A novel image-based retrieval system for characterization of maxillofacial lesions in cone beam CT images

Abstract

Purpose

The objective of medical content-based image retrieval (CBIR) is to assist clinicians in decision making by retrieving the most similar cases to a given query image from a large database. Herein, a new method for content-based image retrieval of cone beam CT (CBCT) scans is presented.

Methods

The introduced framework consists of two main phases: training database construction and querying. The goal of the training phase is database construction, which consists of three main steps. First, automatic segmentation of lesions using 3D symmetry analysis is performed. Embedding the prior shape knowledge of the 3D symmetry characteristics of the healthy human head structure increases the accuracy of automatic segmentation. Then, spatial pyramid matching is used for feature extraction, and the relative importance of each feature is learned using classifiers.

Results

The method was applied to a dataset of 1145 volumetric CBCT images with four classes of maxillofacial lesions. A symmetry-based analysis model for automatic lesion segmentation was evaluated using similarity measures. Mean Dice coefficients of 0.89, 0.85, 0.92, and 0.87 were achieved for maxillary sinus perforation, radiolucent lesion, unerupted tooth, and root fracture classes, respectively. Moreover, the execution time of automatic segmentation was reduced to 3 min per case. The performance of the proposed search engine was evaluated using mean average precision and normalized discounted cumulative gain. A mean average retrieval accuracy and normalized discounted cumulative gain of 0.90 and 0.92, respectively, were achieved.

Conclusion

Quantitative results show that the proposed approach is more effective than previous methods in the literature, and it can facilitate the introduction of CBIR in clinical CBCT applications.

Appendix: Performance evaluation metrics

In the context of CBIR systems, normalized discounted cumulative gain (NDCG) [32] is one of the most frequent measures for retrieval effectiveness. It can be utilized to quantify the ranking of retrieved cases. At first, discounted cumulative gain (DCG) is used to rank the position q of a retrieved result:

$$ {\text{DCG}}_{q} = \sum\limits_{i = 1}^{M} {\frac{{2^{{{\text{rel}}(i)}} - 1}}{{\log_{2} (i + 1)}}} , $$

(10)

where rel_i is the graded dissimilarity of the retrieved result at position i in the ranked list and M is the total number of cases. For each query, normalized discounted cumulative gain (NDCG) is defined as follows:

$$ {\text{NDCG}}_{q} = \frac{{{\text{DCG}}_{q} }}{{{\text{IDCG}}_{q} }} , $$

(11)

where IDCG is ideal discounted cumulative gain and it is computed as:

$$ {\text{IDCG}}_{q} = \sum\limits_{i = 1}^{{N_{\text{rel}} }} {\frac{{2^{{{\text{rel}}_{i} }} - 1}}{{\log_{2} (i + 1)}}} , $$

(12)

and $N_{\text{rel}}$ is the list of relevant images to position q which are ordered by the relevance.