Subject(s)
Incidental Findings , Lung Neoplasms , Tomography, X-Ray Computed , Humans , Lung Neoplasms/diagnostic imaging , Scimitar Syndrome/diagnostic imaging , Early Detection of Cancer , Male , Female , Middle Aged , Aged , Radiation Dosage , Pulmonary Veins/abnormalities , Pulmonary Veins/diagnostic imagingSubject(s)
Humans , Male , Aged , Fistula , Angiography , Pulmonary Artery/diagnostic imaging , UltrasonographySubject(s)
Fistula , Pulmonary Artery , Humans , Pulmonary Artery/diagnostic imaging , Angiography , Ultrasonography , Ultrasonography, DopplerABSTRACT
Machine learning models are increasingly used in the medical domain to study the association between risk factors and diseases to support practitioners in understanding health outcomes. In this paper, we showcase the use of machine-learned staged tree models for investigating complex asymmetric dependence structures in health data. Staged trees are a specific class of generative, probabilistic graphical models that formally model asymmetric conditional independence and non-regular sample spaces. An investigation of the risk factors in invasive fungal infections demonstrates the insights staged trees provide to support medical decision-making.
Subject(s)
Humans , Male , Aged , Lung Neoplasms , Neoplasms, Multiple Primary , Lung/diagnostic imaging , Tomography, X-Ray ComputedABSTRACT
OBJECTIVE: To compare body composition between patients with autonomous cortisol secretion (ACS), those with nonfunctioning adrenal incidentalomas (NFAIs), and control subjects without adrenal tumors. METHODS: A cross-sectional study was performed, incluidng the following 3 groups: patients with ACS (cortisol post-dexamethasone suppression test [DST] >1.8 µg/dL), NFAIs (cortisol post-DST ≤ 1.8 µg/dL), and patients without adrenal tumors (control group). Patients of the 3 groups were matched according to age (±5 years), sex, and body mass index (±5 kg/m2). Body composition was evaluated by bioelectrical impedance and abdominal computed tomography (CT) and urinary steroid profile by gas chromatography mass spectrometry. RESULTS: This study enrolled 25 patients with ACS, 24 with NFAIs, and 24 control subjects. Based on CT images, a weak positive correlation between the serum cortisol level post-DST and subcutaneous fat area (r = 0.3, P =.048) was found. As assessed by bioelectrical impedance, lean mass and bone mass were positively correlated with the excretion of total androgens (r = 0.56, P <.001; and r = 0.58, P <.001, respectively); visceral mass was positively correlated with the excretion of glucocorticoid metabolites and total glucocorticoids (r = 0.28, P =.031; and r = 0.42, P =.001, respectively). Based on CT imaging evaluation, a positive correlation was observed between lean mass and androgen metabolites (r = 0.30, P =.036) and between visceral fat area, total fat area, and visceral/total fat area ratio and the excretion of glucocorticoid metabolites (r = 0.34, P =.014; r = 0.29, P =.042; and r = 0.31, P =.170, respectively). CONCLUSION: The urinary steroid profile observed in adrenal tumors, comprising a low excretion of androgen metabolites and high excretion of glucocorticoid metabolites, is associated with a lower lean mass and bone mass and higher level of visceral mass in patients with adrenal tumors.