Deep learning algorithm performs similarly to radiologists in the assessment of prostate volume on MRI.

OBJECTIVES: Prostate volume (PV) in combination with prostate specific antigen (PSA) yields PSA density which is an increasingly important biomarker. Calculating PV from MRI is a time-consuming, radiologist-dependent task. The aim of this study was to assess whether a deep learning algorithm can replace PI-RADS 2.1 based ellipsoid formula (EF) for calculating PV. METHODS: Eight different measures of PV were retrospectively collected for each of 124 patients who underwent radical prostatectomy and preoperative MRI of the prostate (multicenter and multi-scanner MRI's 1.5 and 3 T). Agreement between volumes obtained from the deep learning algorithm (PVDL) and ellipsoid formula by two radiologists (PVEF1 and PVEF2) was evaluated against the reference standard PV obtained by manual planimetry by an expert radiologist (PVMPE). A sensitivity analysis was performed using a prostatectomy specimen as the reference standard. Inter-reader agreement was evaluated between the radiologists using the ellipsoid formula and between the expert and inexperienced radiologists performing manual planimetry. RESULTS: PVDL showed better agreement and precision than PVEF1 and PVEF2 using the reference standard PVMPE (mean difference [95% limits of agreement] PVDL: -0.33 [-10.80; 10.14], PVEF1: -3.83 [-19.55; 11.89], PVEF2: -3.05 [-18.55; 12.45]) or the PV determined based on specimen weight (PVDL: -4.22 [-22.52; 14.07], PVEF1: -7.89 [-30.50; 14.73], PVEF2: -6.97 [-30.13; 16.18]). Inter-reader agreement was excellent between the two experienced radiologists using the ellipsoid formula and was good between expert and inexperienced radiologists performing manual planimetry. CONCLUSION: Deep learning algorithm performs similarly to radiologists in the assessment of prostate volume on MRI. KEY POINTS: • A commercially available deep learning algorithm performs similarly to radiologists in the assessment of prostate volume on MRI. • The deep-learning algorithm was previously untrained on this heterogenous multicenter day-to-day practice MRI data set.

Patterns of markers of inflammation, coagulation and vasoconstriction during follow-up of abdominal aortic aneurysms.

AIM: The etiology of abdominal aortic aneurysm (AAA) includes inflammation, coagulation, and endothelial dysfunction. We have prospectively evaluated relations between these mechanisms and AAA growth. Tumour necrosis factor (TNF)-α, interleukin (IL)-6, endothelin (ET)-1, CD40 ligand and the complex formed between activated protein C (APC) and protein C inhibitor (PCI) were measured annually and related to AAA growth during up to 5 years in 206 patients with conservatively followed AAA. METHODS: We evaluated 163 patients up to 1 year, 126 patients up to 2 years, 83 patients up to 3 years, 53 patients up to 4 years, and 33 patients up to 5 years. The total number of patient follow-up years was 458. RESULTS: ET-1 remained unchanged except for a tendency to increase in the third and fourth years of follow-up. TNF-α decreased significantly during the first year and thereafter increased back to baseline values. There were no changes in IL-6, CD40 ligand, and APC-PCI complex. When patients in the highest and lowest quartiles of AAA growth up to 5 years follow-up were compared, APC-PCI complex levels tended to be higher (P=0.06) in the highest quartile of growth at three years (0.45 µg/l [i.q.r. 0.40-0.77] versus 0.28 µg/L [i.q.r. 0.14-0.36]). Δ-values of ET-1 and TNF-α did not show any correlation to growth. The 14 AAA patients that ruptured during follow-up did not differ from patients with non-ruptured AAA regarding biomarkers. CONCLUSION: In conclusion, none of the investigated mediators could be used to predict growth or rupture, or help to prolong intervals between ultrasound examinations in follow-up of AAA patients.