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.

Impact of Bone Cement Augmentation on the Fixation Strength of TFNA Blades and Screws.

Background and Objectives: Hip fractures constitute the most debilitating complication of osteoporosis with steadily increasing incidences in the aging population. Their intramedullary nailing can be challenging because of poor anchorage in the osteoporotic femoral head. Cement augmentation of Proximal Femoral Nail Antirotation (PFNA) blades demonstrated promising results by enhancing cut-out resistance in proximal femoral fractures. The aim of this study was to assess the impact of augmentation on the fixation strength of TFN-ADVANCEDTM Proximal Femoral Nailing System (TFNA) blades and screws within the femoral head and compare its effect when they are implanted in centre or anteroposterior off-centre position. Materials and Methods: Eight groups were formed out of 96 polyurethane low-density foam specimens simulating isolated femoral heads with poor bone quality. The specimens in each group were implanted with either non-augmented or cement-augmented TFNA blades or screws in centre or anteroposterior off-centre positions, 7 mm anterior or posterior. Mechanical testing was performed under progressively increasing cyclic loading until failure, in setup simulating an unstable pertrochanteric fracture with a lack of posteromedial support and load sharing at the fracture gap. Varus-valgus and head rotation angles were monitored. A varus collapse of 5° or 10° head rotation was defined as a clinically relevant failure. Results: Failure load (N) for specimens with augmented TFNA head elements (screw/blade centre: 3799 ± 326/3228 ± 478; screw/blade off-centre: 2680 ± 182/2591 ± 244) was significantly higher compared with respective non-augmented specimens (screw/blade centre: 1593 ± 120/1489 ± 41; screw/blade off-centre: 515 ± 73/1018 ± 48), p < 0.001. For both non-augmented and augmented specimens failure load in the centre position was significantly higher compared with the respective off-centre positions, regardless of the head element type, p < 0.001. Augmented off-centre TFNA head elements had significantly higher failure load compared with non-augmented centrally placed implants, p < 0.001. Conclusions: Cement augmentation clearly enhances the fixation stability of TFNA blades and screws. Non-augmented blades outperformed screws in the anteroposterior off-centre position. Positioning of TFNA blades in the femoral head is more forgiving than TFNA screws in terms of failure load.

3D Computer graphical anatomy study of the femur: a basis for a new nail design.

BACKGROUND: Current intramedullary nails with a radius of curvature (ROC) of 1500-2000 mm sometimes cause distal anterior cortical encroachment. Furthermore, clinical data indicate that the proximal nail end is too long for some Asian patients. The objective of our study was to develop a comprehensive 3D measurement protocol that measures both the anatomy of the canal and the proximal region. The protocol was used to obtain measurements from Caucasian and Asian (Japanese and Thai) specimens. MATERIALS AND METHODS: A total of 90 3D bone models representative of hip fracture patients were reconstructed from CT data. RapidForm 2006 was used to generate the reference geometries required for determining radius and angulation of shaft antecurvature as well as measurements of the proximal anatomy. Multiple linear regression analyses were used to determine the relative contribution of height, age, ethnicity, gender, and body side on the total variance. RESULTS: The mean ROC in the natural 3D antecurvature plane was 885 mm overall, 974 mm in Caucasians and 787 mm in Asians. Height, age, ethnicity, gender, and body side significantly predicted ROC (R = 0.53, p = 0.000). The mean values of anteversion measurements for Asians (Japanese: 22.1°; Thai: 22.7°) were significantly larger than those of the Caucasians (14.5°; p = 0.001). There was virtually no difference (p = 0.186) between the measurements pertaining to the length of the proximal nail end between Caucasian and Asian samples. There was no significant difference between the mean neck-to-shaft angles (Caucasian: 126°; Japanese: 128.2°; Thai: 125.7°; p = 0.198 for Asians vs Caucasians). CONCLUSIONS: The developed comprehensive anatomical 3D measurement protocol could serve as standardised approach for anthropometric studies in the future. Our data suggest that the ROC of current nail designs should be reduced from between 1500 and 2000 to 1000 mm to achieve an improved fit for the investigated population.