Deep neural network to locate and segment brain tumors outperformed the expert technicians who created the training data

Training and Comparison of nnU-Net and DeepMedic Methods for Autosegmentation of Pediatric Brain Tumors

BACKGROUND AND PURPOSE

Tumor segmentation is essential in surgical and treatment planning and response assessment and monitoring in pediatric brain tumors, the leading cause of cancer-related death among children. However, manual segmentation is time-consuming and has high interoperator variability, underscoring the need for more efficient methods. After training, we compared 2 deep-learning-based 3D segmentation models, DeepMedic and nnU-Net, with pediatric-specific multi-institutional brain tumor data based on multiparametric MR images.

MATERIALS AND METHODS

Multiparametric preoperative MR imaging scans of 339 pediatric patients (n = 293 internal and n = 46 external cohorts) with a variety of tumor subtypes were preprocessed and manually segmented into 4 tumor subregions, ie, enhancing tumor, nonenhancing tumor, cystic components, and peritumoral edema. After training, performances of the 2 models on internal and external test sets were evaluated with reference to ground truth manual segmentations. Additionally, concordance was assessed by comparing the volume of the subregions as a percentage of the whole tumor between model predictions and ground truth segmentations using the Pearson or Spearman correlation coefficients and the Bland-Altman method.

RESULTS

The mean Dice score for nnU-Net internal test set was 0.9 (SD, 0.07) (median, 0.94) for whole tumor; 0.77 (SD, 0.29) for enhancing tumor; 0.66 (SD, 0.32) for nonenhancing tumor; 0.71 (SD, 0.33) for cystic components, and 0.71 (SD, 0.40) for peritumoral edema, respectively. For DeepMedic, the mean Dice scores were 0.82 (SD, 0.16) for whole tumor; 0.66 (SD, 0.32) for enhancing tumor; 0.48 (SD, 0.27) for nonenhancing tumor; 0.48 (SD, 0.36) for cystic components, and 0.19 (SD, 0.33) for peritumoral edema, respectively. Dice scores were significantly higher for nnU-Net (P ≤ .01). Correlation coefficients for tumor subregion percentage volumes were higher (0.98 versus 0.91 for enhancing tumor, 0.97 versus 0.75 for nonenhancing tumor, 0.98 versus 0.80 for cystic components, 0.95 versus 0.33 for peritumoral edema in the internal test set). Bland-Altman plots were better for nnU-Net compared with DeepMedic. External validation of the trained nnU-Net model on the multi-institutional Brain Tumor Segmentation Challenge in Pediatrics (BraTS-PEDs) 2023 data set revealed high generalization capability in the segmentation of whole tumor, tumor core (a combination of enhancing tumor, nonenhancing tumor, and cystic components), and enhancing tumor with mean Dice scores of 0.87 (SD, 0.13) (median, 0.91), 0.83 (SD, 0.18) (median, 0.89), and 0.48 (SD, 0.38) (median, 0.58), respectively.

CONCLUSIONS

The pediatric-specific data-trained nnU-Net model is superior to DeepMedic for whole tumor and subregion segmentation of pediatric brain tumors.

Adversarial Learning for MRI Reconstruction and Classification of Cognitively Impaired Individuals

Game theory-inspired deep learning using a generative adversarial network provides an environment to competitively interact and accomplish a goal. In the context of medical imaging, most work has focused on achieving single tasks such as improving image resolution, segmenting images, and correcting motion artifacts. We developed a dual-objective adversarial learning framework that simultaneously 1) reconstructs higher quality brain magnetic resonance images (MRIs) that 2) retain disease-specific imaging features critical for predicting progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD). We obtained 3-Tesla, T1-weighted brain MRIs of participants from the Alzheimer's Disease Neuroimaging Initiative (ADNI, N=342) and the National Alzheimer's Coordinating Center (NACC, N = 190) datasets. We simulated MRIs with missing data by removing 50% of sagittal slices from the original scans (i.e., diced scans). The generator was trained to reconstruct brain MRIs using the diced scans as input. We introduced a classifier into the GAN architecture to discriminate between stable (i.e., sMCI) and progressive MCI (i.e., pMCI) based on the generated images to facilitate encoding of disease-related information during reconstruction. The framework was trained using ADNI data and externally validated on NACC data. In the NACC cohort, generated images had better image quality than the diced scans (Structural similarity (SSIM) index: 0.553 ± 0.116 versus 0.348 ± 0.108). Furthermore, a classifier utilizing the generated images distinguished pMCI from sMCI more accurately than with the diced scans (F1-score: 0.634 ± 0.019 versus 0.573 ± 0.028). Competitive deep learning has potential to facilitate disease-oriented image reconstruction in those at risk of developing Alzheimer's disease.

Towards Longitudinal Glioma Segmentation: Evaluating combined pre- and post-treatment MRI training data for automated tumor segmentation using nnU-Net

Identification of key phenotypic regions such as necrosis, contrast enhancement, and edema on magnetic resonance imaging (MRI) is important for understanding disease evolution and treatment response in patients with glioma. Manual delineation is time intensive and not feasible for a clinical workflow. Automating phenotypic region segmentation overcomes many issues with manual segmentation, however, current glioma segmentation datasets focus on pre-treatment, diagnostic scans, where treatment effects and surgical cavities are not present. Thus, existing automatic segmentation models are not applicable to post-treatment imaging that is used for longitudinal evaluation of care. Here, we present a comparison of three-dimensional convolutional neural networks (nnU-Net architecture) trained on large temporally defined pre-treatment, post-treatment, and mixed cohorts. We used a total of 1563 imaging timepoints from 854 patients curated from 13 different institutions as well as diverse public data sets to understand the capabilities and limitations of automatic segmentation on glioma images with different phenotypic and treatment appearance. We assessed the performance of models using Dice coefficients on test cases from each group comparing predictions with manual segmentations generated by trained technicians. We demonstrate that training a combined model can be as effective as models trained on just one temporal group. The results highlight the importance of a diverse training set, that includes images from the course of disease and with effects from treatment, in the creation of a model that can accurately segment glioma MRIs at multiple treatment time points.

Federated learning enables big data for rare cancer boundary detection

Although machine learning (ML) has shown promise across disciplines, out-of-sample generalizability is concerning. This is currently addressed by sharing multi-site data, but such centralization is challenging/infeasible to scale due to various limitations. Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates. Here we present the largest FL study to-date, involving data from 71 sites across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, reporting the largest such dataset in the literature (n = 6, 314). We demonstrate a 33% delineation improvement for the surgically targetable tumor, and 23% for the complete tumor extent, over a publicly trained model. We anticipate our study to: 1) enable more healthcare studies informed by large diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further analyses for glioblastoma by releasing our consensus model, and 3) demonstrate the FL effectiveness at such scale and task-complexity as a paradigm shift for multi-site collaborations, alleviating the need for data-sharing.

Automated brain tumour identification using magnetic resonance imaging: a systematic review and meta-analysis

Background

Automated brain tumor identification facilitates diagnosis and treatment planning. We evaluate the performance of traditional machine learning (TML) and deep learning (DL) in brain tumor detection and segmentation, using MRI.

Methods

A systematic literature search from January 2000 to May 8, 2021 was conducted. Study quality was assessed using the Checklist for Artificial Intelligence in Medical Imaging (CLAIM). Detection meta-analysis was performed using a unified hierarchical model. Segmentation studies were evaluated using a random effects model. Sensitivity analysis was performed for externally validated studies.

Results

Of 224 studies included in the systematic review, 46 segmentation and 38 detection studies were eligible for meta-analysis. In detection, DL achieved a lower false positive rate compared to TML; 0.018 (95% CI, 0.011 to 0.028) and 0.048 (0.032 to 0.072) (P < .001), respectively. In segmentation, DL had a higher dice similarity coefficient (DSC), particularly for tumor core (TC); 0.80 (0.77 to 0.83) and 0.63 (0.56 to 0.71) (P < .001), persisting on sensitivity analysis. Both manual and automated whole tumor (WT) segmentation had “good” (DSC ≥ 0.70) performance. Manual TC segmentation was superior to automated; 0.78 (0.69 to 0.86) and 0.64 (0.53 to 0.74) (P = .014), respectively. Only 30% of studies reported external validation.

Conclusions

The comparable performance of automated to manual WT segmentation supports its integration into clinical practice. However, manual outperformance for sub-compartmental segmentation highlights the need for further development of automated methods in this area. Compared to TML, DL provided superior performance for detection and sub-compartmental segmentation. Improvements in the quality and design of studies, including external validation, are required for the interpretability and generalizability of automated models.

An artificial intelligence framework and its bias for brain tumor segmentation: A narrative review

BACKGROUND

Artificial intelligence (AI) has become a prominent technique for medical diagnosis and represents an essential role in detecting brain tumors. Although AI-based models are widely used in brain lesion segmentation (BLS), understanding their effectiveness is challenging due to their complexity and diversity. Several reviews on brain tumor segmentation are available, but none of them describe a link between the threats due to risk-of-bias (RoB) in AI and its architectures. In our review, we focused on linking RoB and different AI-based architectural Cluster in popular DL framework. Further, due to variance in these designs and input data types in medical imaging, it is necessary to present a narrative review considering all facets of BLS.

APPROACH

The proposed study uses a PRISMA strategy based on 75 relevant studies found by searching PubMed, Scopus, and Google Scholar. Based on the architectural evolution, DL studies were subsequently categorized into four classes: convolutional neural network (CNN)-based, encoder-decoder (ED)-based, transfer learning (TL)-based, and hybrid DL (HDL)-based architectures. These studies were then analyzed considering 32 AI attributes, with clusters including AI architecture, imaging modalities, hyper-parameters, performance evaluation metrics, and clinical evaluation. Then, after these studies were scored for all attributes, a composite score was computed, normalized, and ranked. Thereafter, a bias cutoff (AP(ai)Bias 1.0, AtheroPoint, Roseville, CA, USA) was established to detect low-, moderate- and high-bias studies.

CONCLUSION

The four classes of architectures, from best-to worst-performing, are TL > ED > CNN > HDL. ED-based models had the lowest AI bias for BLS. This study presents a set of three primary and six secondary recommendations for lowering the RoB.

Development and Practical Implementation of a Deep Learning-Based Pipeline for Automated Pre- and Postoperative Glioma Segmentation

BACKGROUND AND PURPOSE

Quantitative volumetric segmentation of gliomas has important implications for diagnosis, treatment, and prognosis. We present a deep-learning model that accommodates automated preoperative and postoperative glioma segmentation with a pipeline for clinical implementation. Developed and engineered in concert, the work seeks to accelerate clinical realization of such tools.

MATERIALS AND METHODS

A deep learning model, autoencoder regularization-cascaded anisotropic, was developed, trained, and tested fusing key elements of autoencoder regularization with a cascaded anisotropic convolutional neural network. We constructed a dataset consisting of 437 cases with 40 cases reserved as a held-out test and the remainder split 80:20 for training and validation. We performed data augmentation and hyperparameter optimization and used a mean Dice score to evaluate against baseline models. To facilitate clinical adoption, we developed the model with an end-to-end pipeline including routing, preprocessing, and end-user interaction.

RESULTS

The autoencoder regularization-cascaded anisotropic model achieved median and mean Dice scores of 0.88/0.83 (SD, 0.09), 0.89/0.84 (SD, 0.08), and 0.81/0.72 (SD, 0.1) for whole-tumor, tumor core/resection cavity, and enhancing tumor subregions, respectively, including both preoperative and postoperative follow-up cases. The overall total processing time per case was ∼10 minutes, including data routing (∼1 minute), preprocessing (∼6 minute), segmentation (∼1-2 minute), and postprocessing (∼1 minute). Implementation challenges were discussed.

CONCLUSIONS

We show the feasibility and advantages of building a coordinated model with a clinical pipeline for the rapid and accurate deep learning segmentation of both preoperative and postoperative gliomas. The ability of the model to accommodate cases of postoperative glioma is clinically important for follow-up. An end-to-end approach, such as used here, may lead us toward successful clinical translation of tools for quantitative volume measures for glioma.

Artificial Intelligence Applications to Improve the Treatment of Locally Advanced Non-Small Cell Lung Cancers

Locally advanced non-small cell lung cancer patients represent around one third of newly diagnosed lung cancer patients. There remains a large unmet need to find treatment strategies that can improve the survival of these patients while minimizing therapeutical side effects. Increasing the availability of patients' data (imaging, electronic health records, patients' reported outcomes, and genomics) will enable the application of AI algorithms to improve therapy selections. In this review, we discuss how artificial intelligence (AI) can be integral to improving clinical decision support systems. To realize this, a roadmap for AI must be defined. We define six milestones involving a broad spectrum of stakeholders, from physicians to patients, that we feel are necessary for an optimal transition of AI into the clinic.