Khurana, Aditya

Abstract Objective To compare the pathology outcomes of image-guided breast biopsies across a multistate health care enterprise by the method of detection (MOD), including the positive predictive value (PPV3) and histological types. Methods Biopsies from October 31, 2017, to July 6, 2023, across 18 imaging sites in 4 states were extracted. Radiologists prospectively assigned MODs, including symptomatic presentations (self-examination, clinical examination) and imaging modalities (mammogram, tomosynthesis, US, MRI, contrast-enhanced mammogram, molecular breast imaging, other imaging). For each MOD, the rates of malignancy and cancer subtypes were calculated. Univariate and multivariate logistic regression analyses were performed. Results In 26 932 biopsies on 22 582 patients (mean age 57.6 years ± 14.2 [SD]; 22 416 women), 6068 (81.1%) cancers were detected by screening, and 1144 (15.3%) were detected by symptoms. Symptomatic presentation MODs had a higher PPV3 (43.0% vs 25.6%; P < .001), larger cancer lesion sizes (x̄ = 27.5 mm vs x̄ = 17.8 mm; P < 001), and a higher association with invasive carcinoma (96.7% vs 76.8%; P <.001) compared with screening imaging-based MODs. Methods of detection with higher PPV3s also tended to have higher rates of invasive carcinoma (Spearman’s ρ = 0.83). Conclusion Compared with breast cancers detected after symptomatic presentations, those identified via image-based screening were smaller and less often invasive, confirming the value of screening in early detection. Positive predictive value and the rate of invasive carcinoma varied in parallel across MODs.

Abstract Objectives The surge in patient portal messages (PPMs) with increasing needs and workloads for efficient PPM triage in healthcare settings has spurred the exploration of AI-driven solutions to streamline the healthcare workflow processes, ensuring timely responses to patients to satisfy their healthcare needs. However, there has been less focus on isolating and understanding patient primary concerns in PPMs—a practice which holds the potential to yield more nuanced insights and enhances the quality of healthcare delivery and patient-centered care. Materials and Methods We propose a fusion framework to leverage pretrained language models (LMs) with different language advantages via a Convolution Neural Network for precise identification of patient primary concerns via multi-class classification. We examined 3 traditional machine learning models, 9 BERT-based language models, 6 fusion models, and 2 ensemble models. Results The outcomes of our experimentation underscore the superior performance achieved by BERT-based models in comparison to traditional machine learning models. Remarkably, our fusion model emerges as the top-performing solution, delivering a notably improved accuracy score of 77.67 ± 2.74% and an F1 score of 74.37 ± 3.70% in macro-average. Discussion This study highlights the feasibility and effectiveness of multi-class classification for patient primary concern detection and the proposed fusion framework for enhancing primary concern detection. Conclusions The use of multi-class classification enhanced by a fusion of multiple pretrained LMs not only improves the accuracy and efficiency of patient primary concern identification in PPMs but also aids in managing the rising volume of PPMs in healthcare, ensuring critical patient communications are addressed promptly and accurately.

Background . Patients increasingly use asynchronous communication platforms to converse with care teams. Natural language processing (NLP) to classify content and automate triage of these messages has great potential to enhance clinical efficiency. We characterize the contents of a corpus of portal messages generated by patients using NLP methods. We aim to demonstrate descriptive analyses of patient text that can contribute to the development of future sophisticated NLP applications. Methods . We collected approximately 3,000 portal messages from the cardiology, dermatology, and gastroenterology departments at Mayo Clinic. After labeling these messages as either Active Symptom, Logistical, Prescription, or Update, we used NER (named entity recognition) to identify medical concepts based on the UMLS library. We hierarchically analyzed the distribution of these messages in terms of departments, message types, medical concepts, and keywords therewithin. Results . Active Symptom and Logistical content types comprised approximately 67% of the message cohort. The “Findings” medical concept had the largest number of keywords across all groupings of content types and departments. “Anatomical Sites” and “Disorders” keywords were more prevalent in Active Symptom messages, while “Drugs” keywords were most prevalent in Prescription messages. Logistical messages tended to have the lower proportions of “Anatomical Sites,”, “Disorders,”, “Drugs,”, and “Findings” keywords when compared to other message content types. Conclusions . This descriptive corpus analysis sheds light on the content and foci of portal messages. The insight into the content and differences among message themes can inform the development of more robust NLP models.

AbstractObjectiveThe content of websites for fellowship programs is an important source of information for residents applying to breast imaging fellowship programs (BIFPs). The purpose of this study is to evaluate the comprehensiveness of online content of BIFPs.MethodsA list of BIFPs was obtained from the Society of Breast Imaging website. Each program’s website was evaluated for the presence of 19 training-relevant content variables. Impact of program characteristics on comprehensiveness scores was determined. For statistical analysis, Kruskal–Wallis tests were used to assess differences in comprehensiveness scores based upon region, and two-tailed t-tests were used to compare based upon program size.ResultsA total of 79 BIFP websites were analyzed. The mean comprehensiveness score of BIFP websites based on meeting the 19 criteria was 44.1% (8.4 ± 2.7/19). Program coordinator contact information (72/79, 91.1%), application requirements (71/79, 89.9%), and faculty information (56/79, 70.9%) appeared on >70% of websites. The majority of fellowships had a dedicated webpage for their program (71/79, 89.9%). Information regarding 12 of the 19 criteria appeared on fewer than 50% of websites. Program region (P = 0.32) and size (P = 0.16) were not associated with any differences in mean comprehensiveness score. Additionally, there was no significant difference in scores associated with filling all available positions for the 2020 match cycle (P = 0.77).ConclusionThere is a paucity of information commonly sought out by applicants on the websites of most BIFPs. Both programs and applicants may mutually benefit from increasing comprehensive online content.