Skin biopsy images can reveal causes and severity of many skin diseases, which is a significant complement for skin surface inspection. Automatic annotation of skin biopsy image is an important problem for increasing efficiency and reducing the subjectiveness in diagnosis. However it is challenging particularly when there exists indirect relationship between annotation terms and local regions of a biopsy image, as well as local structures with different textures. In this paper, a novel method based on a recent proposed machine learning model, named multi-instance multilabel (MIML), is proposed to model the potential knowledge and experience of doctors on skin biopsy image annotation. We first show that the problem of skin biopsy image annotation can naturally be expressed as a MIML problem and then propose an image representation method that can capture both region structure and texture features, and a sparse Bayesian MIML algorithm which can produce probabilities indicating the confidence of annotation. The proposed algorithm framework is evaluated on a real clinical dataset containing 12,700 skin biopsy images. The results show that it is effective and prominent.

As a significant complement for skin surface images, skin biopsy image may reveal causes and severity of many skin diseases, especially in the case of skin cancer inspection. With rapid increment of skin disease patients, computational methods have been introduced for automatic classification of skin images. However, due to the complex relationship among annotation terms and features of local regions, it becomes a great challenge for skin biopsy image feature recognition and annotation. In this paper, we attempt to model the potential knowledge and experience of doctors on skin biopsy image annotation by using a recent proposed machine learning model, named multi-instance multi-label (MIML) model. We show that the relationship among annotation terms and skin biopsy images is naturally consistent with the MIML framework. We further propose a sparse Bayesian MIML algorithm which can produce a probability indicating the confidence of annotating a term. The proposed algorithm framework is evaluated on a real dataset from a large local hospital containing 12,700 skin biopsy images. The results show that the proposed algorithm is effective and prominent.