaneous vertebroplasty associated with puncture or cement leakage.

Due to the minimally invasive character and excellent clinical outcome of percutaneous vertebroplasty (PVP), the procedure is being performed in greatly increasing numbers. While PVP has a low complication rate in general, severe complications can occur. We focus on the imaging appearance of complications of PVP associated with puncture or cement leakage--from harmless to life-threatening.

CSF flow imaging in the management of third ventriculostomy with a reversed fast imaging with steady-state precession sequence.

The aim of this study was evaluation of a fast and slow-flow sensitive 2D steady-state free-precession sequence for its capability to prove the patency of endoscopic third ventriculostomy (TVS) in obstructive hydrocephalus, and to exclude communicating third ventricle prior to TVS. We compared gated and ungated variants of this sequence for this purpose. Twenty-three patients with obstructive hydrocephalus underwent 36 MR examinations with a 2D reversed fast imaging with steady-state precession (PSIF) sequence in a retrospectively cardiac gated (cine) and a faster but ungated version beside T1- and T2-weighted sequences in three planes. Thirteen patients were examined both before and after TVS, 4 patients solely before, and 6 patients solely after TVS. Imaging diagnoses were compared with intraoperative findings and clinical findings after TVS. Preoperative diagnosis of non-communicating third ventricle and cisterns was intraoperatively confirmed in 16 of 17 cases. Preoperative MRI was inconclusive in 1 case. Postoperative MRI revealed sufficient TVS in 16 of 19 cases and obstructed TVS in 3 of 19 cases due to several reasons. Findings at MRI were consistent in 19 of 19 cases with the clinical course and intraoperative results. The faster but ungated PSIF sequence was found to be diagnostically equivalent to the cardiac gated cine sequence. The CSF flow imaging with a 2D reversed fast imaging with steady-state precession sequence in conjunction with conventional T1- and T2-weighted images is a fast and reliable tool for pre- and postoperative functional evaluation in third ventriculostomy.

Flat panel digital radiography compared with storage phosphor computed radiography: assessment of dose versus image quality in phantom studies.

RATIONALE AND OBJECTIVE: To assess and quantify the dose reduction by use of a CsI-flat panel digital radiography (DR)-system compared with digital computed radiography (CR). MATERIALS AND METHODS: A TCDD-test using the CDRAD-phantom was performed at mAs-values of 5, 4, 2.5, 2, 1, and 0.5 mAs for both digital systems. Entrance surface doses were recorded for all images. Images were presented to four independent observers. For quantitative comparison the image quality figure (IQF) was calculated. Statistical analysis was performed using the Pearson correlation and the Wilcoxon test. A ROC analysis was performed using the TRG-phantom. Settings of 4, 2.5, 2 mAs for both systems were used. In addition, 1 and 0.5 mAs were used for the DR system only. Statistical significance was evaluated using Student test. RESULTS: The DR system provided equivalent results compared with CR with respect to high frequency information and superior results with respect to low contrast details. Compared with computed radiography, the flat panel detector demonstrated significantly lower IQFs, ensuring a better image quality with respect to contrast and detail detectability. IQFs for DR and CR were equal at a surface dose reduction of 87% for DR. ROC analysis revealed significantly higher values under the curve for DR up to a surface dose reduction of 70%. CONCLUSIONS: Image quality of DR proved to be far superior to CR in particular for low contrast details. The image quality of CR is similar to that of DR only at high dose levels.