Assessment of automatic decision-support systems for detecting active T2 lesions in multiple sclerosis patients.

BACKGROUND: Active (new/enlarging) T2 lesion counts are routinely used in the clinical management of multiple sclerosis. Thus, automated tools able to accurately identify active T2 lesions would be of high interest to neuroradiologists for assisting in their clinical activity. OBJECTIVE: To compare the accuracy in detecting active T2 lesions and of radiologically active patients based on different visual and automated methods. METHODS: One hundred multiple sclerosis patients underwent two magnetic resonance imaging examinations within 12 months. Four approaches were assessed for detecting active T2 lesions: (1) conventional neuroradiological reports; (2) prospective visual analyses performed by an expert; (3) automated unsupervised tool; and (4) supervised convolutional neural network. As a gold standard, a reference outcome was created by the consensus of two observers. RESULTS: The automated methods detected a higher number of active T2 lesions, and a higher number of active patients, but a higher number of false-positive active patients than visual methods. The convolutional neural network model was more sensitive in detecting active T2 lesions and active patients than the other automated method. CONCLUSION: Automated convolutional neural network models show potential as an aid to neuroradiological assessment in clinical practice, although visual supervision of the outcomes is still required.

Risk Factors, Radiological and Clinical Outcomes in Subclinical and Clinical Pituitary Apoplexy.

BACKGROUND: Pituitary apoplexy (PA) can be symptomatic, namely acute apoplexy (APA), or asymptomatic or subclinical (SPA). OBJECTIVE: To describe the clinical characteristics and evolution of the patients with APA compared to SPA Patients and methods: Retrospective, longitudinal database analysis. RESULTS: We identified 58 patients with PA, and 37 accomplished the inclusion criteria (17 men, median age 47.7 years). A total of 29 (78.4%) had APA (17 underwent surgery, and 12 were conservatively managed), and 8 (21.6%) had SPA. The presence of non-functioning pituitary adenoma (NFPA) odds ratio (OR): 29.36 (95% confidence interval (CI): 1.86-462.36) and the largest size OR 1.10 (95% CI: 1.01-1.2) elevated the risk of having surgery. Hypopituitarism developed in 35.1% without significant differences between APA and SPA. In non-surgical patients, adenoma volume shrunk spontaneously at one year magnetic resonance imaging (MRI), without statistical differences between the conservatively treated and SPA group. CONCLUSIONS: APA is more frequent in larger NFPAs, and this subset of patients has a higher risk of surgery. Hypopituitarism is quite frequent even in patients with SPA, and, therefore, long-term follow-up is mandatory. In the non-surgical group, the pituitary tumour shrinkage is clinically relevant after one year of PA. Consequently, surgery indication in NFPA should be delayed and reassessed if patients remain asymptomatic.

Shrinkage by the third month predicts long-term response of macroprolactinoma after cabergoline.

OBJECTIVE: Transsphenoidal surgery (TSS) is mainly indicated in prolactinomas when dopamine agonist treatment fails. However, there is no established early predictor of cabergoline (CBG) response. The present study was aimed to identify predictors of CBG resistance in order to select patients who may benefit from early TSS. DESIGN: Retrospective longitudinal study. METHODS: We reviewed the medical record of patients diagnosed with prolactinoma after 2010. Inclusion criteria: macroprolactinomas under CBG treatment with serial prolactin levels and MRI before treatment and 3 and 12 months afterwards. The main outcome was tumour size shrinkage ≥ 50% (using the two largest diameters in sagittal view) after 12 months of CBG (TS_50). The capacity of the most important clinical and biochemical variables in predicting the main outcome was examined. RESULTS: A total of 185 prolactinomas where included: 124 (67.0%) were microadenomas and 61 (33.0%) were macroadenomas of which 27 patients meet de inclusion criteria; median age (42.5 years; (IQR: 28.0)). The median follow-up was (67.5 months; (IQR: 30.2)). Ten patients (37.0%) underwent surgery after more than 1 year of CBG. The volume reduction at the first MRI (3-4 months) was the unique valuable predictor: (OR: 1.16 (95% CI: 1.02-1.32)) of TS_50. A tumour volume shrinkage of ≥ 30% in the first 3-4 months of CBG therapy predicts TS_50 with an AUC (0.95 (CI: 0.76-0.99)). CONCLUSION: Tumour shrinkage in the first 3-4 months after starting treatment with CBG is a good tool for predicting the long-term response and can help clinicians to take more appropriated and personalized decisions.