Recognition of rare antinuclear antibody patterns based on a novel attention-based enhancement framework.

Rare antinuclear antibody (ANA) pattern recognition has been a widely applied technology for routine ANA screening in clinical laboratories. In recent years, the application of deep learning methods in recognizing ANA patterns has witnessed remarkable advancements. However, the majority of studies in this field have primarily focused on the classification of the most common ANA patterns, while another subset has concentrated on the detection of mitotic metaphase cells. To date, no prior research has been specifically dedicated to the identification of rare ANA patterns. In the present paper, we introduce a novel attention-based enhancement framework, which was designed for the recognition of rare ANA patterns in ANA-indirect immunofluorescence images. More specifically, we selected the algorithm with the best performance as our target detection network by conducting comparative experiments. We then further developed and enhanced the chosen algorithm through a series of optimizations. Then, attention mechanism was introduced to facilitate neural networks in expediting the learning process, extracting more essential and distinctive features for the target features that belong to the specific patterns. The proposed approach has helped to obtained high precision rate of 86.40%, 82.75% recall, 84.24% F1 score and 84.64% mean average precision for a 9-category rare ANA pattern detection task on our dataset. Finally, we evaluated the potential of the model as medical technologist assistant and observed that the technologist's performance improved after referring to the results of the model prediction. These promising results highlighted its potential as an efficient and reliable tool to assist medical technologists in their clinical practice.

Development of a scoring system for predicting the severity of ulcerative colitis.

BACKGROUND AND STUDY AIMS: Monitoring disease activity in ulcerative colitis (UC) is critical in preventing long-term complications. This study aims to develop a scoring system using non-invasive indicators to predict endoscopic activities for ulcerative colitis (UC) patients. PATIENTS AND METHODS: All enrolled patients with UC admitted to Shanghai Xinhua Hospital between June 2017 and January 2021 were enrolled, and their clinical data were retrospectively collected and a number of serological biomarkers concentrations were analyzed. Patients were categorized into mild and moderate-to-severe disease groups. Univariate and multivariate logistic regression was used to predict moderate-to-severe endoscopic activities, which were then incorporated into a nomogram to establish a prediction scoring model. RESULT: Overall, 231 patients were divided into a mild group (n = 111, 48.0%) and a moderate-to-severe group (n = 120, 52.0%). The following variables were independently associated with the disease severity and were subsequently included into the prediction model: Proteinase 3 antineutrophil cytoplasmic antibody (PR3-ANCA), C-reactive protein (CRP), hemoglobin(Hb), IL-10, stool frequency ≥ 5 times/day and hematochezia. Incorporating these 6 factors, the nomogram showed good discrimination with C-index of 0.819 and reliable calibration. A scoring model was established with the area under the curve 0.818. Moreover, PR3-ANCA and CRP correlated with the duration of hospital stay. CONCLUSION: We developed a predictive model for endoscopic disease activities by using noninvasive factors based on PR3-ANCA, CRP, Hb, IL-10, stool frequency and hematochezia. This prediction model might assist clinicians in managing patients with UC.