Object Detection Improves Tumour Segmentation in MR Images of Rare Brain Tumours

Deep Learning-Based Automated Measurement of Murine Bone Length in Radiographs

Genetic mouse models of skeletal abnormalities have demonstrated promise in the identification of phenotypes relevant to human skeletal diseases. Traditionally, phenotypes are assessed by manually examining radiographs, a tedious and potentially error-prone process. In response, this study developed a deep learning-based model that streamlines the measurement of murine bone lengths from radiographs in an accurate and reproducible manner. A bone detection and measurement pipeline utilizing the Keypoint R-CNN algorithm with an EfficientNet-B3 feature extraction backbone was developed to detect murine bone positions and measure their lengths. The pipeline was developed utilizing 94 X-ray images with expert annotations on the start and end position of each murine bone. The accuracy of our pipeline was evaluated on an independent dataset test with 592 images, and further validated on a previously published dataset of 21,300 mouse radiographs. The results showed that our model performed comparably to humans in measuring tibia and femur lengths (R2 > 0.92, p-value = 0) and significantly outperformed humans in measuring pelvic lengths in terms of precision and consistency. Furthermore, the model improved the precision and consistency of genetic association mapping results, identifying significant associations between genetic mutations and skeletal phenotypes with reduced variability. This study demonstrates the feasibility and efficiency of automated murine bone length measurement in the identification of mouse models of abnormal skeletal phenotypes.

Quantitative and Visual Analysis of Data Augmentation and Hyperparameter Optimization in Deep Learning-Based Segmentation of Low-Grade Glioma Tumors Using Grad-CAM

This study executes a quantitative and visual investigation on the effectiveness of data augmentation and hyperparameter optimization on the accuracy of deep learning-based segmentation of LGG tumors. The study employed the MobileNetV2 and ResNet backbones with atrous convolution in DeepLabV3+ structure. The Grad-CAM tool was also used to interpret the effect of augmentation and network optimization on segmentation performance. A wide investigation was performed to optimize the network hyperparameters. In addition, the study examined 35 different models to evaluate different data augmentation techniques. The results of the study indicated that incorporating data augmentation techniques and optimization can improve the performance of segmenting brain LGG tumors up to 10%. Our extensive investigation of the data augmentation techniques indicated that enlargement of data from 90° and 225° rotated data,up to down and left to right flipping are the most effective techniques. MobilenetV2 as the backbone,"Focal Loss" as the loss function and "Adam" as the optimizer showed the superior results. The optimal model (DLG-Net) achieved an overall accuracy of 96.1% with a loss value of 0.006. Specifically, the segmentation performance for Whole Tumor (WT), Tumor Core (TC), and Enhanced Tumor (ET) reached a Dice Similarity Coefficient (DSC) of 89.4%, 70.1%, and 49.9%, respectively. Simultaneous visual and quantitative assessment of data augmentation and network optimization can lead to an optimal model with a reasonable performance in segmenting the LGG tumors.

In-Domain Transfer Learning Strategy for Tumor Detection on Brain MRI

Transfer learning has gained importance in areas where there is a labeled data shortage. However, it is still controversial as to what extent natural image datasets as pre-training sources contribute scientifically to success in different fields, such as medical imaging. In this study, the effect of transfer learning for medical object detection was quantitatively compared using natural and medical image datasets. Within the scope of this study, transfer learning strategies based on five different weight initialization methods were discussed. A natural image dataset MS COCO and brain tumor dataset BraTS 2020 were used as the transfer learning source, and Gazi Brains 2020 was used for the target. Mask R-CNN was adopted as a deep learning architecture for its capability to effectively handle both object detection and segmentation tasks. The experimental results show that transfer learning from the medical image dataset was found to be 10% more successful and showed 24% better convergence performance than the MS COCO pre-trained model, although it contains fewer data. While the effect of data augmentation on the natural image pre-trained model was 5%, the same domain pre-trained model was measured as 2%. According to the most widely used object detection metric, transfer learning strategies using MS COCO weights and random weights showed the same object detection performance as data augmentation. The performance of the most effective strategies identified in the Mask R-CNN model was also tested with YOLOv8. Results showed that even if the amount of data is less than the natural dataset, in-domain transfer learning is more efficient than cross-domain transfer learning. Moreover, this study demonstrates the first use of the Gazi Brains 2020 dataset, which was generated to address the lack of labeled and qualified brain MRI data in the medical field for in-domain transfer learning. Thus, knowledge transfer was carried out from the deep neural network, which was trained with brain tumor data and tested on a different brain tumor dataset.

An Ensemble of Deep Learning Object Detection Models for Anatomical and Pathological Regions in Brain MRI

This paper proposes ensemble strategies for the deep learning object detection models carried out by combining the variants of a model and different models to enhance the anatomical and pathological object detection performance in brain MRI. In this study, with the help of the novel Gazi Brains 2020 dataset, five different anatomical parts and one pathological part that can be observed in brain MRI were identified, such as the region of interest, eye, optic nerves, lateral ventricles, third ventricle, and a whole tumor. Firstly, comprehensive benchmarking of the nine state-of-the-art object detection models was carried out to determine the capabilities of the models in detecting the anatomical and pathological parts. Then, four different ensemble strategies for nine object detectors were applied to boost the detection performance using the bounding box fusion technique. The ensemble of individual model variants increased the anatomical and pathological object detection performance by up to 10% in terms of the mean average precision (mAP). In addition, considering the class-based average precision (AP) value of the anatomical parts, an up to 18% AP improvement was achieved. Similarly, the ensemble strategy of the best different models outperformed the best individual model by 3.3% mAP. Additionally, while an up to 7% better FAUC, which is the area under the TPR vs. FPPI curve, was achieved on the Gazi Brains 2020 dataset, a 2% better FAUC score was obtained on the BraTS 2020 dataset. The proposed ensemble strategies were found to be much more efficient in finding the anatomical and pathological parts with a small number of anatomic objects, such as the optic nerve and third ventricle, and producing higher TPR values, especially at low FPPI values, compared to the best individual methods.

Multimodal MRI radiomic models to predict genomic mutations in diffuse intrinsic pontine glioma with missing imaging modalities

Purpose

Predicting H3.1, TP53, and ACVR1 mutations in DIPG could aid in the selection of therapeutic options. The contribution of clinical data and multi-modal MRI were studied for these three predictive tasks. To keep the maximum number of subjects, which is essential for a rare disease, missing data were considered. A multi-modal model was proposed, collecting all available data for each patient, without performing any imputation.

Methods

A retrospective cohort of 80 patients with confirmed DIPG and at least one of the four MR modalities (T1w, T1c, T2w, and FLAIR), acquired with two different MR scanners was built. A pipeline including standardization of MR data and extraction of radiomic features within the tumor was applied. The values of radiomic features between the two MR scanners were realigned using the ComBat method. For each prediction task, the most robust features were selected based on a recursive feature elimination with cross-validation. Five different models, one based on clinical data and one per MR modality, were developed using logistic regression classifiers. The prediction of the multi-modal model was defined as the average of all possible prediction results among five for each patient. The performances of the models were compared using a leave-one-out approach.

Results

The percentage of missing modalities ranged from 6 to 11% across modalities and tasks. The performance of each individual model was dependent on each specific task, with an AUC of the ROC curve ranging from 0.63 to 0.80. The multi-modal model outperformed the clinical model for each prediction tasks, thus demonstrating the added value of MRI. Furthermore, regardless of performance criteria, the multi-modal model came in the first place or second place (very close to first). In the leave-one-out approach, the prediction of H3.1 (resp. ACVR1 and TP53) mutations achieved a balanced accuracy of 87.8% (resp. 82.1 and 78.3%).

Conclusion

Compared with a single modality approach, the multi-modal model combining multiple MRI modalities and clinical features was the most powerful to predict H3.1, ACVR1, and TP53 mutations and provided prediction, even in the case of missing modality. It could be proposed in the absence of a conclusive biopsy.