Rationale: Cholesterol granuloma of the breast is a rare disease defined as chronic reactive inflammation of cholesterol crystals and foreign body giant cells. This disease can mimic breast cancer in the clinic as painless palpable hard nodules, and imaging shows irregular hypoechoic nodules with unclear boundaries. Therefore, the uncommon lesions can be easily misdiagnosed as breast cancer. Meanwhile, it can be easily neglected by clinicians because of poor understanding.Patient concerns: In this report, we present a rare case of multiple cholesterol granulomas of the breast, which was analyzed retrospectively and combined with all 11 relevant available studies in the last 50 years. Interventions:The patient had undergone multiple breast imaging inspections for breast nodules and had the local resection of nodules.Diagnoses: The patient was confirmed to have a final diagnosis of benign cholesterol granulomas but was initially considered as breast cancer.Outcomes: The patient did not complain of discomfort after surgery, and ultrasound reexamination 5 months after surgery showed no recurrence.Lessons: By retrospective analysis, dynamic contrast-enhanced magnetic resonance imaging and core needle biopsy can synergistically help clinicians distinguish it from other breast disease. To raise awareness of such a rare disease and reduce related misdiagnoses, we summarize the characteristics of cholesterol granulomas and recommend appropriate novel diagnosis and treatment regimens for patients with cholesterol granulomas.

Since Correlation Filter appeared in the field of video object tracking, it is very popular due to its excellent performance. The Correlation Filter-based tracking algorithms are very competitive in terms of accuracy and speed as well as robustness. However, there are still some fields for improvement in the Correlation Filter-based tracking algorithms. First, during the training of the classifier, the background information that can be utilized is very limited. Moreover, the introduction of the cosine window further reduces the background information. These reasons reduce the discriminating power of the classifier. This paper introduces more global background information on the basis of the DCF tracker to improve the discriminating ability of the classifier. Then, in some complex scenes, tracking loss is easy to occur. At this point, the tracker will be treated the background information as the object. To solve this problem, this paper introduces a novel re-detection component. Finally, the current Correlation Filter-based tracking algorithms use the linear interpolation model update method, which cannot adapt to the object changes in time. This paper proposes an adaptive model update strategy to improve the robustness of the tracker. The experimental results on multiple datasets can show that the tracking algorithm proposed in this paper is an excellent algorithm.

Since Correlation Filter appeared in the field of video object tracking, it is great popular due to its excellent performance. The Correlation Filter based tracking algorithms are very competitive in terms of accuracy and speed as well as robustness. However, there are still some fields for improvement in the Correlation Filter based tracking algorithms. First, during the training of the classifier, the background information that can be utilized is very limited. Moreover, the introduction of the cosine window further reduces the background information. These reasons reduce the discriminating power of the classifier. This paper introduces more global background information on the basis of the DCF tracker to improve the discriminating ability of the classifier. Then, in some complex scenes, tracking loss is easy to occur. At this point, the tracker will be treated the background information as the object. To solve this problem, this paper proposes a novel re-detection component. Finally, the current Correlation Filter based tracking algorithms use the linear interpolation model update method, which cannot adapt to the object changes in time. This paper proposes an adaptive model update strategy to improve the robustness of the tracker.

No abstract

Fibroadenomas (FAs) are the most common breast tumors in women. No pharmacological agents are currently approved for FA intervention owing to its unclear mechanisms and a shortage of reproducible human models. Here, using single-cell RNA sequencing of human FAs and normal breast tissues, we observe distinct cellular composition and epithelial structural changes in FAs. Interestingly, epithelial cells exhibit hormone-responsive functional signatures and synchronous activation of estrogen-sensitive and hormone-resistant mechanisms (ERBB2, BCL2 and CCND1 pathways). We develop a human expandable FA organoid system and observe that most organoids seem to be resistant to tamoxifen. Individualized combinations of tamoxifen with ERBB2, BCL2 or CCND1 inhibitors could significantly suppress the viability of tamoxifen-resistant organoids. Thus, our study presents an overview of human FA at single-cell resolution that outlines the structural and functional differences between FA and normal breast epithelium and, in particular, provides a potential therapeutic strategy for breast FAs.