Clinical significance of abdominal computed tomography and colonoscopy in the evaluation of phlebosclerotic colitis.

Clinical manifestations of phlebosclerotic colitis (PC) exhibit significant variability, necessitating diverse treatment strategies depending on disease severity. However, there is limited research exploring the relationship between imaging findings and disease severity. Hence, this retrospective study aimed to analyze the correlation between computed tomography (CT) findings, colonoscopic features, and disease severity. This study compared the abdominal CT characteristics, colonoscopy findings, and treatment modalities of 45 PC patients. CT images were assessed for the severity of mesenteric venous calcification, maximum colonic wall thickness, number of involved colonic segments, and presence of pericolic inflammation. Colonoscopic images were assessed for dark purple discoloration mucosa, erosive and ulcerative lesions, mucosal edema, luminal narrowing, and the number of involved colonic segments. In addition, patients were categorized into three groups: the observation (n = 15), medical treatment (n = 19), and operation (n = 11) groups. In CT images, a significant difference in pericolic inflammation (p = 0.039) was observed among groups. Further, significant differences in dark purple discoloration mucosa (p = 0.033), erosive or ulcerative lesions (p < 0.001), mucosal edema (p < 0.001), luminal narrowing (p = 0.012), and the number of involved colonic segments (p = 0.001) were observed in colonoscopy. Moreover, we found positive correlations between CT and colonoscopy features. In conclusion, CT manifestations and colonoscopy findings exhibited correlation with disease severity in PC. When limited to one diagnostic tool, observations from that tool can infer potential manifestations of the alternative tool.

Deep Learning Approach for Negation and Speculation Detection for Automated Important Finding Flagging and Extraction in Radiology Report: Internal Validation and Technique Comparison Study.

BACKGROUND: Negation and speculation unrelated to abnormal findings can lead to false-positive alarms for automatic radiology report highlighting or flagging by laboratory information systems. OBJECTIVE: This internal validation study evaluated the performance of natural language processing methods (NegEx, NegBio, NegBERT, and transformers). METHODS: We annotated all negative and speculative statements unrelated to abnormal findings in reports. In experiment 1, we fine-tuned several transformer models (ALBERT [A Lite Bidirectional Encoder Representations from Transformers], BERT [Bidirectional Encoder Representations from Transformers], DeBERTa [Decoding-Enhanced BERT With Disentangled Attention], DistilBERT [Distilled version of BERT], ELECTRA [Efficiently Learning an Encoder That Classifies Token Replacements Accurately], ERNIE [Enhanced Representation through Knowledge Integration], RoBERTa [Robustly Optimized BERT Pretraining Approach], SpanBERT, and XLNet) and compared their performance using precision, recall, accuracy, and F1-scores. In experiment 2, we compared the best model from experiment 1 with 3 established negation and speculation-detection algorithms (NegEx, NegBio, and NegBERT). RESULTS: Our study collected 6000 radiology reports from 3 branches of the Chi Mei Hospital, covering multiple imaging modalities and body parts. A total of 15.01% (105,755/704,512) of words and 39.45% (4529/11,480) of important diagnostic keywords occurred in negative or speculative statements unrelated to abnormal findings. In experiment 1, all models achieved an accuracy of >0.98 and F1-score of >0.90 on the test data set. ALBERT exhibited the best performance (accuracy=0.991; F1-score=0.958). In experiment 2, ALBERT outperformed the optimized NegEx, NegBio, and NegBERT methods in terms of overall performance (accuracy=0.996; F1-score=0.991), in the prediction of whether diagnostic keywords occur in speculative statements unrelated to abnormal findings, and in the improvement of the performance of keyword extraction (accuracy=0.996; F1-score=0.997). CONCLUSIONS: The ALBERT deep learning method showed the best performance. Our results represent a significant advancement in the clinical applications of computer-aided notification systems.