ABSTRACT
Studies of accuracy and reaction time in decision making often observe a speed-accuracy tradeoff, where either accuracy or reaction time is sacrificed for the other. While this effect may mask certain multisensory benefits in performance when accuracy and reaction time are separately measured, drift diffusion models (DDMs) are able to consider both simultaneously. However, drift diffusion models are often limited by large sample size requirements for reliable parameter estimation. One solution to this restriction is the use of hierarchical Bayesian estimation for DDM parameters. Here, we utilize hierarchical drift diffusion models (HDDMs) to reveal a multisensory advantage in auditory-visual numerosity discrimination tasks. By fitting this model with a modestly sized dataset, we also demonstrate that large sample sizes are not necessary for reliable parameter estimation.
ABSTRACT
BACKGROUND: Multiparametric magnetic resonance imaging (mpMRI) is now recommended prebiopsy in numerous healthcare regions based on the findings of high-quality studies from expert centres. Concern remains about reproducibility of mpMRI to rule out clinically significant prostate cancer (csPCa) in real-world settings. OBJECTIVE: To assess the diagnostic performance of mpMRI for csPCa in a real-world setting. DESIGN, SETTING, AND PARTICIPANTS: A multicentre, retrospective cohort study, including men referred with raised prostate-specific antigen (PSA) or an abnormal digital rectal examination who had undergone mpMRI followed by transrectal or transperineal biopsy, was conducted. Patients could be biopsy naïve or have had previous negative biopsies. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: The primary definition for csPCa was International Society of Urological Pathology (ISUP) grade group (GG) ≥2 (any Gleason ≥7); the accuracy for other definitions was also evaluated. RESULTS AND LIMITATIONS: Across ten sites, 2642 men were included (January 2011-November 2018). Mean age and PSA were 65.3yr (standard deviation [SD] 7.8yr) and 7.5ng/ml (SD 3.3ng/ml), respectively. Of the patients, 35.9% had "negative MRI" (scores 1-2); 51.9% underwent transrectal biopsy and 48.1% had transperineal biopsy, with 43.4% diagnosed with csPCa overall. The sensitivity and negative predictive value (NPV) for ISUP GG≥2 were 87.3% and 87.5%, respectively. The NPVs were 87.4% and 88.1% for men undergoing transrectal and transperineal biopsy, respectively. Specificity and positive predictive value of MRI were 49.8% and 49.2%, respectively. The sensitivity and NPV increased to 96.6% and 90.6%, respectively, when a PSA density threshold of 0.15ng/ml/ml was used in MRI scores 1-2; these metrics increased to 97.5% and 91.2%, respectively, for PSA density 0.12ng/ml/ml. ISUP GG≥3 (Gleason ≥4+3) was found in 2.4% (15/617) of men with MRI scores 1-2. They key limitations of this study are the heterogeneity and retrospective nature of the data. CONCLUSIONS: Multiparametric MRI when used in real-world settings is able to rule out csPCa accurately, suggesting that about one-third of men might avoid an immediate biopsy. Men should be counselled about the risk of missing some significant cancers. PATIENT SUMMARY: Multiparametric magnetic resonance imaging (MRI) is a useful tool for ruling out prostate cancer, especially when combined with prostate-specific antigen density (PSAD). Previous results published from specialist centres can be reproduced at smaller institutions. However, patients and their clinicians must be aware that an early diagnosis of clinically significant prostate cancer could be missed in nearly 10% of patients by relying on MRI and PSAD alone.
