RESUMO
Purpose: The purpose of this work was to investigate the use of a segmentation approach that could potentially improve the speed and reproducibility of contouring during magnetic resonance-guided adaptive radiation therapy. Methods and Materials: The segmentation algorithm was based on a hybrid deep neural network and graph optimization approach that also allows rapid user intervention (Deep layered optimal graph image segmentation of multiple objects and surfaces [LOGISMOS] + just enough interaction [JEI]). A total of 115 magnetic resonance-data sets were used for training and quantitative assessment. Expert segmentations were used as the independent standard for the prostate, seminal vesicles, bladder, rectum, and femoral heads for all 115 data sets. In addition, 3 independent radiation oncologists contoured the prostate, seminal vesicles, and rectum for a subset of patients such that the interobserver variability could be quantified. Consensus contours were then generated from these independent contours using a simultaneous truth and performance level estimation approach, and the deviation of Deep LOGISMOS + JEI contours to the consensus contours was evaluated and compared with the interobserver variability. Results: The absolute accuracy of Deep LOGISMOS + JEI generated contours was evaluated using median absolute surface-to-surface distance which ranged from a minimum of 0.20 mm for the bladder to a maximum of 0.93 mm for the prostate compared with the independent standard across all data sets. The median relative surface-to-surface distance was less than 0.17 mm for all organs, indicating that the Deep LOGISMOS + JEI algorithm did not exhibit a systematic under- or oversegmentation. Interobserver variability testing yielded a mean absolute surface-to-surface distance of 0.93, 1.04, and 0.81 mm for the prostate, seminal vesicles, and rectum, respectively. In comparison, the deviation of Deep LOGISMOS + JEI from consensus simultaneous truth and performance level estimation contours was 0.57, 0.64, and 0.55 mm for the same organs. On average, the Deep LOGISMOS algorithm took less than 26 seconds for contour segmentation. Conclusions: Deep LOGISMOS + JEI segmentation efficiently generated clinically acceptable prostate and normal tissue contours, potentially limiting the need for time intensive manual contouring with each fraction.
RESUMO
Gait-stabilizing devices (GSDs) are effective at preventing falls, but people are often reluctant to use them until after experiencing a fall. Inexpensive, convenient, and effective methods for predicting which patients need GSDs could help improve adoption. The purpose of this study was to determine if a Wii Balance Board (WBB) can be used to determine whether or not patients use a GSD. We prospectively recruited participants ages 70-100, some who used GSDs and some who did not. Participants first answered questions from the Modified Vulnerable Elders Survey, and then completed a grip-strength test using a handgrip dynamometer. Finally, they were asked to complete a series of four 30-second balance tests on a WBB in random order: (1) eyes open, feet apart; (2) eyes open, feet together; (3) eyes closed, feet apart; and (4) eyes closed, feet together. The four-test series was repeated a second time in the same random order. The resulting data, represented as 25 features extracted from the questionnaires and the grip test, and data from the eight balance tests, were used to predict a subject's GSD use using generalized functional linear models based on the Bernoulli distribution. 268 participants were consented; 62 were missing data elements and were removed from analysis; 109 were not GSD users and 97 were GSD users. The use of velocity and acceleration information from the WBB improved upon predictions based solely on grip strength, demographic, and survey variables. The WBB is a convenient, inexpensive, and easy-to-use device that can be used to recommend whether or not patients should be using a GSD.
Assuntos
Força da Mão , Jogos de Vídeo , Idoso , Humanos , Marcha , Equilíbrio Postural , Reprodutibilidade dos Testes , Idoso de 80 Anos ou maisRESUMO
Purpose: Explaining deep learning model decisions, especially those for medical image segmentation, is a critical step toward the understanding and validation that will enable these powerful tools to see more widespread adoption in healthcare. We introduce kernel-weighted contribution, a visual explanation method for three-dimensional medical image segmentation models that produces accurate and interpretable explanations. Unlike previous attribution methods, kernel-weighted contribution is explicitly designed for medical image segmentation models and assesses feature importance using the relative contribution of each considered activation map to the predicted segmentation. Approach: We evaluate our method on a synthetic dataset that provides complete knowledge over input features and a comprehensive explanation quality metric using this ground truth. Our method and three other prevalent attribution methods were applied to five different model layer combinations to explain segmentation predictions for 100 test samples and compared using this metric. Results: Kernel-weighted contribution produced superior explanations of obtained image segmentations when applied to both encoder and decoder sections of a trained model as compared to other layer combinations (p<0.0005). In between-method comparisons, kernel-weighted contribution produced superior explanations compared with other methods using the same model layers in four of five experiments (p<0.0005) and showed equivalently superior performance to GradCAM++ when only using non-transpose convolution layers of the model decoder (p=0.008). Conclusion: The reported method produced explanations of superior quality uniquely suited to fully utilize the specific architectural considerations present in image and especially medical image segmentation models. Both the synthetic dataset and implementation of our method are available to the research community.
RESUMO
Quantitative analysis of neurite growth and morphology is essential for understanding the determinants of neural development and regeneration, however, it is complicated by the labor-intensive process of measuring diverse parameters of neurite outgrowth. Consequently, automated approaches have been developed to study neurite morphology in a high-throughput and comprehensive manner. These approaches include computer-automated algorithms known as 'convolutional neural networks' (CNNs)-powerful models capable of learning complex tasks without the biases of hand-crafted models. Nevertheless, their complexity often relegates them to functioning as 'black boxes.' Therefore, research in the field of explainable AI is imperative to comprehend the relationship between CNN image analysis output and predefined morphological parameters of neurite growth in order to assess the applicability of these machine learning approaches. In this study, drawing inspiration from the field of automated feature selection, we investigate the correlation between quantified metrics of neurite morphology and the image analysis results from NeuriteNet-a CNN developed to analyze neurite growth. NeuriteNet accurately distinguishes images of neurite growth based on different treatment groups within two separate experimental systems. These systems differentiate between neurons cultured on different substrate conditions and neurons subjected to drug treatment inhibiting neurite outgrowth. By examining the model's function and patterns of activation underlying its classification decisions, we discover that NeuriteNet focuses on aspects of neuron morphology that represent quantifiable metrics distinguishing these groups. Additionally, it incorporates factors that are not encompassed by neuron morphology tracing analyses. NeuriteNet presents a novel tool ideally suited for screening morphological differences in heterogeneous neuron groups while also providing impetus for targeted follow-up studies.
Assuntos
Neuritos , Neurogênese , Neurônios , Algoritmos , BenchmarkingRESUMO
BACKGROUND: During development or regeneration, neurons extend processes (i.e., neurites) via mechanisms that can be readily analyzed in culture. However, defining the impact of a drug or genetic manipulation on such mechanisms can be challenging due to the complex arborization and heterogeneous patterns of neurite growth in vitro. New Method: NeuriteNet is a Convolutional Neural Network (CNN) sorting model that uses a novel adaptation of the XRAI saliency map overlay, which is a region-based attribution method. NeuriteNet compares neuronal populations based on differences in neurite growth patterns, sorts them into respective groups, and overlays a saliency map indicating which areas differentiated the image for the sorting procedure. RESULTS: In this study, we demonstrate that NeuriteNet effectively sorts images corresponding to dissociated neurons into control and treatment groups according to known morphological differences. Furthermore, the saliency map overlay highlights the distinguishing features of the neuron when sorting the images into treatment groups. NeuriteNet also identifies novel morphological differences in neurons cultured from control and genetically modified mouse strains. Comparison with Existing Methods: Unlike other neurite analysis platforms, NeuriteNet does not require manual manipulations, such as segmentation of neurites prior to analysis, and is more accurate than experienced researchers for categorizing neurons according to their pattern of neurite growth. CONCLUSIONS: NeuriteNet can be used to effectively screen for morphological differences in a heterogeneous group of neurons and to provide feedback on the key features distinguishing those groups via the saliency map overlay.