Advancements in diagnosing oral potentially malignant disorders: leveraging Vision transformers for multi-class detection.

OBJECTIVES: Diagnosing oral potentially malignant disorders (OPMD) is critical to prevent oral cancer. This study aims to automatically detect and classify the most common pre-malignant oral lesions, such as leukoplakia and oral lichen planus (OLP), and distinguish them from oral squamous cell carcinomas (OSCC) and healthy oral mucosa on clinical photographs using vision transformers. METHODS: 4,161 photographs of healthy mucosa, leukoplakia, OLP, and OSCC were included. Findings were annotated pixel-wise and reviewed by three clinicians. The photographs were divided into 3,337 for training and validation and 824 for testing. The training and validation images were further divided into five folds with stratification. A Mask R-CNN with a Swin Transformer was trained five times with cross-validation, and the held-out test split was used to evaluate the model performance. The precision, F1-score, sensitivity, specificity, and accuracy were calculated. The area under the receiver operating characteristics curve (AUC) and the confusion matrix of the most effective model were presented. RESULTS: The detection of OSCC with the employed model yielded an F1 of 0.852 and AUC of 0.974. The detection of OLP had an F1 of 0.825 and AUC of 0.948. For leukoplakia the F1 was 0.796 and the AUC was 0.938. CONCLUSIONS: OSCC were effectively detected with the employed model, whereas the detection of OLP and leukoplakia was moderately effective. CLINICAL RELEVANCE: Oral cancer is often detected in advanced stages. The demonstrated technology may support the detection and observation of OPMD to lower the disease burden and identify malignant oral cavity lesions earlier.

A deep image classification model based on prior feature knowledge embedding and application in medical diagnosis.

Aiming at the problem of image classification with insignificant morphological structural features, strong target correlation, and low signal-to-noise ratio, combined with prior feature knowledge embedding, a deep learning method based on ResNet and Radial Basis Probabilistic Neural Network (RBPNN) is proposed model. Taking ResNet50 as a visual modeling network, it uses feature pyramid and self-attention mechanism to extract appearance and semantic features of images at multiple scales, and associate and enhance local and global features. Taking into account the diversity of category features, channel cosine similarity attention and dynamic C-means clustering algorithms are used to select representative sample features in different category of sample subsets to implicitly express prior category feature knowledge, and use them as the kernel centers of radial basis probability neurons (RBPN) to realize the embedding of diverse prior feature knowledge. In the RBPNN pattern aggregation layer, the outputs of RBPN are selectively summed according to the category of the kernel center, that is, the subcategory features are combined into category features, and finally the image classification is implemented based on Softmax. The functional module of the proposed method is designed specifically for image characteristics, which can highlight the significance of local and structural features of the image, form a non-convex decision-making area, and reduce the requirements for the completeness of the sample set. Applying the proposed method to medical image classification, experiments were conducted based on the brain tumor MRI image classification public dataset and the actual cardiac ultrasound image dataset, and the accuracy rate reached 85.82% and 83.92% respectively. Compared with the three mainstream image classification models, the performance indicators of this method have been significantly improved.

Spatial and geometric learning for classification of breast tumors from multi-center ultrasound images: a hybrid learning approach.

BACKGROUND: Breast cancer is the most common cancer among women, and ultrasound is a usual tool for early screening. Nowadays, deep learning technique is applied as an auxiliary tool to provide the predictive results for doctors to decide whether to make further examinations or treatments. This study aimed to develop a hybrid learning approach for breast ultrasound classification by extracting more potential features from local and multi-center ultrasound data. METHODS: We proposed a hybrid learning approach to classify the breast tumors into benign and malignant. Three multi-center datasets (BUSI, BUS, OASBUD) were used to pretrain a model by federated learning, then every dataset was fine-tuned at local. The proposed model consisted of a convolutional neural network (CNN) and a graph neural network (GNN), aiming to extract features from images at a spatial level and from graphs at a geometric level. The input images are small-sized and free from pixel-level labels, and the input graphs are generated automatically in an unsupervised manner, which saves the costs of labor and memory space. RESULTS: The classification AUCROC of our proposed method is 0.911, 0.871 and 0.767 for BUSI, BUS and OASBUD. The balanced accuracy is 87.6%, 85.2% and 61.4% respectively. The results show that our method outperforms conventional methods. CONCLUSIONS: Our hybrid approach can learn the inter-feature among multi-center data and the intra-feature of local data. It shows potential in aiding doctors for breast tumor classification in ultrasound at an early stage.

Automatic analysis of nuclear features reveals a non-tumoral predictor of tumor grade in bladder cancer.

BACKGROUND & OBJECTIVES: Tumor grade determines prognosis in urothelial carcinoma. The classification of low and high grade is based on nuclear morphological features that include nuclear size, hyperchromasia and pleomorphism. These features are subjectively assessed by the pathologists and are not numerically measured, which leads to high rates of interobserver variability. The purpose of this study is to assess the value of a computer-based image analysis tool for identifying predictors of tumor grade in bladder cancer. METHODS: Four hundred images of urothelial tumors were graded by five pathologists and two expert genitourinary pathologists using a scale of 1 (lowest grade) to 5 (highest grade). A computer algorithm was used to automatically segment the nuclei and to provide morphometric parameters for each nucleus, which were used to establish the grading algorithm. Grading algorithm was compared to pathologists' agreement. RESULTS: Comparison of the grading scores of the five pathologists with the expert genitourinary pathologists score showed agreement rates between 88.5% and 97.5%.The agreement rate between the two expert genitourinary pathologists was 99.5%. The quantified algorithm based conventional parameters that determine the grade (nuclear size, pleomorphism and hyperchromasia) showed > 85% agreement with the expert genitourinary pathologists. Surprisingly, the parameter that was most associated with tumor grade was the 10th percentile of the nuclear area, and high grade was associated with lower 10th percentile nuclei, caused by the presence of more inflammatory cells in the high-grade tumors. CONCLUSION: Quantitative nuclear features could be applied to determine urothelial carcinoma grade and explore new biologically explainable parameters with better correlation to grade than those currently used.

Comprehensive analysis of clinical images contributions for melanoma classification using convolutional neural networks.

BACKGROUND: Timely diagnosis plays a critical role in determining melanoma prognosis, prompting the development of deep learning models to aid clinicians. Questions persist regarding the efficacy of clinical images alone or in conjunction with dermoscopy images for model training. This study aims to compare the classification performance for melanoma of three types of CNN models: those trained on clinical images, dermoscopy images, and a combination of paired clinical and dermoscopy images from the same lesion. MATERIALS AND METHODS: We divided 914 image pairs into training, validation, and test sets. Models were built using pre-trained Inception-ResNetV2 convolutional layers for feature extraction, followed by binary classification. Training comprised 20 models per CNN type using sets of random hyperparameters. Best models were chosen based on validation AUC-ROC. RESULTS: Significant AUC-ROC differences were found between clinical versus dermoscopy models (0.661 vs. 0.869, p < 0.001) and clinical versus clinical + dermoscopy models (0.661 vs. 0.822, p = 0.001). Significant sensitivity differences were found between clinical and dermoscopy models (0.513 vs. 0.799, p = 0.01), dermoscopy versus clinical + dermoscopy models (0.799 vs. 1.000, p = 0.02), and clinical versus clinical + dermoscopy models (0.513 vs. 1.000, p < 0.001). Significant specificity differences were found between dermoscopy versus clinical + dermoscopy models (0.800 vs. 0.288, p < 0.001) and clinical versus clinical + dermoscopy models (0.650 vs. 0.288, p < 0.001). CONCLUSION: CNN models trained on dermoscopy images outperformed those relying solely on clinical images under our study conditions. The potential advantages of incorporating paired clinical and dermoscopy images for CNN-based melanoma classification appear less clear based on our findings.

Enhancing brain tumor detection in MRI images through explainable AI using Grad-CAM with Resnet 50.

This study addresses the critical challenge of detecting brain tumors using MRI images, a pivotal task in medical diagnostics that demands high accuracy and interpretability. While deep learning has shown remarkable success in medical image analysis, there remains a substantial need for models that are not only accurate but also interpretable to healthcare professionals. The existing methodologies, predominantly deep learning-based, often act as black boxes, providing little insight into their decision-making process. This research introduces an integrated approach using ResNet50, a deep learning model, combined with Gradient-weighted Class Activation Mapping (Grad-CAM) to offer a transparent and explainable framework for brain tumor detection. We employed a dataset of MRI images, enhanced through data augmentation, to train and validate our model. The results demonstrate a significant improvement in model performance, with a testing accuracy of 98.52% and precision-recall metrics exceeding 98%, showcasing the model's effectiveness in distinguishing tumor presence. The application of Grad-CAM provides insightful visual explanations, illustrating the model's focus areas in making predictions. This fusion of high accuracy and explainability holds profound implications for medical diagnostics, offering a pathway towards more reliable and interpretable brain tumor detection tools.

Comparison of medical image interpretation time between conventional and automated methods of breast ultrasound.

Objective: To compare the medical image interpretation's time between the conventional and automated methods of breast ultrasound in patients with breast lesions. Secondarily, to evaluate the agreement between the two methods and interobservers. Methods: This is a cross-sectional study with prospective data collection. The agreement's degrees were established in relation to the breast lesions's ultrasound descriptors. To determine the accuracy of each method, a biopsy of suspicious lesions was performed, considering the histopathological result as the diagnostic gold standard. Results: We evaluated 27 women. Conventional ultrasound used an average medical time of 10.77 minutes (± 2.55) greater than the average of 7.38 minutes (± 2.06) for automated ultrasound (p<0.001). The degrees of agreement between the methods ranged from 0.75 to 0.95 for researcher 1 and from 0.71 to 0.98 for researcher 2. Among the researchers, the degrees of agreement were between 0.63 and 1 for automated ultrasound and between 0.68 and 1 for conventional ultrasound. The area of the ROC curve for the conventional method was 0.67 (p=0.003) for researcher 1 and 0.72 (p<0.001) for researcher 2. The area of the ROC curve for the automated method was 0. 69 (p=0.001) for researcher 1 and 0.78 (p<0.001) for researcher 2. Conclusion: We observed less time devoted by the physician to automated ultrasound compared to conventional ultrasound, maintaining accuracy. There was substantial or strong to perfect interobserver agreement and substantial or strong to almost perfect agreement between the methods.

Brain tumor detection with integrating traditional and computational intelligence approaches across diverse imaging modalities - Challenges and future directions.

Brain tumor segmentation and classification play a crucial role in the diagnosis and treatment planning of brain tumors. Accurate and efficient methods for identifying tumor regions and classifying different tumor types are essential for guiding medical interventions. This study comprehensively reviews brain tumor segmentation and classification techniques, exploring various approaches based on image processing, machine learning, and deep learning. Furthermore, our study aims to review existing methodologies, discuss their advantages and limitations, and highlight recent advancements in this field. The impact of existing segmentation and classification techniques for automated brain tumor detection is also critically examined using various open-source datasets of Magnetic Resonance Images (MRI) of different modalities. Moreover, our proposed study highlights the challenges related to segmentation and classification techniques and datasets having various MRI modalities to enable researchers to develop innovative and robust solutions for automated brain tumor detection. The results of this study contribute to the development of automated and robust solutions for analyzing brain tumors, ultimately aiding medical professionals in making informed decisions and providing better patient care.

CauDR: A causality-inspired domain generalization framework for fundus-based diabetic retinopathy grading.

Diabetic retinopathy (DR) is the most common diabetic complication, which usually leads to retinal damage, vision loss, and even blindness. A computer-aided DR grading system has a significant impact on helping ophthalmologists with rapid screening and diagnosis. Recent advances in fundus photography have precipitated the development of novel retinal imaging cameras and their subsequent implementation in clinical practice. However, most deep learning-based algorithms for DR grading demonstrate limited generalization across domains. This inferior performance stems from variance in imaging protocols and devices inducing domain shifts. We posit that declining model performance between domains arises from learning spurious correlations in the data. Incorporating do-operations from causality analysis into model architectures may mitigate this issue and improve generalizability. Specifically, a novel universal structural causal model (SCM) was proposed to analyze spurious correlations in fundus imaging. Building on this, a causality-inspired diabetic retinopathy grading framework named CauDR was developed to eliminate spurious correlations and achieve more generalizable DR diagnostics. Furthermore, existing datasets were reorganized into 4DR benchmark for DG scenario. Results demonstrate the effectiveness and the state-of-the-art (SOTA) performance of CauDR. Diabetic retinopathy (DR) is the most common diabetic complication, which usually leads to retinal damage, vision loss, and even blindness. A computer-aided DR grading system has a significant impact on helping ophthalmologists with rapid screening and diagnosis. Recent advances in fundus photography have precipitated the development of novel retinal imaging cameras and their subsequent implementation in clinical practice. However, most deep learning-based algorithms for DR grading demonstrate limited generalization across domains. This inferior performance stems from variance in imaging protocols and devices inducing domain shifts. We posit that declining model performance between domains arises from learning spurious correlations in the data. Incorporating do-operations from causality analysis into model architectures may mitigate this issue and improve generalizability. Specifically, a novel universal structural causal model (SCM) was proposed to analyze spurious correlations in fundus imaging. Building on this, a causality-inspired diabetic retinopathy grading framework named CauDR was developed to eliminate spurious correlations and achieve more generalizable DR diagnostics. Furthermore, existing datasets were reorganized into 4DR benchmark for DG scenario. Results demonstrate the effectiveness and the state-of-the-art (SOTA) performance of CauDR.

Diabetic retinopathy prediction based on vision transformer and modified capsule network.

Diabetic retinopathy is considered one of the most common diseases that can lead to blindness in the working age, and the chance of developing it increases as long as a person suffers from diabetes. Protecting the sight of the patient or decelerating the evolution of this disease depends on its early detection as well as identifying the exact levels of this pathology, which is done manually by ophthalmologists. This manual process is very consuming in terms of the time and experience of an expert ophthalmologist, which makes developing an automated method to aid in the diagnosis of diabetic retinopathy an essential and urgent need. In this paper, we aim to propose a new hybrid deep learning method based on a fine-tuning vision transformer and a modified capsule network for automatic diabetic retinopathy severity level prediction. The proposed approach consists of a new range of computer vision operations, including the power law transformation technique and the contrast-limiting adaptive histogram equalization technique in the preprocessing step. While the classification step builds up on a fine-tuning vision transformer, a modified capsule network, and a classification model combined with a classification model, The effectiveness of our approach was evaluated using four datasets, including the APTOS, Messidor-2, DDR, and EyePACS datasets, for the task of severity levels of diabetic retinopathy. We have attained excellent test accuracy scores on the four datasets, respectively: 88.18%, 87.78%, 80.36%, and 78.64%. Comparing our results with the state-of-the-art, we reached a better performance.

Improving multiple sclerosis lesion segmentation across clinical sites: A federated learning approach with noise-resilient training.

Accurately measuring the evolution of Multiple Sclerosis (MS) with magnetic resonance imaging (MRI) critically informs understanding of disease progression and helps to direct therapeutic strategy. Deep learning models have shown promise for automatically segmenting MS lesions, but the scarcity of accurately annotated data hinders progress in this area. Obtaining sufficient data from a single clinical site is challenging and does not address the heterogeneous need for model robustness. Conversely, the collection of data from multiple sites introduces data privacy concerns and potential label noise due to varying annotation standards. To address this dilemma, we explore the use of the federated learning framework while considering label noise. Our approach enables collaboration among multiple clinical sites without compromising data privacy under a federated learning paradigm that incorporates a noise-robust training strategy based on label correction. Specifically, we introduce a Decoupled Hard Label Correction (DHLC) strategy that considers the imbalanced distribution and fuzzy boundaries of MS lesions, enabling the correction of false annotations based on prediction confidence. We also introduce a Centrally Enhanced Label Correction (CELC) strategy, which leverages the aggregated central model as a correction teacher for all sites, enhancing the reliability of the correction process. Extensive experiments conducted on two multi-site datasets demonstrate the effectiveness and robustness of our proposed methods, indicating their potential for clinical applications in multi-site collaborations to train better deep learning models with lower cost in data collection and annotation.

Application of artificial intelligence in pancreas endoscopic ultrasound imaging- A systematic review.

The pancreas is a vital organ in digestive system which has significant health implications. It is imperative to evaluate and identify malignant pancreatic lesions promptly in light of the high mortality rate linked to such malignancies. Endoscopic Ultrasound (EUS) is a non-invasive precise technique to detect pancreas disorders, but it is highly operator dependent. Artificial intelligence (AI), including traditional machine learning (ML) and deep learning (DL) techniques can play a pivotal role to enhancing the performance of EUS regardless of operator. AI performs a critical function in the detection, classification, and segmentation of medical images. The utilization of AI-assisted systems has improved the accuracy and productivity of pancreatic analysis, including the detection of diverse pancreatic disorders (e.g., pancreatitis, masses, and cysts) as well as landmarks and parenchyma. This systematic review examines the rapidly developing domain of AI-assisted system in EUS of the pancreas. Its objective is to present a thorough study of the present research status and developments in this area. This paper explores the significant challenges of AI-assisted system in pancreas EUS imaging, highlights the potential of AI techniques in addressing these challenges, and suggests the scope for future research in domain of AI-assisted EUS systems.

A novel multi-task learning network for skin lesion classification based on multi-modal clues and label-level fusion.

In this paper, we propose a multi-task learning (MTL) network based on the label-level fusion of metadata and hand-crafted features by unsupervised clustering to generate new clustering labels as an optimization goal. We propose a MTL module (MTLM) that incorporates an attention mechanism to enable the model to learn more integrated, variable information. We propose a dynamic strategy to adjust the loss weights of different tasks, and trade off the contributions of multiple branches. Instead of feature-level fusion, we propose label-level fusion and combine the results of our proposed MTLM with the results of the image classification network to achieve better lesion prediction on multiple dermatological datasets. We verify the effectiveness of the proposed model by quantitative and qualitative measures. The MTL network using multi-modal clues and label-level fusion can yield the significant performance improvement for skin lesion classification.

Automated Prediction of Malignant Melanoma using Two-Stage Convolutional Neural Network.

PURPOSE: A skin lesion refers to an area of the skin that exhibits anomalous growth or distinctive visual characteristics compared to the surrounding skin. Benign skin lesions are noncancerous and generally pose no threat. These irregular skin growths can vary in appearance. On the other hand, malignant skin lesions correspond to skin cancer, which happens to be the most prevalent form of cancer in the United States. Skin cancer involves the unusual proliferation of skin cells anywhere on the body. The conventional method for detecting skin cancer is relatively more painful. METHODS: This work involves the automated prediction of skin cancer and its types using two stage Convolutional Neural Network (CNN). The first stage of CNN extracts low level features and second stage extracts high level features. Feature selection is done using these two CNN and ABCD (Asymmetry, Border irregularity, Colour variation, and Diameter) technique. The features extracted from the two CNNs are fused with ABCD features and fed into classifiers for the final prediction. The classifiers employed in this work include ensemble learning methods such as gradient boosting and XG boost, as well as machine learning classifiers like decision trees and logistic regression. This methodology is evaluated using the International Skin Imaging Collaboration (ISIC) 2018 and 2019 dataset. RESULTS: As a result, the first stage CNN which is used for creation of new dataset achieved an accuracy of 97.92%. Second stage CNN which is used for feature selection achieved an accuracy of 98.86%. Classification results are obtained for both with and without fusion of features. CONCLUSION: Therefore, two stage prediction model achieved better results with feature fusion.

Refining neural network algorithms for accurate brain tumor classification in MRI imagery.

Brain tumor diagnosis using MRI scans poses significant challenges due to the complex nature of tumor appearances and variations. Traditional methods often require extensive manual intervention and are prone to human error, leading to misdiagnosis and delayed treatment. Current approaches primarily include manual examination by radiologists and conventional machine learning techniques. These methods rely heavily on feature extraction and classification algorithms, which may not capture the intricate patterns present in brain MRI images. Conventional techniques often suffer from limited accuracy and generalizability, mainly due to the high variability in tumor appearance and the subjective nature of manual interpretation. Additionally, traditional machine learning models may struggle with the high-dimensional data inherent in MRI images. To address these limitations, our research introduces a deep learning-based model utilizing convolutional neural networks (CNNs).Our model employs a sequential CNN architecture with multiple convolutional, max-pooling, and dropout layers, followed by dense layers for classification. The proposed model demonstrates a significant improvement in diagnostic accuracy, achieving an overall accuracy of 98% on the test dataset. The proposed model demonstrates a significant improvement in diagnostic accuracy, achieving an overall accuracy of 98% on the test dataset. The precision, recall, and F1-scores ranging from 97 to 98% with a roc-auc ranging from 99 to 100% for each tumor category further substantiate the model's effectiveness. Additionally, the utilization of Grad-CAM visualizations provides insights into the model's decision-making process, enhancing interpretability. This research addresses the pressing need for enhanced diagnostic accuracy in identifying brain tumors through MRI imaging, tackling challenges such as variability in tumor appearance and the need for rapid, reliable diagnostic tools.

A bi-directional segmentation method for prostate ultrasound images under semantic constraints.

Due to the lack of sufficient labeled data for the prostate and the extensive and complex semantic information in ultrasound images, accurately and quickly segmenting the prostate in transrectal ultrasound (TRUS) images remains a challenging task. In this context, this paper proposes a solution for TRUS image segmentation using an end-to-end bidirectional semantic constraint method, namely the BiSeC model. The experimental results show that compared with classic or popular deep learning methods, this method has better segmentation performance, with the Dice Similarity Coefficient (DSC) of 96.74% and the Intersection over Union (IoU) of 93.71%. Our model achieves a good balance between actual boundaries and noise areas, reducing costs while ensuring the accuracy and speed of segmentation.

Optimising the diagnostic accuracy of First post-contrAst SubtracTed breast MRI (FAST MRI) through interpretation-training: a multicentre e-learning study, mapping the learning curve of NHS Breast Screening Programme (NHSBSP) mammogram readers using an enriched dataset.

BACKGROUND: Abbreviated breast MRI (FAST MRI) is being introduced into clinical practice to screen women with mammographically dense breasts or with a personal history of breast cancer. This study aimed to optimise diagnostic accuracy through the adaptation of interpretation-training. METHODS: A FAST MRI interpretation-training programme (short presentations and guided hands-on workstation teaching) was adapted to provide additional training during the assessment task (interpretation of an enriched dataset of 125 FAST MRI scans) by giving readers feedback about the true outcome of each scan immediately after each scan was interpreted (formative assessment). Reader interaction with the FAST MRI scans used developed software (RiViewer) that recorded reader opinions and reading times for each scan. The training programme was additionally adapted for remote e-learning delivery. STUDY DESIGN: Prospective, blinded interpretation of an enriched dataset by multiple readers. RESULTS: 43 mammogram readers completed the training, 22 who interpreted breast MRI in their clinical role (Group 1) and 21 who did not (Group 2). Overall sensitivity was 83% (95%CI 81-84%; 1994/2408), specificity 94% (95%CI 93-94%; 7806/8338), readers' agreement with the true outcome kappa = 0.75 (95%CI 0.74-0.77) and diagnostic odds ratio = 70.67 (95%CI 61.59-81.09). Group 1 readers showed similar sensitivity (84%) to Group 2 (82% p = 0.14), but slightly higher specificity (94% v. 93%, p = 0.001). Concordance with the ground truth increased significantly with the number of FAST MRI scans read through the formative assessment task (p = 0.002) but by differing amounts depending on whether or not a reader had previously attended FAST MRI training (interaction p = 0.02). Concordance with the ground truth was significantly associated with reading batch size (p = 0.02), tending to worsen when more than 50 scans were read per batch. Group 1 took a median of 56 seconds (range 8-47,466) to interpret each FAST MRI scan compared with 78 (14-22,830, p < 0.0001) for Group 2. CONCLUSIONS: Provision of immediate feedback to mammogram readers during the assessment test set reading task increased specificity for FAST MRI interpretation and achieved high diagnostic accuracy. Optimal reading-batch size for FAST MRI was 50 reads per batch. Trial registration (25/09/2019): ISRCTN16624917.

Texture analysis using non-contrast MRI to predict estrogen receptor status in breast cancer lesions: Is it feasible?

PURPOSE: We aimed to evaluate the feasibility of non-contrast-enhanced T1 sequence in texture analysis of breast cancer lesions to predict their estrogen receptor status. METHODS: The study included 85 pathologically proven breast cancer lesions in 53 patients. Immunohistochemical studies were performed to determine the estrogen receptor status (ER). Lesions were divided into two groups: ER + ve status and ER-ve status. Texture analysis using the second-order analysis features [The Co-occurrence matrix (11 features)] was applied on both T1 and dynamic contrast-enhanced (DCE) MRI images for each lesion. Texture features gained from both T1 and DCE images were analyzed to obtain cut-off values using ROC curves to sort lesions according to their estrogen receptor status. RESULTS: Angular second momentum and some of the entropy-based features showed statistically significant cut-off values in differentiation between the two groups [P-values for pre- and post-contrast images for AngSecMom (0.001, 0.008), sum entropy (0.003,0.005), and entropy (0.033,0.019), respectively]. On comparing the AUCs between pre- and post-contrast images, we found that differences were statistically insignificant. Sum of squares, sum variance, and sum average showed statistically significant cut-off points only on pre-contrast images [P-values for sum of squares (0.018), sum variance (0.024), and sum average (0.039)]. CONCLUSIONS: Texture analysis features showed promising results in predicting estrogen receptor status of breast cancer lesions on non-contrast T1 images.

Investigating the discrimination ability of 3D convolutional neural networks applied to altered brain MRI parametric maps.

Convolutional neural networks (CNNs) are gradually being recognized in the neuroimaging community as a powerful tool for image analysis. Despite their outstanding performances, some aspects of CNN functioning are still not fully understood by human operators. We postulated that the interpretability of CNNs applied to neuroimaging data could be improved by investigating their behavior when they are fed data with known characteristics. We analyzed the ability of 3D CNNs to discriminate between original and altered whole-brain parametric maps derived from diffusion-weighted magnetic resonance imaging. The alteration consisted in linearly changing the voxel intensity of either one (monoregion) or two (biregion) anatomical regions in each brain volume, but without mimicking any neuropathology. Performing ten-fold cross-validation and using a hold-out set for testing, we assessed the CNNs' discrimination ability according to the intensity of the altered regions, comparing the latter's size and relative position. Monoregion CNNs showed that the larger the modified region, the smaller the intensity increase needed to achieve good performances. Biregion CNNs systematically outperformed monoregion CNNs, but could only detect one of the two target regions when tested on the corresponding monoregion images. Exploiting prior information on training data allowed for a better understanding of CNN behavior, especially when altered regions were combined. This can inform about the complexity of CNN pattern retrieval and elucidate misclassified examples, particularly relevant for pathological data. The proposed analytical approach may serve to gain insights into CNN behavior and guide the design of enhanced detection systems exploiting our prior knowledge.