Evaluation of significantly changed chemokine factors of idiopathic granulomatous mastitis in non-puerperal patients.

Idiopathic granulomatous mastitis (IGM), a recurrent inflammation disease of the non-lactating breast, has had an increasing clinical morbidity rate in recent years, and its complicated symptoms and unclear etiology make it challenging to treat. This rare benign inflammatory breast disease, centered on the lobules, represents the most challenging type of non-puerperal mastitis (NPM), also known as non-lactating mastitis. In this study, patients diagnosed with IGM (M, n = 23) were recruited as cases, and patients with benign control breast disease (C, n = 17) were enrolled as controls. Cytokine microarray detection measured and analyzed the differentially expressed cytokine factors between IGM and control patients. Then, we verified the mRNA and protein expression levels of the significantly changed cytokine factors using Q-RT-PCR, ELISA, western blot, and IHC experiments. The cytokine factor expression levels significantly changed compared to the control group. We observed a significant increase between IGM and control patients in cytokine factors expression, such as interleukin-1ß (IL-1ß), monokine induced by gamma interferon (MIG), macrophage inflammatory protein (MIP)-1α, MIP-1ß, tumor necrosis factor receptor 2 (TNF RII). Then, we verified the expression of these top five dysregulated factors in both mRNA and protein levels. Our results demonstrated the cytokine map in IGM and indicated that several cytokines, especially chemokines, were associated with and significantly dysregulated in IGM tissues compared to the control group. The chemokine factors involved might be essential in developing and treating IGM. These findings would be helpful for a better understanding of IGM and offer valuable insights for devising novel diagnostic and therapeutic strategies.

Automatic ICD-10-CM coding via Lambda-Scaled attention based deep learning model.

The International Classification of Diseases (ICD) serves as a global healthcare administration standard, with one of its editions being ICD-10-CM, an enhanced diagnostic classification system featuring numerous new codes for specific anatomic sites, co-morbidities, and causes. These additions facilitate conveying the complexities of various diseases. Currently, ICD-10 coding is widely adopted worldwide. However, public hospitals in Pakistan have yet to implement it and automate the coding process. In this research, we implemented ICD-10-CM coding for a private database and named it Clinical Pool of Liver Transplant (CPLT). Additionally, we proposed a novel deep learning model called Deep Recurrent-Convolution Neural Network with a lambda-scaled Attention module (DRCNN-ATT) using the CPLT database to achieve automatic ICD-10-CM coding. DRCNN-ATT combines a bi-directional long short-term memory network (bi-LSTM), a multi-scale convolutional neural network (MS-CNN), and a lambda-scaled attention module. Experimental results demonstrate that deep recurrent convolutional neural network (DRCNN) faces attention score vanishing problem with a standard attention module for automatic ICD coding. However, adding a lambda-scaled attention module resolves this issue. We evaluated DRCNN-ATT model using two distinct datasets: a private CPLT dataset and a public MIMIC III top 50 dataset. The results indicate that the DRCNN-ATT model outperformed various baselines by generating 0.862 micro F1 and 0.25 macro F1 scores on CPLT dataset and 0.705 micro F1 and 0.655 macro F1 scores on MIMIC III top 50 dataset. Furthermore, we also deployed our model for automatic ICD-10-CM coding using ngrok and the Flask APIs, which receives input, processes it, and then returns the results.

Retrieve and rerank for automated ICD coding via Contrastive Learning.

Automated ICD coding is a multi-label prediction task aiming at assigning patient diagnoses with the most relevant subsets of disease codes. In the deep learning regime, recent works have suffered from large label set and heavy imbalance distribution. To mitigate the negative effect in such scenarios, we propose a retrieve and rerank framework that introduces the Contrastive Learning (CL) for label retrieval, allowing the model to make more accurate prediction from a simplified label space. Given the appealing discriminative power of CL, we adopt it as the training strategy to replace the standard cross-entropy objective and retrieve a small subset by taking the distance between clinical notes and ICD codes into account. After properly training, the retriever could implicitly capture the code co-occurrence, which makes up for the deficiency of cross-entropy assigning each label independently of the others. Further, we evolve a powerful model via a Transformer variant for refining and reranking the candidate set, which can extract semantically meaningful features from long clinical sequences. Applying our method on well-known models, experiments show that our framework provides more accurate results guaranteed by preselecting a small subset of candidates before fine-level reranking. Relying on the framework, our proposed model achieves 0.590 and 0.990 in terms of Micro-F1 and Micro-AUC on benchmark MIMIC-III.

Dermis-retained breast dermo-glandular flap: a new surgical approach for granulomatous lobular mastitis.

Background: Granulomatous lobular mastitis (GLM) is characterized by nonspecific chronic inflammation concentrated in breast lobules. Surgical resection is one of the most common treatment options for GLM. On the basis of our previous use of Breast Dermo-Glandular Flap (BDGF), we designed a new surgical approach for GLM, especially for cases where the focus is close to the nipple. Here we describe this new treatment approach. Methods: In Peking Union Medical College Hospital (PUMCH) and Beijing Dangdai Hospital during January 2020-June 2021, we enrolled all 18 GLM patients who underwent surgery with the use of Dermis-Retained BDGF. All patients were women; most of the patients were 18-50 years old (88%); and the most common clinical manifestation of GLM was breast mass (60%). Then, we collected and analyzed data about the surgery and outcomes (drainage tubes moving time, relapse, patients' shape satisfaction). We regarded GLM recurrence on the same side as relapse. If there was no complication and the patient's satisfaction was excellent or good, we rated the surgery as successful. We recorded the occurrence of all common postsurgical complications of the breast. Results: The debridement area was 3-5.5 (4.3 ± 0.7) cm; surgery time was 78-119 (95.6 ± 11.6) min; and mean debridement time (27.8 ± 8.9 min) was shorter than the time to obtain and transplant the flap (47.5 ± 12.9 min). Blood loss was less than 139 ml. As for bacterial culture, two patients had positive results, but they had no symptoms. No surgery-related complications happened. In terms of the outcomes, all of the drainage tubes were removed in less than 5 days, and only one patient experienced relapse after 1 year of surgery during the follow-up. The patients' satisfaction with the breast shape was as follows: excellent (50%), good (22%), acceptable (22%), and poor (6%). Conclusion: For GLM patients refractory to conservative therapy or former unsatisfactory surgical management whose lesion is in the vicinity of the nipple and larger than 3 cm, Dermis-Retained BDGF is a suitable approach to fill the after-debridement defect below the nipple-areola and achieve a relatively satisfactory cosmetic outcome.

A systematic surgical approach for the treatment of idiopathic granulomatous mastitis: a case series.

BACKGROUND: Surgical resection can be performed for idiopathic granulomatous mastitis (IGM), but recurrence and tissue defects remain issues. Here we report our 6-year experience with a four-pattern surgical approach for IGM that involves the use of a random breast dermo-glandular flap (BDGF). METHODS: Sixty-eight consecutive patients with IGM were prospectively enrolled from 01/2012 and 03/2017. Based on the extent, shape, and location of the lesions, four different patterns of surgery based on BDGF were used to remove the lesion and repair the defect. Operative data (time, blood loss, and intraoperative complications), primary healing time, recurrence, and patient-reported outcomes (cosmetic outcome, and improvement in dressing change and bathing) at 2 years were evaluated. RESULTS: Patients' median age was 35 (range, 22-55) years. Duration of IGM was 3-22 months, with a median lesion size of 3.5 (range, 0.9-9.1) cm. The operative time was significantly longer, and blood loss was more important with the increasing lesion size (both P<0.05). No significant intraoperative complications occurred. All wounds healed by primary intention. IGM relapsed in three patients (3/68, 4.4%); they were treated successfully with a second operation. The self-evaluated cosmetic outcome was "much better" in 45 patients (66.2%), "a little better" in 18 (26.5%), and "same or worse" in five (7.3%). The self-evaluated improvement in dressing change and bathing was "improved a lot" in 51 patients (75.0%), "improved a little" in 11 (16.2%), and "not improved or getting worse" in six (8.8%). CONCLUSIONS: The BDGF-based systematic four-pattern surgical approach is effective in the treatment of IGM. Recurrence rate is low, there are no complications, and the cosmetic results and improvement in dressing change and bathing are generally favourable.