PDF file - 5.2MB, Figure S1: Decreased Flow. Representative example of patient with flow decrease. (A) Anatomic MR imaging showing decrease in the contrast enhanced tumor area. (B) Flow maps showing decreasing flow. The blue ovals indicate region of tumor. (C) Histogram analysis of enhancing tumor showing decrease of flow compared to reference tissue.

Supplementary Figure 2 from Glioblastoma Recurrence after Cediranib Therapy in Patients: Lack of “Rebound” Revascularization as Mode of Escape

Supplementary Table 1 from Glioblastoma Recurrence after Cediranib Therapy in Patients: Lack of “Rebound” Revascularization as Mode of Escape

Objective Functional and morphologic changes in extracranial organs can occur after acute brain injury. The neuroanatomic correlates of such changes are not fully known. Herein, we tested the hypothesis that brain infarcts are associated with cardiac and systemic abnormalities (CSAs) in a regionally specific manner. Methods We generated voxelwise p value maps of brain infarcts for poststroke plasma cardiac troponin T (cTnT) elevation, QTc prolongation, in‐hospital infection, and acute stress hyperglycemia (ASH) in 1,208 acute ischemic stroke patients prospectively recruited into the Heart–Brain Interactions Study. We examined the relationship between infarct location and CSAs using a permutation‐based approach and identified clusters of contiguous voxels associated with p < 0.05. Results cTnT elevation not attributable to a known cardiac reason was detected in 5.5%, QTc prolongation in the absence of a known provoker in 21.2%, ASH in 33.9%, and poststroke infection in 13.6%. We identified significant, spatially segregated voxel clusters for each CSA. The clusters for troponin elevation and QTc prolongation mapped to the right hemisphere. There were 3 clusters for ASH, the largest of which was in the left hemisphere. We found 2 clusters for poststroke infection, one associated with pneumonia in the left and one with urinary tract infection in the right hemisphere. The relationship between infarct location and CSAs persisted after adjusting for infarct volume. Interpretation Our results show that there are discrete regions of brain infarcts associated with CSAs. This information could be used to bootstrap toward new markers for better differentiation between neurogenic and non‐neurogenic mechanisms of poststroke CSAs. ANN NEUROL 2023;94:1155–1163

Supplementary Appendix, Table 1, Figure Legends 1-4 from Serial Magnetic Resonance Spectroscopy Reveals a Direct Metabolic Effect of Cediranib in Glioblastoma

Supplementary Table 1 from A “Vascular Normalization Index” as Potential Mechanistic Biomarker to Predict Survival after a Single Dose of Cediranib in Recurrent Glioblastoma Patients

PDF file - 680K, Figure S6: Reproducibility Analysis. Bland-Altman plots showing (A) the test-retest variability of measurement of microvessel flow (Pearson correlation; =1.00) and (B) the variability of flow measurement between the day -5 scan and the day -1 scan (two baselines) (=0.93).

Supplementary Figures 1-4 from A “Vascular Normalization Index” as Potential Mechanistic Biomarker to Predict Survival after a Single Dose of Cediranib in Recurrent Glioblastoma Patients

PDF file - 382K, Figure S7: Relationship Between Increased Microvessel Flow and Metabolic Status of Tumors. Averaged changes in MRS ratios for patients with an increase in microvessel flow (n=5; 2 patients did not have MRS data). Relative to pretreatment values, metabolic ratios for NAA/norCre were significantly higher at days +28, +56 and at day +28 for Cho/norCre (*Wilcoxon signed-rank; P<0.05). After day +56, the observed anti-tumor response was reversed. Numerical data show log-scaled averaged values (SEM) and values at day -1 were set as 100% in all lesions.

Supplemental Table 1: Participating centers that enrolled patients; Supplemental Table 2: Serial FMISO PET Imaging.

PDF file - 1.3MB, Figure S4: Macrovessel Flow and Response to Treatment. (A) Similar to microvessel flow, analysis of total (macrovessel) flow show three types of response to anti-angiogenic treatment: increase in flow (6 patients), stable flow (13 patients) or decrease in flow (11 patients). (B) Applying the patients groups from microvessel tumor flow analysis on the macrovessel tumor flow data. Figures show log-scaled averaged values (SEM) and the day -1 values were set as 100% in all lesions. P-values show results of Kruskal-Wallis tests for the difference between the three patient groups at each study day (Holm-Bonferroni corrected).

PDF file - 894K, Figure S5: Arterial Spin Labeling (ASL) Flow and Response to Treatment. (A) Analysis of blood flow as measured by arterial spin labeling (ASL) when applying the patients groups from the macrovessel tumor flow analysis. (B) ASL blood flow when applying the patients groups from the microvessel tumor flow analysis. Figures show log-scaled averaged values (SEM) and the day -1 values were set as 100% in all lesions. P-values show results of Mann-Whitney tests for the difference between patients with increase in flow and stable or decreased flow (Holm-Bonferroni corrected).

C. Catana, T. Benner, A. van der Kouwe, D. L. Jennings, M. Hamm, P-J. Chen, O. C. Andronesi, E. R. Gerstner, L. Byars, C. Michel, J. Pfeuffer, M. Schmand, B. R. Rosen, and A. G. Sorensen MGH, Radiology, A.A. Martinos Center for Biomedical Imaging, Charlestown, MA, United States, Siemens Medical Solutions USA Inc., Charlestown, MA, United States, Nuclear Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA, United States, Massachusetts General Hospital, Boston, MA, United States, Siemens Medical Solutions USA Inc., Knoxville, TN, United States

Cover image supplementary movie: added 7-15-09 from A “Vascular Normalization Index” as Potential Mechanistic Biomarker to Predict Survival after a Single Dose of Cediranib in Recurrent Glioblastoma Patients

PDF file - 1.7MB, Figure S2: Changes in Imaging Parameters over Time. Based on the three patients groups from the microvessel tumor flow analysis, the plots show: (A) contrast-enhanced T1-weighted tumor volumes, (B) FLAIR tumor volumes and (C) permeability (Ktrans) of the tumor over time. Patients with an increase and decrease in flow showed the same changes in contrast-enhanced tumor volume and peritumoral vasogenic edema during treatment. Indeed, redefining PFS using the Response Assessment in Neuro-Oncology (RANO) Working Group criteria instead of Macdonald did not change the outcome of any test in our study. Numerical data show log-scaled averaged values (SEM) and values at day -1 were set as 100% in all lesions. P-values show results of Mann-Whitney tests for the difference between patients with stable flow and increased or decreased flow (Holm-Bonferroni corrected).

Supplementary Methods, Figures 1-4 from Serial Magnetic Resonance Spectroscopy Reveals a Direct Metabolic Effect of Cediranib in Glioblastoma

PDF file - 1MB, Figure S3: Individual Flow Data. Plots showing patients with (A) increased (n=7), (B) stable (n=12) and (C) decreased (n=11) normalized tumor flow after anti-angiogenic treatment onset. Compared to both baseline values, patients with an increase (decrease) in flow showed elevated (decreased) flow values at a minimum of two consecutive time points after treatment onset. Tumor flow equal to reference tissue was set as 100%. By pair-wise analysis, at day +1 only, patients with a decrease in flow had significantly lower absolute flow in tumor compared to reference tissue (Wilcoxon signed-rank; P<0.01, Holm-Bonferroni corrected). The lack of a consistent significant difference between absolute tumor flow and reference tissue flow indicates that vascular changes occur not only in the tumor, but also in surrounding areas. Anti-angiogenic treatment effects on global flow and the observed variations in baseline flow between patients warrant further studies.

ImportanceWith a shortfall in fellowship-trained breast radiologists, mammography screening programs are looking toward artificial intelligence (AI) to increase efficiency and diagnostic accuracy. External validation studies provide an initial assessment of how promising AI algorithms perform in different practice settings.ObjectiveTo externally validate an ensemble deep-learning model using data from a high-volume, distributed screening program of an academic health system with a diverse patient population.Design, Setting, and ParticipantsIn this diagnostic study, an ensemble learning method, which reweights outputs of the 11 highest-performing individual AI models from the Digital Mammography Dialogue on Reverse Engineering Assessment and Methods (DREAM) Mammography Challenge, was used to predict the cancer status of an individual using a standard set of screening mammography images. This study was conducted using retrospective patient data collected between 2010 and 2020 from women aged 40 years and older who underwent a routine breast screening examination and participated in the Athena Breast Health Network at the University of California, Los Angeles (UCLA).Main Outcomes and MeasuresPerformance of the challenge ensemble method (CEM) and the CEM combined with radiologist assessment (CEM+R) were compared with diagnosed ductal carcinoma in situ and invasive cancers within a year of the screening examination using performance metrics, such as sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC).ResultsEvaluated on 37 317 examinations from 26 817 women (mean [SD] age, 58.4 [11.5] years), individual model AUROC estimates ranged from 0.77 (95% CI, 0.75-0.79) to 0.83 (95% CI, 0.81-0.85). The CEM model achieved an AUROC of 0.85 (95% CI, 0.84-0.87) in the UCLA cohort, lower than the performance achieved in the Kaiser Permanente Washington (AUROC, 0.90) and Karolinska Institute (AUROC, 0.92) cohorts. The CEM+R model achieved a sensitivity (0.813 [95% CI, 0.781-0.843] vs 0.826 [95% CI, 0.795-0.856]; P = .20) and specificity (0.925 [95% CI, 0.916-0.934] vs 0.930 [95% CI, 0.929-0.932]; P = .18) similar to the radiologist performance. The CEM+R model had significantly lower sensitivity (0.596 [95% CI, 0.466-0.717] vs 0.850 [95% CI, 0.766-0.923]; P < .001) and specificity (0.803 [95% CI, 0.734-0.861] vs 0.945 [95% CI, 0.936-0.954]; P < .001) than the radiologist in women with a prior history of breast cancer and Hispanic women (0.894 [95% CI, 0.873-0.910] vs 0.926 [95% CI, 0.919-0.933]; P = .004).Conclusions and RelevanceThis study found that the high performance of an ensemble deep-learning model for automated screening mammography interpretation did not generalize to a more diverse screening cohort, suggesting that the model experienced underspecification. This study suggests the need for model transparency and fine-tuning of AI models for specific target populations prior to their clinical adoption.

BACKGROUND:Artificial intelligence (AI) may improve cancer detection and risk prediction during mammography screening, but radiologists' preferences regarding its characteristics and implementation are unknown.PURPOSE:To quantify how different attributes of AI-based cancer detection and risk prediction tools affect radiologists' intentions to use AI during screening mammography interpretation.MATERIALS AND METHODS:Through qualitative interviews with radiologists, we identified five primary attributes for AI-based breast cancer detection and four for breast cancer risk prediction. We developed a discrete choice experiment based on these attributes and invited 150 US-based radiologists to participate. Each respondent made eight choices for each tool between three alternatives: two hypothetical AI-based tools versus screening without AI. We analyzed samplewide preferences using random parameters logit models and identified subgroups with latent class models.RESULTS:Respondents (n = 66; 44% response rate) were from six diverse practice settings across eight states. Radiologists were more interested in AI for cancer detection when sensitivity and specificity were balanced (94% sensitivity with <25% of examinations marked) and AI markup appeared at the end of the hanging protocol after radiologists complete their independent review. For AI-based risk prediction, radiologists preferred AI models using both mammography images and clinical data. Overall, 46% to 60% intended to adopt any of the AI tools presented in the study; 26% to 33% approached AI enthusiastically but were deterred if the features did not align with their preferences.CONCLUSION:Although most radiologists want to use AI-based decision support, short-term uptake may be maximized by implementing tools that meet the preferences of dissuadable users.