Chen, David

The analysis and interpretation of cardiac magnetic resonance (CMR) images are often time-consuming. The automated segmentation of cardiac structures can reduce the time required for image analysis. Spatial similarities between different CMR image types were leveraged to jointly segment multiple sequences using a segmentation model termed a multi-image type UNet (MI-UNet). This model was developed from 72 exams (46% female, mean age 63 ± 11 years) performed on patients with hypertrophic cardiomyopathy. The MI-UNet for steady-state free precession (SSFP) images achieved a superior Dice similarity coefficient (DSC) of 0.92 ± 0.06 compared to 0.87 ± 0.08 for a single-image type UNet (p < 0.001). The MI-UNet for late gadolinium enhancement (LGE) images also had a superior DSC of 0.86 ± 0.11 compared to 0.78 ± 0.11 for a single-image type UNet (p = 0.001). The difference across image types was most evident for the left ventricular myocardium in SSFP images and for both the left ventricular cavity and the left ventricular myocardium in LGE images. For the right ventricle, there were no differences in DCS when comparing the MI-UNet with single-image type UNets. The joint segmentation of multiple image types increases segmentation accuracy for CMR images of the left ventricle compared to single-image models. In clinical practice, the MI-UNet model may expedite the analysis and interpretation of CMR images of multiple types.

PURPOSE Time to event is an important aspect of clinical decision making. This is particularly true when diseases have highly heterogeneous presentations and prognoses, as in chronic lymphocytic lymphoma (CLL). Although machine learning methods can readily learn complex nonlinear relationships, many methods are criticized as inadequate because of limited interpretability. We propose using unsupervised clustering of the continuous output of machine learning models to provide discrete risk stratification for predicting time to first treatment in a cohort of patients with CLL. PATIENTS AND METHODS A total of 737 treatment-naïve patients with CLL diagnosed at Mayo Clinic were included in this study. We compared predictive abilities for two survival models (Cox proportional hazards and random survival forest) and four classification methods (logistic regression, support vector machines, random forest, and gradient boosting machine). Probability of treatment was then stratified. RESULTS Machine learning methods did not yield significantly more accurate predictions of time to first treatment. However, automated risk stratification provided by clustering was able to better differentiate patients who were at risk for treatment within 1 year than models developed using standard survival analysis techniques. CONCLUSION Clustering the posterior probabilities of machine learning models provides a way to better interpret machine learning models.

560 Background: Level I evidence exists demonstrating the efficacy of the mTOR inhibitor everolimus (EVE) in decreasing tumor volume of syndromic angiomyolipomas (AMLs) among patients with Tuberous Sclerosis. No prospective data are available regarding the effect of mTOR inhibition on growth kinetics in patients with sporadic AMLs. Methods: We conducted a multi-institutional, prospective, phase 2 trial with an optimal two-stage Simon design in patients presenting with > 3cm sporadic AMLs who were candidates for surgical or percutaneous intervention. Response was defined as ≥25% volumetric reduction of the AML. Planned enrollment was 43 patients to test the null hypothesis at a 5% level of significance with 80% power. Baseline, 4- and 6-month volumetric analysis was performed by dynamic contrast-enhanced MRI (DCE-MRI). Patients received EVE 10mg for four 28-day cycles, at which point EVE was discontinued in those with < 25% volumetric reduction. Those with ≥ 25% volumetric reduction received two additional cycles of EVE. Dose reductions and interruptions were allowed to 5 mg QOD. Conservative stopping rules were established for toxicity, given the benign nature of AMLs. Results: The early stopping rules for both efficacy and toxicity were invoked. We enrolled 20 patients (median age = 68) from 5 centers with 21 sporadic AMLs. 11/20 (55%) patients completed 4 cycles of EVE, while 7/20 (35%) completed 6 cycles. Median days on treatment was 88 (2 cycles). 4/20 (20%) patients were withdrawn due to toxicity, while 8/20 (40%) withdrew due to personal preference. Dose reductions were required in 6/20 (30%) patients, and 5/20 (25%) patients had grade 3 toxicities which resolved upon discontinuation or dose reduction of EVE. At 4-month MRI, 10/16 (62.5%) patients had a ≥25% reduction in volume (mean = 54.1% decrease). At 6-month MRI, 8/12 (66.6%) patients had a ≥25% reduction in volume (mean = 51.5% decrease). Conclusions: EVE was effective in reducing tumor volume in patients with sporadic AMLs but was associated with a high rate of treatment termination due to patient preference or prespecified AEs. Neoadjuvant EVE may be useful in potentiating surgical resection of large or anatomically complex AMLs. Clinical trial information: NCT02539459.