ABSTRACT
Automatic segmentation of histopathology whole-slide images (WSI) usually involves supervised training of deep learning models with pixel-level labels to classify each pixel of the WSI into tissue regions such as benign or cancerous. However, fully supervised segmentation requires large-scale data manually annotated by experts, which can be expensive and time-consuming to obtain. Non-fully supervised methods, ranging from semi-supervised to unsupervised, have been proposed to address this issue and have been successful in WSI segmentation tasks. But these methods have mainly been focused on technical advancements in algorithmic performance rather than on the development of practical tools that could be used by pathologists or researchers in real-world scenarios. In contrast, we present DEPICTER (Deep rEPresentatIon ClusTERing), an interactive segmentation tool for histopathology annotation that produces a patch-wise dense segmentation map at WSI level. The interactive nature of DEPICTER leverages self- and semi-supervised learning approaches to allow the user to participate in the segmentation producing reliable results while reducing the workload. DEPICTER consists of three steps: first, a pretrained model is used to compute embeddings from image patches. Next, the user selects a number of benign and cancerous patches from the multi-resolution image. Finally, guided by the deep representations, label propagation is achieved using our novel seeded iterative clustering method or by directly interacting with the embedding space via feature space gating. We report both real-time interaction results with three pathologists and evaluate the performance on three public cancer classification dataset benchmarks through simulations. The code and demos of DEPICTER are publicly available at https://github.com/eduardchelebian/depicter.
Subject(s)
Benchmarking , Supervised Machine Learning , Cluster Analysis , Workload , Image Processing, Computer-AssistedABSTRACT
Background and objectives: Spatially resolved techniques for exploring the molecular landscape of tissue samples, such as spatial transcriptomics, often result in millions of data points and images too large to view on a regular desktop computer, limiting the possibilities in visual interactive data exploration. TissUUmaps is a free, open-source browser-based tool for GPU-accelerated visualization and interactive exploration of 107+ data points overlaying tissue samples. Methods: Herein we describe how TissUUmaps 3 provides instant multiresolution image viewing and can be customized, shared, and also integrated into Jupyter Notebooks. We introduce new modules where users can visualize markers and regions, explore spatial statistics, perform quantitative analyses of tissue morphology, and assess the quality of decoding in situ transcriptomics data. Results: We show that thanks to targeted optimizations the time and cost associated with interactive data exploration were reduced, enabling TissUUmaps 3 to handle the scale of today's spatial transcriptomics methods. Conclusion: TissUUmaps 3 provides significantly improved performance for large multiplex datasets as compared to previous versions. We envision TissUUmaps to contribute to broader dissemination and flexible sharing of largescale spatial omics data.
ABSTRACT
Large-scale multiplex tissue analysis aims to understand processes such as development and tumor formation by studying the occurrence and interaction of cells in local environments in, for example, tissue samples from patient cohorts. A typical procedure in the analysis is to delineate individual cells, classify them into cell types, and analyze their spatial relationships. All steps come with a number of challenges, and to address them and identify the bottlenecks of the analysis, it is necessary to include quality control tools in the analysis workflow. This makes it possible to optimize the steps and adjust settings in order to get better and more precise results. Additionally, the development of automated approaches for tissue analysis requires visual verification to reduce skepticism with regard to the accuracy of the results. Quality control tools could be used to build users' trust in automated approaches. In this paper, we present three plugins for visualization and quality control in large-scale multiplex tissue analysis of microscopy images. The first plugin focuses on the quality of cell staining, the second one was made for interactive evaluation and comparison of different cell classification results, and the third one serves for reviewing interactions of different cell types.
ABSTRACT
Prostate cancer is a common cancer type in men, yet some of its traits are still under-explored. One reason for this is high molecular and morphological heterogeneity. The purpose of this study was to develop a method to gain new insights into the connection between morphological changes and underlying molecular patterns. We used artificial intelligence (AI) to analyze the morphology of seven hematoxylin and eosin (H&E)-stained prostatectomy slides from a patient with multi-focal prostate cancer. We also paired the slides with spatially resolved expression for thousands of genes obtained by a novel spatial transcriptomics (ST) technique. As both spaces are highly dimensional, we focused on dimensionality reduction before seeking associations between them. Consequently, we extracted morphological features from H&E images using an ensemble of pre-trained convolutional neural networks and proposed a workflow for dimensionality reduction. To summarize the ST data into genetic profiles, we used a previously proposed factor analysis. We found that the regions were automatically defined, outlined by unsupervised clustering, associated with independent manual annotations, in some cases, finding further relevant subdivisions. The morphological patterns were also correlated with molecular profiles and could predict the spatial variation of individual genes. This novel approach enables flexible unsupervised studies relating morphological and genetic heterogeneity using AI to be carried out.
ABSTRACT
INTRODUCTION: IDH1/2 wt glioblastoma (GB) represents the most lethal tumour of the central nervous system. Tumour vascularity is associated with overall survival (OS), and the clinical relevance of vascular markers, such as rCBV, has already been validated. Nevertheless, molecular and clinical factors may have different influences on the beneficial effect of a favourable vascular signature. PURPOSE: To evaluate the association between the rCBV and OS of IDH1/2 wt GB patients for long-term survivors (LTSs) and short-term survivors (STSs). Given that initial high rCBV may affect the patient's OS in follow-up stages, we will assess whether a moderate vascularity is beneficial for OS in both groups of patients. MATERIALS AND METHODS: Ninety-nine IDH1/2 wt GB patients were divided into LTSs (OS ≥ 400 days) and STSs (OS < 400 days). Mann-Whitney and Fisher, uni- and multiparametric Cox, Aalen's additive regression and Kaplan-Meier tests were carried out. Tumour vascularity was represented by the mean rCBV of the high angiogenic tumour (HAT) habitat computed through the haemodynamic tissue signature methodology (available on the ONCOhabitats platform). RESULTS: For LTSs, we found a significant association between a moderate value of rCBVmean and higher OS (uni- and multiparametric Cox and Aalen's regression) (p = 0.0140, HR = 1.19; p = 0.0085, HR = 1.22) and significant stratification capability (p = 0.0343). For the STS group, no association between rCBVmean and survival was observed. Moreover, no significant differences (p > 0.05) in gender, age, resection status, chemoradiation, or MGMT methylation were observed between LTSs and STSs. CONCLUSION: We have found different prognostic and stratification effects of the vascular marker for the LTS and STS groups. We propose the use of rCBVmean at HAT as a vascular marker clinically relevant for LTSs with IDH1/2 wt GB and maybe as a potential target for randomized clinical trials focused on this group of patients.
Subject(s)
Brain Neoplasms/blood supply , Cancer Survivors , Glioblastoma/blood supply , Isocitrate Dehydrogenase/genetics , Blood Volume , Brain Neoplasms/genetics , Brain Neoplasms/mortality , Cerebrovascular Circulation , DNA Modification Methylases/genetics , DNA Repair Enzymes/genetics , Female , Glioblastoma/genetics , Glioblastoma/mortality , Humans , Male , Middle Aged , Proportional Hazards Models , Tumor Suppressor Proteins/geneticsABSTRACT
OBJECTIVES: To assess the combined role of tumor vascularity, estimated from perfusion MRI, and MGMT methylation status on overall survival (OS) in patients with glioblastoma. METHODS: A multicentric international dataset including 96 patients from NCT03439332 clinical study were used to study the prognostic relationships between MGMT and perfusion markers. Relative cerebral blood volume (rCBV) in the most vascularized tumor regions was automatically obtained from preoperative MRIs using ONCOhabitats online analysis service. Cox survival regression models and stratification strategies were conducted to define a subpopulation that is particularly favored by MGMT methylation in terms of OS. RESULTS: rCBV distributions did not differ significantly (p > 0.05) in the methylated and the non-methylated subpopulations. In patients with moderately vascularized tumors (rCBV < 10.73), MGMT methylation was a positive predictive factor for OS (HR = 2.73, p = 0.003, AUC = 0.70). In patients with highly vascularized tumors (rCBV > 10.73), however, there was no significant effect of MGMT methylation (HR = 1.72, p = 0.10, AUC = 0.56). CONCLUSIONS: Our results indicate the existence of complementary prognostic information provided by MGMT methylation and rCBV. Perfusion markers could identify a subpopulation of patients who will benefit the most from MGMT methylation. Not considering this information may lead to bias in the interpretation of clinical studies. KEY POINTS: ⢠MRI perfusion provides complementary prognostic information to MGMT methylation. ⢠MGMT methylation improves prognosis in glioblastoma patients with moderate vascular profile. ⢠Failure to consider these relations may lead to bias in the interpretation of clinical studies.
Subject(s)
Brain Neoplasms , Glioblastoma , Antineoplastic Agents, Alkylating/therapeutic use , Brain Neoplasms/diagnostic imaging , Brain Neoplasms/drug therapy , Brain Neoplasms/genetics , DNA Methylation , DNA Modification Methylases/genetics , DNA Repair Enzymes/genetics , Glioblastoma/diagnostic imaging , Glioblastoma/genetics , Humans , Prognosis , Promoter Regions, Genetic , Temozolomide/therapeutic use , Tumor Suppressor Proteins/geneticsABSTRACT
BACKGROUND AND PURPOSE: Genetic classifications are crucial for understanding the heterogeneity of glioblastoma. Recently, perfusion MRI techniques have demonstrated associations molecular alterations. In this work, we investigated whether perfusion markers within infiltrated peripheral edema were associated with proneural, mesenchymal, classical and neural subtypes. MATERIALS AND METHODS: ONCOhabitats open web services were used to obtain the cerebral blood volume at the infiltrated peripheral edema for MRI studies of 50 glioblastoma patients from The Cancer Imaging Archive: TCGA-GBM. ANOVA and Kruskal-Wallis tests were carried out in order to assess the association between vascular features and the Verhaak subtypes. For assessing specific differences, Mann-Whitney U-test was conducted. Finally, the association of overall survival with molecular and vascular features was assessed using univariate and multivariate Cox models. RESULTS: ANOVA and Kruskal-Wallis tests for the maximum cerebral blood volume at the infiltrated peripheral edema between the four subclasses yielded false discovery rate corrected p-values of <0.001 and 0.02, respectively. This vascular feature was significantly higher (p = 0.0043) in proneural patients compared to the rest of the subtypes while conducting Mann-Whitney U-test. The multivariate Cox model pointed to redundant information provided by vascular features at the peripheral edema and proneural subtype when analyzing overall survival. CONCLUSIONS: Higher relative cerebral blood volume at infiltrated peripheral edema is associated with proneural glioblastoma subtype suggesting underlying vascular behavior related to molecular composition in that area.
Subject(s)
Brain Edema/physiopathology , Brain Neoplasms/diagnostic imaging , Glioblastoma/diagnostic imaging , Magnetic Resonance Angiography/methods , Adolescent , Adult , Aged , Aged, 80 and over , Brain Edema/diagnostic imaging , Brain Neoplasms/physiopathology , Cerebral Blood Volume , Female , Glioblastoma/physiopathology , Humans , Male , Middle Aged , Retrospective Studies , Survival Analysis , Young AdultABSTRACT
BACKGROUND: Glioblastoma (GBM) is the most aggressive primary brain tumor, characterized by a heterogeneous and abnormal vascularity. Subtypes of vascular habitats within the tumor and edema can be distinguished: high angiogenic tumor (HAT), low angiogenic tumor (LAT), infiltrated peripheral edema (IPE), and vasogenic peripheral edema (VPE). PURPOSE: To validate the association between hemodynamic markers from vascular habitats and overall survival (OS) in glioblastoma patients, considering the intercenter variability of acquisition protocols. STUDY TYPE: Multicenter retrospective study. POPULATION: In all, 184 glioblastoma patients from seven European centers participating in the NCT03439332 clinical study. FIELD STRENGTH/SEQUENCE: 1.5T (for 54 patients) or 3.0T (for 130 patients). Pregadolinium and postgadolinium-based contrast agent-enhanced T1 -weighted MRI, T2 - and FLAIR T2 -weighted, and dynamic susceptibility contrast (DSC) T2 * perfusion. ASSESSMENT: We analyzed preoperative MRIs to establish the association between the maximum relative cerebral blood volume (rCBVmax ) at each habitat with OS. Moreover, the stratification capabilities of the markers to divide patients into "vascular" groups were tested. The variability in the markers between individual centers was also assessed. STATISTICAL TESTS: Uniparametric Cox regression; Kaplan-Meier test; Mann-Whitney test. RESULTS: The rCBVmax derived from the HAT, LAT, and IPE habitats were significantly associated with patient OS (P < 0.05; hazard ratio [HR]: 1.05, 1.11, 1.28, respectively). Moreover, these markers can stratify patients into "moderate-" and "high-vascular" groups (P < 0.05). The Mann-Whitney test did not find significant differences among most of the centers in markers (HAT: P = 0.02-0.685; LAT: P = 0.010-0.769; IPE: P = 0.093-0.939; VPE: P = 0.016-1.000). DATA CONCLUSION: The rCBVmax calculated in HAT, LAT, and IPE habitats have been validated as clinically relevant prognostic biomarkers for glioblastoma patients in the pretreatment stage. This study demonstrates the robustness of the hemodynamic tissue signature (HTS) habitats to assess the GBM vascular heterogeneity and their association with patient prognosis independently of intercenter variability. LEVEL OF EVIDENCE: 3 Technical Efficacy Stage: 2 J. Magn. Reson. Imaging 2020;51:1478-1486.