Enhancing classification of cells procured from bone marrow aspirate smears using generative adversarial networks and sequential convolutional neural network

Improving the generalizability of white blood cell classification with few-shot domain adaptation

The morphological classification of nucleated blood cells is fundamental for the diagnosis of hematological diseases. Many Deep Learning algorithms have been implemented to automatize this classification task, but most of the time they fail to classify images coming from different sources. This is known as "domain shift". Whereas some research has been conducted in this area, domain adaptation techniques are often computationally expensive and can introduce significant modifications to initial cell images. In this article, we propose an easy-to-implement workflow where we trained a model to classify images from two datasets, and tested it on images coming from eight other datasets. An EfficientNet model was trained on a source dataset comprising images from two different datasets. It was afterwards fine-tuned on each of the eight target datasets by using 100 or less-annotated images from these datasets. Images from both the source and the target dataset underwent a color transform to put them into a standardized color style. The importance of color transform and fine-tuning was evaluated through an ablation study and visually assessed with scatter plots, and an extensive error analysis was carried out. The model achieved an accuracy higher than 80% for every dataset and exceeded 90% for more than half of the datasets. The presented workflow yielded promising results in terms of generalizability, significantly improving performance on target datasets, whereas keeping low computational cost and maintaining consistent color transformations. Source code is available at: https://github.com/mc2295/WBC_Generalization.

Models for the marrow: A comprehensive review of AI-based cell classification methods and malignancy detection in bone marrow aspirate smears

Given the high prevalence of artificial intelligence (AI) research in medicine, the development of deep learning (DL) algorithms based on image recognition, such as the analysis of bone marrow aspirate (BMA) smears, is rapidly increasing in the field of hematology and oncology. The models are trained to identify the optimal regions of the BMA smear for differential cell count and subsequently detect and classify a number of cell types, which can ultimately be utilized for diagnostic purposes. Moreover, AI is capable of identifying genetic mutations phenotypically. This pipeline has the potential to offer an accurate and rapid preliminary analysis of the bone marrow in the clinical routine. However, the intrinsic complexity of hematological diseases presents several challenges for the automatic morphological assessment. To ensure general applicability across multiple medical centers and to deliver high accuracy on prospective clinical data, AI models would require highly heterogeneous training datasets. This review presents a systematic analysis of models for cell classification and detection of hematological malignancies published in the last 5 years (2019-2024). It provides insight into the challenges and opportunities of these DL-assisted tasks.

Engineered feature embeddings meet deep learning: A novel strategy to improve bone marrow cell classification and model transparency

Cytomorphology evaluation of bone marrow cell is the initial step to diagnose different hematological diseases. This assessment is still manually performed by trained specialists, who may be a bottleneck within the clinical process. Deep learning algorithms are a promising approach to automate this bone marrow cell evaluation. These artificial intelligence models have focused on limited cell subtypes, mainly associated to a particular disease, and are frequently presented as black boxes. The herein introduced strategy presents an engineered feature representation, the region-attention embedding, which improves the deep learning classification performance of a cytomorphology with 21 bone marrow cell subtypes. This embedding is built upon a specific organization of cytology features within a squared matrix by distributing them after pre-segmented cell regions, i.e., cytoplasm, nucleus, and whole-cell. This novel cell image representation, aimed to preserve spatial/regional relations, is used as input of the network. Combination of region-attention embedding and deep learning networks (Xception and ResNet50) provides local relevance associated to image regions, adding up interpretable information to the prediction. Additionally, this approach is evaluated in a public database with the largest number of cell subtypes (21) by a thorough evaluation scheme with three iterations of a 3-fold cross-validation, performed in 80% of the images (n = 89,484), and a testing process in an unseen set of images composed by the remaining 20% of the images (n = 22,371). This evaluation process demonstrates the introduced strategy outperforms previously published approaches in an equivalent validation set, with a f1-score of 0.82, and presented competitive results on the unseen data partition with a f1-score of 0.56.

Efficient leukocytes detection and classification in microscopic blood images using convolutional neural network coupled with a dual attention network

Leukocytes, also called White Blood Cells (WBCs) or leucocytes, are the cells that play a pivotal role in human health and are vital indicators of diseases such as malaria, leukemia, AIDS, and other viral infections. WBCs detection and classification in blood smears offers insights to pathologists, aiding diagnosis across medical conditions. Traditional techniques, including manual counting, detection, classification, and visual inspection of microscopic images by medical professionals, pose challenges due to their labor-intensive nature. However, traditional methods are time consuming and sometimes susceptible to errors. Here, we propose a high-performance convolutional neural network (CNN) coupled with a dual-attention network that efficiently detects and classifies WBCs in microscopic thick smear images. The main aim of this study was to enhance clinical hematology systems and expedite medical diagnostic processes. In the proposed technique, we utilized a deep convolutional generative adversarial network (DCGAN) to overcome the limitations imposed by limited training data and employed a dual attention mechanism to improve accuracy, efficiency, and generalization. The proposed technique achieved overall accuracy rates of 99.83%, 99.35%, and 99.60% for the peripheral blood cell (PBC), leukocyte images for segmentation and classification (LISC), and Raabin-WBC benchmark datasets, respectively. Our proposed approach outperforms state-of-the-art methods in terms of accuracy, highlighting the effectiveness of the strategies employed and their potential to enhance diagnostic capabilities and advance real-world healthcare practices and diagnostic systems.

Prediction of Myocardial Infarction Using a Combined Generative Adversarial Network Model and Feature-Enhanced Loss Function

Accurate risk prediction for myocardial infarction (MI) is crucial for preventive strategies, given its significant impact on global mortality and morbidity. Here, we propose a novel deep-learning approach to enhance the prediction of incident MI cases by incorporating metabolomics alongside clinical risk factors. We utilized data from the KORA cohort, including the baseline S4 and follow-up F4 studies, consisting of 1454 participants without prior history of MI. The dataset comprised 19 clinical variables and 363 metabolites. Due to the imbalanced nature of the dataset (78 observed MI cases and 1376 non-MI individuals), we employed a generative adversarial network (GAN) model to generate new incident cases, augmenting the dataset and improving feature representation. To predict MI, we further utilized multi-layer perceptron (MLP) models in conjunction with the synthetic minority oversampling technique (SMOTE) and edited nearest neighbor (ENN) methods to address overfitting and underfitting issues, particularly when dealing with imbalanced datasets. To enhance prediction accuracy, we propose a novel GAN for feature-enhanced (GFE) loss function. The GFE loss function resulted in an approximate 2% improvement in prediction accuracy, yielding a final accuracy of 70%. Furthermore, we evaluated the contribution of each clinical variable and metabolite to the predictive model and identified the 10 most significant variables, including glucose tolerance, sex, and physical activity. This is the first study to construct a deep-learning approach for producing 7-year MI predictions using the newly proposed loss function. Our findings demonstrate the promising potential of our technique in identifying novel biomarkers for MI prediction.

Evaluation of deep learning training strategies for the classification of bone marrow cell images

BACKGROUND AND OBJECTIVE

The classification of bone marrow (BM) cells by light microscopy is an important cornerstone of hematological diagnosis, performed thousands of times a day by highly trained specialists in laboratories worldwide. As the manual evaluation of blood or BM smears is very time-consuming and prone to inter-observer variation, new reliable automated systems are needed.

METHODS

We aim to improve the automatic classification performance of hematological cell types. Therefore, we evaluate four state-of-the-art Convolutional Neural Network (CNN) architectures on a dataset of 171,374 microscopic cytological single-cell images obtained from BM smears from 945 patients diagnosed with a variety of hematological diseases. We further evaluate the effect of an in-domain vs. out-of-domain pre-training, and assess whether class activation maps provide human-interpretable explanations for the models' predictions.

RESULTS

The best performing pre-trained model (Regnet_y_32gf) yields a mean precision, recall, and F1 scores of 0.787±0.060, 0.755±0.061, and 0.762±0.050, respectively. This is a 53.5% improvement in precision and 7.3% improvement in recall over previous results with CNNs (ResNeXt-50) that were trained from scratch. The out-of-domain pre-training apparently yields general feature extractors/filters that apply very well to the BM cell classification use case. The class activation maps on cell types with characteristic morphological features were found to be consistent with the explanations of a human domain expert. For example, the Auer rods in the cytoplasm were the predictive cellular feature for correctly classified images of faggot cells.

CONCLUSIONS

Our study provides data that can help hematology laboratories to choose the optimal training strategy for blood cell classification deep learning models to improve computer-assisted blood and bone marrow cell identification. It also highlights the need for more specific training data, i.e. images of difficult-to-classify classes, including cells labeled with disease information.

Increasing-Margin Adversarial (IMA) training to improve adversarial robustness of neural networks

BACKGROUND AND OBJECTIVE

Deep neural networks (DNNs) are vulnerable to adversarial noises. Adversarial training is a general and effective strategy to improve DNN robustness (i.e., accuracy on noisy data) against adversarial noises. However, DNN models trained by the current existing adversarial training methods may have much lower standard accuracy (i.e., accuracy on clean data), compared to the same models trained by the standard method on clean data, and this phenomenon is known as the trade-off between accuracy and robustness and is commonly considered unavoidable. This issue prevents adversarial training from being used in many application domains, such as medical image analysis, as practitioners do not want to sacrifice standard accuracy too much in exchange for adversarial robustness. Our objective is to lift (i.e., alleviate or even avoid) this trade-off between standard accuracy and adversarial robustness for medical image classification and segmentation.

METHODS

We propose a novel adversarial training method, named Increasing-Margin Adversarial (IMA) Training, which is supported by an equilibrium state analysis about the optimality of adversarial training samples. Our method aims to preserve accuracy while improving robustness by generating optimal adversarial training samples. We evaluate our method and the other eight representative methods on six publicly available image datasets corrupted by noises generated by AutoAttack and white-noise attack.

RESULTS

Our method achieves the highest adversarial robustness for image classification and segmentation with the smallest reduction in accuracy on clean data. For one of the applications, our method improves both accuracy and robustness.

CONCLUSIONS

Our study has demonstrated that our method can lift the trade-off between standard accuracy and adversarial robustness for the image classification and segmentation applications. To our knowledge, it is the first work to show that the trade-off is avoidable for medical image segmentation.

Efficient diagnosis of hematologic malignancies using bone marrow microscopic images: A method based on MultiPathGAN and MobileViTv2

BACKGROUND AND OBJECTIVES

Hematologic malignancies, including the associated multiple subtypes, are critically threatening to human health. The timely detection of malignancies is crucial for their effective treatment. In this regard, the examination of bone marrow smears constitutes a crucial step. Nonetheless, the conventional approach to cell identification and enumeration is laborious and time-intensive. Therefore, the present study aimed to develop a method for the efficient diagnosis of these malignancies directly from bone marrow microscopic images.

METHODS

A deep learning-based framework was developed to facilitate the diagnosis of common hematologic malignancies. First, a total of 2033 microscopic images of bone marrow analysis, including the images for 6 disease types and 1 healthy control, were collected from two Chinese medical websites. Next, the collected images were classified into the training, validation, and test datasets in the ratio of 7:1:2. Subsequently, a method of stain normalization to multi-domains (stain domain augmentation) based on the MultiPathGAN model was developed to equalize the stain styles and expand the image datasets. Afterward, a lightweight hybrid model named MobileViTv2, which integrates the strengths of both CNNs and ViTs, was developed for disease classification. The resulting model was trained and utilized to diagnose patients based on multiple microscopic images of their bone marrow smears, obtained from a cohort of 61 individuals.

RESULTS

MobileViTv2 exhibited an average accuracy of 94.28% when applied to the test set, with multiple myeloma, acute lymphocytic leukemia, and lymphoma revealed as the three diseases diagnosed with the highest accuracy values of 98%, 96%, and 96%, respectively. Regarding patient-level prediction, the average accuracy of MobileViTv2 was 96.72%. This model outperformed both CNN and ViT models in terms of accuracy, despite utilizing only 9.8 million parameters. When applied to two public datasets, MobileViTv2 exhibited accuracy values of 99.75% and 99.72%, respectively, and outperformed previous methods.

CONCLUSIONS

The proposed framework could be applied directly to bone marrow microscopic images with different stain styles to efficiently establish the diagnosis of common hematologic malignancies.