NIMG-44. INTEGRATING AUTOMATED LESION SEGMENTATIONS FROM SINGLE-IMAGES INTO ROUTINE CLINICAL WORKFLOW FOR VOLUMETRIC RESPONSE ASSESSMENT

Abstract INTRODUCTION Volume calculations have not been adopted into glioma response assessment due to lengthy times for manual definition and unreliable measures provided by automated algorithms. Relatively new artificial intelligence approaches such as convolutional neural networks have significantly improved lesion segmentation with performance accuracies >90%. However, their adoption into routine practice remains limited due to poor generalizability and failure rates approaching 25% when incorporated into clinical workflow. The latter can be attributed to 1) the requirement of four different types of anatomic images (T2, T2-FLAIR, T1 pre- and post-contrast); 2) cumbersome preprocessing including alignment, reformatting, and skull removal; and 3) the lack of a well-integrated clinical deployment system. The goal of this study was to demonstrate how simple modifications to a robust network coupled with an integrated workflow can provide reliable measures of tumor volume for real-time use in the reading room. METHODS Leveraging NVIDIA’s Clara-Train software and a molecularly diverse dataset of 400 labeled images for training, we modified a top-performing ensembled 2D-U-Net to require a single image-volume input (T2-FLAIR or post-contrast T1 for the T2-hyperintense or contrast-enhancing lesions) and deployed the results in the clinic to provide quantitative volumetrics. Inference was performed on a mix of image orientations without any reformatting or skull-stripping. RESULTS Training on only 115 of our 400 datasets, we achieved Dice Coefficients of 90% and 81% overlap of our auto-segmented T2 and contrast-enhancing lesions with manual labels in our 25-patient validation cohort (11 enhancing), compared to 91% and 83% overlap with the original model that required four anatomic images to segment each lesion. Radiologists can view segmentations directly from PACS as contours or overlays and provide numerical feedback for model refinement. The workflow has been applied on 50 cases to date without any failures and can be easily shared for deployment on any clinical PACS.

Download Full-text

Flare phenomenon in prostate cancer: recent evidence on new drugs and next generation imaging

Therapeutic Advances in Medical Oncology ◽

10.1177/1758835920987654 ◽

2021 ◽

Vol 13 ◽

pp. 175883592098765

Author(s):

Vincenza Conteduca ◽

Giulia Poti ◽

Paola Caroli ◽

Sabino Russi ◽

Nicole Brighi ◽

...

Keyword(s):

Prostate Cancer ◽

Response Assessment ◽

New Drugs ◽

Biological Mechanisms ◽

Early Discontinuation ◽

Flare Phenomenon ◽

Discontinuation Of Treatment ◽

Different Types ◽

Treatment Response Assessment ◽

New Evidence

Over the years, an increasing proportion of metastatic prostate cancer patients has been found to experience an initial bone flare phenomenon under both standard therapies (androgen deprivation therapy, chemotherapy, radiotherapy, abiraterone, enzalutamide) and novel agents (immunotherapy, bone-targeting radioisotopes). The underlying biological mechanisms of the flare phenomenon are still elusive and need further clarification, particularly in relation to different types of treatment and their treatment response assessment. Flare phenomenon is often underestimated and, in some cases, can negatively affect clinical outcome. In cases with suspected bone flare, the treatment should be continued for a minimum of 12 more weeks before further decisions about efficacy can be taken. Physicians and patients should be aware of this effect to avoid unwarranted anxiety and inadequate early discontinuation of treatment. This review aims at highlighting new evidence on flare phenomenon arising after the introduction of new drugs extending across the biochemical, radiographic and clinical spectrum of the disease.

Download Full-text

Use of radiomics for the prediction of local control of brain metastases after stereotactic radiosurgery

Neuro-Oncology ◽

10.1093/neuonc/noaa007 ◽

2020 ◽

Vol 22 (6) ◽

pp. 797-805 ◽

Cited By ~ 5

Author(s):

Andrei Mouraviev ◽

Jay Detsky ◽

Arjun Sahgal ◽

Mark Ruschin ◽

Young K Lee ◽

...

Keyword(s):

Brain Metastases ◽

Stereotactic Radiosurgery ◽

Clinical Features ◽

Characteristic Curve ◽

Response Prediction ◽

Response Assessment ◽

Local Failure ◽

Post Contrast ◽

Axial Diameter ◽

Fluid Attenuated Inversion Recovery

Abstract Background Local response prediction for brain metastases (BM) after stereotactic radiosurgery (SRS) is challenging, particularly for smaller BM, as existing criteria are based solely on unidimensional measurements. This investigation sought to determine whether radiomic features provide additional value to routinely available clinical and dosimetric variables to predict local recurrence following SRS. Methods Analyzed were 408 BM in 87 patients treated with SRS. A total of 440 radiomic features were extracted from the tumor core and the peritumoral regions, using the baseline pretreatment volumetric post-contrast T1 (T1c) and volumetric T2 fluid-attenuated inversion recovery (FLAIR) MRI sequences. Local tumor progression was determined based on Response Assessment in Neuro-Oncology‒BM criteria, with a maximum axial diameter growth of >20% on the follow-up T1c indicating local failure. The top radiomic features were determined based on resampled random forest (RF) feature importance. An RF classifier was trained using each set of features and evaluated using the area under the receiver operating characteristic curve (AUC). Results The addition of any one of the top 10 radiomic features to the set of clinical features resulted in a statistically significant (P < 0.001) increase in the AUC. An optimized combination of radiomic and clinical features resulted in a 19% higher resampled AUC (mean = 0.793; 95% CI = 0.792–0.795) than clinical features alone (0.669, 0.668–0.671). Conclusions The increase in AUC of the RF classifier, after incorporating radiomic features, suggests that quantitative characterization of tumor appearance on pretreatment T1c and FLAIR adds value to known clinical and dosimetric variables for predicting local failure.

Download Full-text

Four types of urban austerity: Public land privatisations in French and Italian cities

Urban Studies ◽

10.1177/0042098019827517 ◽

2019 ◽

Vol 57 (1) ◽

pp. 75-92 ◽

Cited By ~ 3

Author(s):

Félix Adisson ◽

Francesca Artioli

Keyword(s):

Case Studies ◽

Analytical Model ◽

Ad Hoc ◽

Public Land ◽

Routine Practice ◽

Local Policy ◽

Policy Capacity ◽

Different Types ◽

Local Patterns

This article contributes to current debates on urban austerity by comparing public land privatisations in French and Italian cities. These privatisations have emerged in several countries during the last two decades as a recurring austerity measure. However, current research does not explain how similar national austerity policies result in diverse urban outcomes. This article tackles this limitation by developing an analytical model of the different types of urban austerity. It uses the intergovernmental system and local policy capacity as the main variables to explain four local patterns of austerity, that is, gridlock austerity, nationally mitigated austerity, locally mitigated austerity and opportunistic austerity. Drawing on nine case studies covering two public landowners, the article shows that public land austerity policies have become routine practice based on compromises in French cities, but conflictual and based on ad hoc solutions in Italian cities.

Download Full-text

Treatment Planning and Volumetric Response Assessment for Yttrium-90 Radioembolization: Semiautomated Determination of Liver Volume and Volume of Tumor Necrosis in Patients with Hepatic Malignancy

CardioVascular and Interventional Radiology ◽

10.1007/s00270-010-9938-3 ◽

2010 ◽

Vol 34 (2) ◽

pp. 306-318 ◽

Cited By ~ 18

Author(s):

Wayne L. Monsky ◽

Armando S. Garza ◽

Isaac Kim ◽

Shaun Loh ◽

Tzu-Chun Lin ◽

...

Keyword(s):

Treatment Planning ◽

Tumor Necrosis ◽

Liver Volume ◽

Response Assessment ◽

Hepatic Malignancy ◽

Volumetric Response ◽

Yttrium 90

Download Full-text

Non-disruptive collagen characterization in clinical histopathology using cross-modality image synthesis

Communications Biology ◽

10.1038/s42003-020-01151-5 ◽

2020 ◽

Vol 3 (1) ◽

Cited By ~ 1

Author(s):

Adib Keikhosravi ◽

Bin Li ◽

Yuming Liu ◽

Matthew W. Conklin ◽

Agnes G. Loeffler ◽

...

Keyword(s):

Second Harmonic ◽

Image Synthesis ◽

Fibrillar Collagen ◽

Clinical Workflow ◽

Synthesis Algorithm ◽

Different Types ◽

Stromal Collagen ◽

Diagnostic Work ◽

Second Harmonic Generation Imaging ◽

Picrosirius Red

AbstractThe importance of fibrillar collagen topology and organization in disease progression and prognostication in different types of cancer has been characterized extensively in many research studies. These explorations have either used specialized imaging approaches, such as specific stains (e.g., picrosirius red), or advanced and costly imaging modalities (e.g., second harmonic generation imaging (SHG)) that are not currently in the clinical workflow. To facilitate the analysis of stromal biomarkers in clinical workflows, it would be ideal to have technical approaches that can characterize fibrillar collagen on standard H&E stained slides produced during routine diagnostic work. Here, we present a machine learning-based stromal collagen image synthesis algorithm that can be incorporated into existing H&E-based histopathology workflow. Specifically, this solution applies a convolutional neural network (CNN) directly onto clinically standard H&E bright field images to extract information about collagen fiber arrangement and alignment, without requiring additional specialized imaging stains, systems or equipment.

Download Full-text

Tumor response assessment: comparison between unstructured free text reporting in routine clinical workflow and computer-aided evaluation based on RECIST 1.1 criteria

Journal of Cancer Research and Clinical Oncology ◽

10.1007/s00432-017-2488-1 ◽

2017 ◽

Vol 143 (12) ◽

pp. 2527-2533 ◽

Cited By ~ 4

Author(s):

Juliane Goebel ◽

Julia Hoischen ◽

Carolin Gramsch ◽

Haemi P. Schemuth ◽

Andreas-Claudius Hoffmann ◽

...

Keyword(s):

Tumor Response ◽

Response Assessment ◽

Free Text ◽

Clinical Workflow ◽

Recist 1.1 ◽

Computer Aided

Download Full-text

Role of CMR Mapping Techniques in Cardiac Hypertrophic Phenotype

Diagnostics ◽

10.3390/diagnostics10100770 ◽

2020 ◽

Vol 10 (10) ◽

pp. 770

Author(s):

Andrea Baggiano ◽

Alberico Del Torto ◽

Marco Guglielmo ◽

Giuseppe Muscogiuri ◽

Laura Fusini ◽

...

Keyword(s):

Relaxation Time ◽

Tissue Characterization ◽

Myocardial Dysfunction ◽

Longitudinal Relaxation Time ◽

Clinical Workflow ◽

Post Contrast ◽

Functional Features ◽

Longitudinal Relaxation ◽

Mapping Techniques

Non-ischemic cardiomyopathies represent a heterogeneous group of myocardial diseases potentially leading to heart failure, life-threatening arrhythmias, and eventually death. Myocardial dysfunction is associated with different underlying pathological processes, ultimately inducing changes in morphological appearance. Thus, classification based on presenting morphological phenotypes has been proposed, i.e., dilated, hypertrophic, restrictive, and right ventricular cardiomyopathies. In light of the key diagnostic and prognostic role of morphological and functional features, cardiovascular imaging has emerged as key element in the clinical workflow of suspected cardiomyopathies, and above all, cardiovascular magnetic resonance (CMR) represents the ideal technique to be used: thanks to its physical principles, besides optimal spatial and temporal resolutions, incomparable contrast resolution allows to assess myocardial tissue abnormalities in detail. Traditionally, weighted images and late enhancement images after gadolinium-based contrast agent administration have been used to perform tissue characterization, but in the last decade quantitative assessment of pre-contrast longitudinal relaxation time (native T1), post-contrast longitudinal relaxation time (post-contrast T1) and transversal relaxation time (T2), all displayed with dedicated pixel-wise color-coded maps (mapping), has contributed to give precious knowledge insight, with positive influence of diagnostic accuracy and prognosis assessment, mostly in the setting of the hypertrophic phenotype. This review aims to describe the available evidence of the role of mapping techniques in the assessment of hypertrophic phenotype, and to suggest their integration in the routine CMR evaluation of newly diagnosed cardiomyopathies with increased wall thickness.

Download Full-text

Response assessment in neuro-oncology (RANO) criteria: application in clinical research and routine practice

Pharmateca ◽

10.18565/pharmateca.2021.11.21-33 ◽

2021 ◽

Vol 11_2021 ◽

pp. 21-33

Author(s):

E.A. Kobyakova Kobyakova ◽

D.Yu. Usachev Usachev ◽

O.V. Absalyamova Absalyamova ◽

N.G. Kobyakov Kobyakov ◽

K.S. Lodygina Lodygina ◽

...

Keyword(s):

Clinical Research ◽

Response Assessment ◽

Routine Practice ◽

Rano Criteria

Download Full-text

Contrast-enhanced T1-weighted subtraction maps for response assessment in recurrent glioblastoma treated with bevacizumab.

Journal of Clinical Oncology ◽

10.1200/jco.2013.31.15_suppl.2055 ◽

2013 ◽

Vol 31 (15_suppl) ◽

pp. 2055-2055 ◽

Cited By ~ 3

Author(s):

Benjamin M. Ellingson ◽

Hyun J. Kim ◽

Davis C. Woodworth ◽

Whitney B. Pope ◽

Jonathan N. Cloughesy ◽

...

Keyword(s):

Antiangiogenic Therapy ◽

Positive Contrast ◽

Response Assessment ◽

Recurrent Glioblastoma ◽

Tumor Segmentation ◽

Post Contrast ◽

Contrast Enhanced ◽

Standard Response ◽

Intensity Normalization ◽

Recurrent Gbm

2055 Background: Antiangiogenic therapy in glioblastoma (GBM) results in decreased enhancement on post-contrast T1w images, which complicates standard response assessment and is likely the reason no studies have found predictive value in enhancing tumor size or change in size. The current study examined whether contrast-enhanced T1-weighted subtraction maps (CE-ΔT1w) calculated from subtracting pre-contrast (T1) from post-contrast T1w images (T1+C) can improve quantification and predict response in GBM patients treated with bevacizumab. Methods: Recurrent GBM patients (n=160) from the BRAIN trial (AVF3708g), a multicenter Phase II trial evaluating bevacvizumab, were used in the current study. CE-ΔT1w maps were calculated 2 weeks before and 6 weeks after the first dose of bevacizumab by: 1) performing registration between T1 and T1+C images, 2) image intensity normalization of T1 and T1+C images, and 3) subtraction of T1 from T1+C images. The volume of tumor regions with positive contrast enhancement after subtraction was retained for analysis. Results: CE-ΔT1w maps greatly improved detectability of subtly enhancing lesions, particularly post-treatment. Results for PFS and OS are summarized in the table below. In all scenarios, CE-ΔT1w maps outperformed conventional tumor segmentation. Results show that size and change in size are both predictive of PFS and OS. Conclusions: CE-ΔT1w maps improve visualization and quantification of contrast enhancing tumor regions in recurrent GBM, allowing for more accurate response assessment. [Table: see text]

Download Full-text