Multi-Learner Based Deep Meta-Learning for Few-Shot Medical Image Classification

A systematic review of few-shot learning in medical imaging

The lack of annotated medical images limits the performance of deep learning models, which usually need large-scale labelled datasets. Few-shot learning techniques can reduce data scarcity issues and enhance medical image analysis speed and robustness. This systematic review gives a comprehensive overview of few-shot learning methods for medical image analysis, aiming to establish a standard methodological pipeline for future research reference. With a particular emphasis on the role of meta-learning, we analysed 80 relevant articles published from 2018 to 2023, conducting a risk of bias assessment and extracting relevant information, especially regarding the employed learning techniques. From this, we delineated a comprehensive methodological pipeline shared among all studies. In addition, we performed a statistical analysis of the studies' results concerning the clinical task and the meta-learning method employed while also presenting supplemental information such as imaging modalities and model robustness evaluation techniques. We discussed the findings of our analysis, providing a deep insight into the limitations of the state-of-the-art methods and the most promising approaches. Drawing on our investigation, we yielded recommendations on potential future research directions aiming to bridge the gap between research and clinical practice.

A few-shot learning framework for the diagnosis of osteopenia and osteoporosis using knee X-ray images

OBJECTIVE

We developed a few-shot learning (FSL) framework for the diagnosis of osteopenia and osteoporosis in knee X-ray images.

METHODS

Computer vision models containing deep convolutional neural networks were fine-tuned to enable generalization from natural images (ImageNet) to chest X-ray images (normal vs. pneumonia, base images). Then, a series of automated machine learning classifiers based on the Euclidean distances of base images were developed to make predictions for novel images (normal vs. osteopenia vs. osteoporosis). The performance of the FSL framework was compared with that of junior and senior radiologists. In addition, the gradient-weighted class activation mapping algorithm was used for visual interpretation.

RESULTS

In Cohort #1, the mean accuracy (0.728) and sensitivity (0.774) of the FSL models were higher than those of the radiologists (0.512 and 0.448). A diagnostic pipeline of FSL model (first)-radiologists (second) achieved better performance (0.653 accuracy, 0.582 sensitivity, and 0.816 specificity) than radiologists alone. In Cohort #2, the diagnostic pipeline also showed improved performance.

CONCLUSIONS

The FSL framework yielded practical performance with respect to the diagnosis of osteopenia and osteoporosis in comparison with radiologists. This retrospective study supports the use of promising FSL methods in computer-aided diagnosis tasks involving limited samples.

A few-shot disease diagnosis decision making model based on meta-learning for general practice

BACKGROUND

Diagnostic errors have become the biggest threat to the safety of patients in primary health care. General practitioners, as the "gatekeepers" of primary health care, have a responsibility to accurately diagnose patients. However, many general practitioners have insufficient knowledge and clinical experience in some diseases. Clinical decision making tools need to be developed to effectively improve the diagnostic process in primary health care. The long-tailed class distributions of medical datasets are challenging for many popular decision making models based on deep learning, which have difficulty predicting few-shot diseases. Meta-learning is a new strategy for solving few-shot problems.

METHODS AND MATERIALS

In this study, a few-shot disease diagnosis decision making model based on a model-agnostic meta-learning algorithm (FSDD-MAML) is proposed. The MAML algorithm is applied in a knowledge graph-based disease diagnosis model to find the optimal model parameters. Moreover, FSDD-MAML can learn learning rates for all modules of the knowledge graph-based disease diagnosis model. For n-way, k-shot learning tasks, the inner loop of FSDD-MAML performs multiple gradient update steps to learn internal features in disease classification tasks using n×k examples, and the outer loop of FSDD-MAML optimizes the meta-objective to find the associated optimal parameters and learning rates. FSDD-MAML is compared with the original knowledge graph-based disease diagnosis model and other meta-learning algorithms based on an abdominal disease dataset.

RESULT

Meta-learning algorithms can greatly improve the performance of models in top-1 evaluation compared with top-3, top-5, and top-10 evaluations. The proposed decision making model FSDD-MAML outperforms all the other models, with a precision@1 of 90.02 %. We achieve state-of-the-art performance in the diagnosis of all diseases, and the prediction performance for few-shot diseases is greatly improved. For the two groups with the fewest examples of diseases, FSDD-MAML achieves relative increases in precision@1 of 29.13 % and 21.63 % compared with the original knowledge graph-based disease diagnosis model. In addition, we analyze the reasoning process of several few-shot disease predictions and provide an explanation for the results.

CONCLUSION

The decision making model based on meta-learning proposed in this paper can support the rapid diagnosis of diseases in general practice and is especially capable of helping general practitioners diagnose few-shot diseases. This study is of profound significance for the exploration and application of meta-learning to few-shot disease assessment in general practice.

Dynamic feature splicing for few-shot rare disease diagnosis

Annotated images for rare disease diagnosis are extremely hard to collect. Therefore, identifying rare diseases under a few-shot learning (FSL) setting is significant. Existing FSL methods transfer useful and global knowledge from base classes with abundant training samples to enrich features of novel classes with few training samples, but still face difficulties when being applied to medical images due to the complex lesion characteristics and large intra-class variance. In this paper, we propose a dynamic feature splicing (DNFS) framework for few-shot rare disease diagnosis. Under DNFS, both low-level features (i.e., the output of three convolutional blocks) and high-level features (i.e., the output of the last fully connected layer) of novel classes are dynamically enriched. We construct the position coherent DNFS (P-DNFS) module to perform low-level feature splicing, where a lesion-oriented Transformer is designed to detect lesion regions. Thus, novel-class channels are replaced by similar base-class channels within the detected lesion regions to achieve disease-related feature enrichment. We also devise a semantic coherent DNFS (S-DNFS) module to perform high-level feature splicing. It explores cross-image channel relations and selects base-class channels with semantic consistency for explicit knowledge transfer. Both low-level and high-level feature splicings are performed dynamically and iteratively. Consequently, abundant spliced features are generated for disease diagnosis, leading to more accurate decision boundary and improved diagnosis performance. Extensive experiments have been conducted on three medical image classification datasets. Our results suggest that the proposed DNFS achieves superior performance against state-of-the-art approaches.

A visual transformer-based smart textual extraction method for financial invoices

In era of big data, the computer vision-assisted textual extraction techniques for financial invoices have been a major concern. Currently, such tasks are mainly implemented via traditional image processing techniques. However, they highly rely on manual feature extraction and are mainly developed for specific financial invoice scenes. The general applicability and robustness are the major challenges faced by them. As consequence, deep learning can adaptively learn feature representation for different scenes and be utilized to deal with the above issue. As a consequence, this work introduces a classic pre-training model named visual transformer to construct a lightweight recognition model for this purpose. First, we use image processing technology to preprocess the bill image. Then, we use a sequence transduction model to extract information. The sequence transduction model uses a visual transformer structure. In the stage target location, the horizontal-vertical projection method is used to segment the individual characters, and the template matching is used to normalize the characters. In the stage of feature extraction, the transformer structure is adopted to capture relationship among fine-grained features through multi-head attention mechanism. On this basis, a text classification procedure is designed to output detection results. Finally, experiments on a real-world dataset are carried out to evaluate performance of the proposal and the obtained results well show the superiority of it. Experimental results show that this method has high accuracy and robustness in extracting financial bill information.

Discriminative ensemble meta-learning with co-regularization for rare fundus diseases diagnosis

Deep neural networks (DNNs) have been widely applied in the medical image community, contributing to automatic ophthalmic screening systems for some common diseases. However, the incidence of fundus diseases patterns exhibits a typical long-tailed distribution. In clinic, a small number of common fundus diseases have sufficient observed cases for large-scale analysis while most of the fundus diseases are infrequent. For these rare diseases with extremely low-data regimes, it is challenging to train DNNs to realize automatic diagnosis. In this work, we develop an automatic diagnosis system for rare fundus diseases, based on the meta-learning framework. The system incorporates a co-regularization loss and the ensemble-learning strategy into the meta-learning framework, fully leveraging the advantage of multi-scale hierarchical feature embedding. We initially conduct comparative experiments on our newly-constructed lightweight multi-disease fundus images dataset for the few-shot recognition task (namely, FundusData-FS). Moreover, we verify the cross-domain transferability from miniImageNet to FundusData-FS, and further confirm our method's good repeatability. Rigorous experiments demonstrate that our method can detect rare fundus diseases, and is superior to the state-of-the-art methods. These investigations demonstrate that the potential of our method for the real clinical practice is promising.