Development and internal validation of diagnostic prediction models using machine-learning algorithms in dogs with hypothyroidism.

Introduction: Hypothyroidism can be easily misdiagnosed in dogs, and prediction models can support clinical decision-making, avoiding unnecessary testing and treatment. The aim of this study is to develop and internally validate diagnostic prediction models for hypothyroidism in dogs by applying machine-learning algorithms. Methods: A single-institutional cross-sectional study was designed searching the electronic database of a Veterinary Teaching Hospital for dogs tested for hypothyroidism. Hypothyroidism was diagnosed based on suggestive clinical signs and thyroid function tests. Dogs were excluded if medical records were incomplete or a definitive diagnosis was lacking. Predictors identified after data processing were dermatological signs, alopecia, lethargy, hematocrit, serum concentrations of cholesterol, creatinine, total thyroxine (tT4), and thyrotropin (cTSH). Four models were created by combining clinical signs and clinicopathological variables expressed as quantitative (models 1 and 2) and qualitative variables (models 3 and 4). Models 2 and 4 included tT4 and cTSH, models 1 and 3 did not. Six different algorithms were applied to each model. Internal validation was performed using a 10-fold cross-validation. Apparent performance was evaluated by calculating the area under the receiver operating characteristic curve (AUROC). Results: Eighty-two hypothyroid and 233 euthyroid client-owned dogs were included. The best performing algorithms were naive Bayes in model 1 (AUROC = 0.85; 95% confidence interval [CI] = 0.83-0.86) and in model 2 (AUROC = 0.98; 95% CI = 0.97-0.99), logistic regression in model 3 (AUROC = 0.88; 95% CI = 0.86-0.89), and random forest in model 4 (AUROC = 0.99; 95% CI = 0.98-0.99). Positive predictive value was 0.76, 0.84, 0.93, and 0.97 in model 1, 2, 3, and 4, respectively. Negative predictive value was 0.89, 0.89, 0.99, and 0.99 in model 1, 2, 3, and 4, respectively. Discussion: Machine learning-based prediction models were accurate in predicting and quantifying the likelihood of hypothyroidism in dogs based on internal validation performed in a single-institution, but external validation is required to support the clinical applicability of these models.

Prediction of vascular invasion using a 7-point scale computed tomography grading system in adrenal tumors in dogs.

BACKGROUND: Previous studies evaluating the accuracy of computed tomography (CT) in detecting caudal vena cava (CVC) invasion by adrenal tumors (AT) used a binary system and did not evaluate for other vessels. OBJECTIVE: Test a 7-point scale CT grading system for accuracy in predicting vascular invasion and for repeatability among radiologists. Build a decision tree based on CT criteria to predict tumor type. METHODS: Retrospective observational cross-sectional case study. Abdominal CT studies were analyzed by 3 radiologists using a 7-point CT grading scale for vascular invasion and by 1 radiologist for CT features of AT. ANIMALS: Dogs with AT that underwent adrenalectomy and had pre- and postcontrast CT. RESULTS: Ninety-one dogs; 45 adrenocortical carcinomas (50%), 36 pheochromocytomas (40%), 9 adrenocortical adenomas (10%) and 1 unknown tumor. Carcinoma and pheochromocytoma differed in pre- and postcontrast attenuation, contralateral adrenal size, tumor thrombus short- and long-axis, and tumor and thrombus mineralization. A decision tree was built based on these differences. Adenoma and malignant tumors differed in contour irregularity. Probability of vascular invasion was dependent on CT grading scale, and a large equivocal zone existed between 3 and 6 scores, lowering CT accuracy to detect vascular invasion. Radiologists' agreement for detecting abnormalities (evaluated by chance-corrected weighted kappa statistics) was excellent for CVC and good to moderate for other vessels. The quality of postcontrast CT study had a negative impact on radiologists' performance and agreement. CONCLUSIONS AND CLINICAL IMPORTANCE: Features of CT may help radiologists predict AT type and provide probabilistic information on vascular invasion.

Comparison between a flash glucose monitoring system and a portable blood glucose meter for monitoring dogs with diabetes mellitus.

BACKGROUND: Flash glucose monitoring system (FGMS; FreeStyle Libre) was recently validated for use in diabetic dogs (DD). It is not known if this system is clinically useful in monitoring DD. OBJECTIVE: To compare the clinical utility of FGMS against blood glucose curves (BGCs) obtained with a portable blood glucose meter (PBGM) in monitoring DD. ANIMALS: Twenty dogs with diabetes mellitus. METHODS: Prospective study. Dogs with diabetes mellitus on insulin treatment for at least 1 month were included. Comparisons of insulin dose recommendations based on the in-hospital GCs acquired using FGMS and a PBGM, consecutive-day interstitial GCs (IGCs) acquired at home using the FGMS, and consecutive-day, home vs hospital IGCs acquired using the FGMS were made using concordance analysis. RESULTS: There was good concordance between insulin dose recommendations based on FGMS and PBGM generated GCs and IGCs obtained in the 2 different environments on 2 consecutive days, but almost absent concordance between IGCs obtained on 2 consecutive days at home. Glucose nadirs were detected in 34/43 (79%) of Ambulatory Glucose Profile (AGP) reports of the FGMS. In comparison, concordant glucose nadirs were identified in 14/34 (41%) BGCs using PBGM. The individual FGMS scans and PBGM identified 60% and 9% of low IG/hypoglycemic episodes, respectively. CONCLUSIONS AND CLINICAL IMPORTANCE: Insulin dose adjustments based on BGCs can be suboptimal. The FGMS allows a more accurate identification of the glucose nadirs and hypoglycemic episodes compared to the use of a PBGM and assessment of day-to-day variations in glycemic control.

Accuracy of a flash glucose monitoring system in dogs with diabetic ketoacidosis.

BACKGROUND: A factory-calibrated flash glucose monitoring system (FGMS; FreeStyle Libre) recently was evaluated in dogs with uncomplicated diabetes mellitus. It is not known if this system is reliable during diabetic ketoacidosis (DKA). OBJECTIVES: To assess the performance of the FGMS in dogs with DKA and to determine the effect of severity of ketosis and acidosis, lactate concentration, body condition score (BCS), and time wearing the sensor on the accuracy of the device. ANIMALS: Fourteen client-owned dogs with DKA. METHODS: The interstitial glucose (IG) measurements were compared with blood glucose (BG) measurements obtained using a validated portable glucometer. The influence of changes in metabolic variables (ß-hydroxybutyrate, pH, bicarbonate, and lactate) and the effect of BCS and time wearing on sensor performance were evaluated. Accuracy was determined by fulfillment of ISO15197:2013 criteria. RESULTS: Metabolic variables, BCS, and time wearing were not associated with the accuracy of the sensor. Good agreement between IG measurements and BG was obtained both before and after DKA resolution (r = .88 and r = .93, respectively). Analytical accuracy was not achieved, whereas clinical accuracy was demonstrated with 100% and 99.6% of results in zones A + B of the Parkes consensus error grid analysis before and after DKA resolution, respectively. CONCLUSIONS AND CLINICAL IMPORTANCE: Changes in metabolic variables, BCS, and time wearing do not seem to affect agreement between IG and BG. Despite not fulfilling the ISO requirements, the FGMS provides clinically accurate estimates of BG in dogs with DKA.