Identifying out of distribution samples for skin cancer and malaria images. (September 2022)

• Record Type: Journal Article
• Authors: Zaid, Muhammad; Ali, Shafaqat; Ali, Mohsen; Hussein, Sarfaraz; Saadia, Asma; Sultani, Waqas
• Is Part Of: Biomedical signal processing and control. Volume 78(2022)
• Publication Date: 2022-09
• Subjects: Skin cancer -- Malaria -- Out of distribution -- Unsupervised approach -- Tuplet loss -- K-reciprocal neighbor; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.103882
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract: Deep neural networks have shown promising results in disease detection and classification using medical image data. However, they still suffer from the challenges of handling real-world scenarios especially reliably detecting out-of-distribution (OoD) samples. We propose an approach to robustly classify OoD samples in skin and malaria images without the need to access labeled OoD samples during training. Specifically, we use metric learning along with logistic regression to force the deep networks to learn much rich class representative features. To guide the learning process against the OoD examples, we generate in distribution's similar-looking examples by either removing class-specific salient regions in the image or permuting image parts and distancing them away from in-distribution (ID) samples. During inference time, the K-reciprocal nearest neighbor is employed to detect out-of-distribution samples. For skin cancer OoD detection, we employ two standard benchmark skin cancer ISIC datasets as ID, and six different datasets with varying difficulty levels were taken as out of distribution. For malaria OoD detection, we use the BBBC041 malaria dataset as ID and five different challenging datasets as out of distribution. We achieve state-of-the-art results, improving 5% and 4% in TNR @ TPR95% over the previous state-of-the-art OoD detection methods for skin cancer and malaria images while preserving the ID classification ability. Highlights: The approach does not have any hyperparameters to tune on the labeled OoD. Employed novel ways to generate OoD surrogates from ID samples. Does not require access to any real OoD during training. A unique way to improve the OoD score using K-reciprocal neighbors. Proposed approach has good results when compared with the baselines.