Tarabishy, Abdul

The translation of AI-generated brain metastases (BM) segmentation into clinical practice relies heavily on diverse, high-quality annotated medical imaging datasets. The BraTS-METS 2023 challenge has gained momentum for testing and benchmarking algorithms using rigorously annotated internationally compiled real-world datasets. This study presents the results of the segmentation challenge and characterizes the challenging cases that impacted the performance of the winning algorithms. Untreated brain metastases on standard anatomic MRI sequences (T1, T2, FLAIR, T1PG) from eight contributed international datasets were annotated in stepwise method: published UNET algorithms, student, neuroradiologist, final approver neuroradiologist. Segmentations were ranked based on lesion-wise Dice and Hausdorff distance (HD95) scores. False positives (FP) and false negatives (FN) were rigorously penalized, receiving a score of 0 for Dice and a fixed penalty of 374 for HD95. The mean scores for the teams were calculated. Eight datasets comprising 1303 studies were annotated, with 402 studies (3076 lesions) released on Synapse as publicly available datasets to challenge competitors. Additionally, 31 studies (139 lesions) were held out for validation, and 59 studies (218 lesions) were used for testing. Segmentation accuracy was measured as rank across subjects, with the winning team achieving a LesionWise mean score of 7.9. The Dice score for the winning team was 0.65 ± 0.25. Common errors among the leading teams included false negatives for small lesions and misregistration of masks in space. The Dice scores and lesion detection rates of all algorithms diminished with decreasing tumor size, particularly for tumors smaller than 100 mm3. In conclusion, algorithms for BM segmentation require further refinement to balance high sensitivity in lesion detection with the minimization of false positives and negatives. The BraTS-METS 2023 challenge successfully curated well- annotated, diverse datasets and identified common errors, facilitating the translation of BM segmentation across varied environments and providing the tools for future development of personalized volumetric reports to patients undergoing BM treatment.

Introduction: Heart failure with low ejection fraction (EF) is a known risk factor for stroke. Low EF is associated with increased risk of thrombus formation and is accompanied with 2 to 3 fold increased risk of stroke. This can happen even in the absence of atrial fibrillation. Hypothesis: Our hypothesis was that size of the stroke will be larger in patients with reduction in EF as compared with cardioembolic strokes in preserved EF. The worse the EF, the more stagnation of blood and poor cardiac output leading higher risk of thrombus formation. It’s also likely that the group with reduced EF form a larger thrombus leading to larger stroke as compared to the group with preserved EF. Methods: For our analysis, we retrospectively reviewed charts for 49 acute ischemic stroke patients without atrial fibrillation. We used transthoracic echocardiogram to determine the EF. There are 25 patients with preserved EF, cutoff > 50%. There were 24 patients in reduced EF group. Both groups were matched for age and cardiovascular risk factors. Infarct volume was manually calculated from T1 MRI using BrainLab software. Results: There is an inverse correlation between EF and infarct volume (r=-.283, p=.048) meaning patients with reduced EF had greater infarct volume. Median infarct volume is higher in reduced EF group (median volume = 36 cm 3 ) compared to patients with preserved EF (median volume = 11 cm 3 ) (p=.117). Conclusions: In our patient’s sample, there seems to be an inverse correlation between EF and size of stroke. Patients with reduced EF were associated with larger strokes. Larger studies need to be performed to establish a correlation.