Characterizing Sentinel Lymph Node Status in Breast Cancer Patients Using a Deep-Learning Model Compared With Radiologists' Analysis of Grayscale Ultrasound and Lymphosonography.

ABSTRACT: The objective of the study was to use a deep learning model to differentiate between benign and malignant sentinel lymph nodes (SLNs) in patients with breast cancer compared to radiologists' assessments.Seventy-nine women with breast cancer were enrolled and underwent lymphosonography and contrast-enhanced ultrasound (CEUS) examination after subcutaneous injection of ultrasound contrast agent around their tumor to identify SLNs. Google AutoML was used to develop image classification model. Grayscale and CEUS images acquired during the ultrasound examination were uploaded with a data distribution of 80% for training/20% for testing. The performance metric used was area under precision/recall curve (AuPRC). In addition, 3 radiologists assessed SLNs as normal or abnormal based on a clinical established classification. Two-hundred seventeen SLNs were divided in 2 for model development; model 1 included all SLNs and model 2 had an equal number of benign and malignant SLNs. Validation results model 1 AuPRC 0.84 (grayscale)/0.91 (CEUS) and model 2 AuPRC 0.91 (grayscale)/0.87 (CEUS). The comparison between artificial intelligence (AI) and readers' showed statistical significant differences between all models and ultrasound modes; model 1 grayscale AI versus readers, P = 0.047, and model 1 CEUS AI versus readers, P < 0.001. Model 2 r grayscale AI versus readers, P = 0.032, and model 2 CEUS AI versus readers, P = 0.041.The interreader agreement overall result showed κ values of 0.20 for grayscale and 0.17 for CEUS.In conclusion, AutoML showed improved diagnostic performance in balance volume datasets. Radiologist performance was not influenced by the dataset's distribution.

Understanding nationwide readmissions after thyroid surgery.

BACKGROUND: The 30-day readmission rate is increasingly utilized as a metric of quality that impacts reimbursement. To date, there are no nationally representative data on readmission rates after thyroid surgery. We aimed to determine national readmission rates after inpatient thyroidectomy operations and whether select clinical factors were associated with increased odds of postthyroidectomy readmission. METHODS: Using the 2014 Nationwide Readmissions Database, we identified patients undergoing inpatient thyroid surgery as defined by the International Classification of Diseases, Ninth Revision, procedure codes for thyroid lobectomy, partial thyroidectomy, complete thyroidectomy, and substernal thyroidectomy. Descriptive statistics were used to report readmission rates, most common diagnosis and causes of readmission, and timing of presentation after discharge. Multivariable logistic regression models controlling for potential confounders were used to determine whether select factors were associated with 30-day readmission. RESULTS: A total of 22,654 patients underwent inpatient thyroid surgery during the study period, 990 of whom (4.4%) were readmitted within 30 days. Among these, the most common diagnoses during readmission were disorders of mineral metabolism and hypocalcemia, accounting for 36.0% and 26.6% of readmissions, respectively. This held true regardless of the apparent indication for thyroid surgery (goiter, cancer, or thyroid function disorder) or timing of readmission after discharge. Calcium-related abnormalities were the top diagnoses at readmissions (22.1%). Most readmissions (54.6%) occurred within 7 days of discharge, with 24.6% within the first 2 days Factors associated with an increased odds of readmission included having Medicare (adjusted odds ratio [AOR] 1.47 and 95% confidence interval [CI] 1.03-2.11) or Medicaid insurance (AOR 1.44 [CI 1.04-1.99]), being discharged to inpatient post acute care (AOR 2.31 [CI 1.48-3.62]) or to home health care (AOR 1.78 [CI 1.21-2.63]), having an Elixhauser comorbidity score ≥ 4 (AOR 2.04 [CI 1.27-3.26]), and a duration of stay ≥2 days after the thyroid surgery (AOR 2.7 [CI 1.9-3.82]). The only complication during index admission associated with increased odds of readmission was hypocalcemia (AOR 1.5 [CI 1.1-2.06]. Indications for thyroid surgery were not associated with increased odds of readmission. CONCLUSION: Readmissions after thyroid surgery are relatively low and occur early after surgery. The most common diagnoses identified on readmission were calcium and mineral metabolism disorders, which also were the most common cause of readmission. Socioeconomic factors, comorbidities, and complications during the index admissions were found to be associated with nonelective, postthyroidectomy readmissions. Recognition of these risk factors may guide the development of interventions and protocols to decrease readmissions.