Accurate classification of Cushing's Syndrome (CS) plays a critical role in providing the early and correct diagnosis of CS that may facilitate treatment and improve patient outcomes. Diagnosis of CS is a complex process, which requires careful and concurrent interpretation of signs and symptoms, multiple biochemical test results, and findings of medical imaging by physicians with a high degree of specialty and knowledge to make correct judgments. In this article, we explore the state of the art machine learning algorithms to demonstrate their potential as a clinical decision support system to analyze and classify CS to facilitate the diagnosis, prognosis, and treatment of CS. Prominent algorithms are compared using nested cross-validation and various class comparison strategies including multiclass, one vs. all, and one vs. one binary classification. Our findings show that Random Forest (RF) algorithm is most suitable for the classification of CS. We demonstrate that the proposed approach can classify CS with an average accuracy of 92% and an average F1 score of 91.5%, depending on the class comparison strategy and selected features. RF-based one vs. all binary classification model achieves sensitivity of 97.6%, precision of 91.1%, and specificity of 87.1% to discriminate CS from non-CS on the test dataset. RF-based multiclass classification model achieves average per class sensitivity of 91.8%, average per class specificity of 97.1%, and average per class precision of 92.1% to classify different subtypes of CS on the test dataset. Clinical performance evaluation suggests that the developed models can help improve physicians' judgment in diagnosing CS.
Accurate classification of Cushing's Syndrome (CS) plays a critical role in providing early and correct diagnosis of CS that may facilitate treatment and improve patient outcomes. Diagnosis of CS is a complex process, which requires careful and concurrent interpretation of signs and symptoms, multiple biochemical test results, and findings of medical imaging by physicians with a high degree of specialty and knowledge to make correct judgments. In this article, we explore the state of the art machine learning algorithms to demonstrate their potential as a clinical decision support system to analyze and classify CS in order to facilitate the diagnosis, prognosis, and treatment of CS. Prominent algorithms are compared using nested cross-validation and various class comparison strategies including multiclass, one vs. all, and one vs. one binary classification. Our findings show that Random Forest (RF) algorithm is most suitable for the classification of CS. We demonstrate that the proposed approach can classify CS subjects with an average accuracy of 92% and an average F1 score of 91.5%, depending on the class comparison strategy and selected features. RF-based one vs. all binary classification model achieves sensitivity of 97.6%, precision of 91.1%, and specificity of 87.1% to discriminate CS from non-CS on the test dataset. RF-based multiclass classification model achieves average per class sensitivity of 91.8%, average per class specificity of 97.1%, and average per class precision of 92.1% to classify different subtypes of CS on the test dataset. Clinical performance evaluation suggests that the developed models can help improve physician's judgment in diagnosing CS.
Introduction Obesity is a known condition when the energy taken is more than the energy consumed. According to the World Obesity Atlas data, 1 out of every five women and 1 out of every seven men will live with obesity by 2030.The prevalence of obesity was 21.1% in Turkey (2019). Although environmental factors and changes in dietary patterns are essential in this increase, genetic factors also have an effect. Mutations in genes that affect catecholamine function, which play an essential role in energy consumption and lipolysis, also lead to obesity. One of these critical genes is the β2-adrenergic receptor gene (ADRB2). It was previously reported that it is responsible for regulating lipolysis and thermogenesis and is mainly expressed in the adipose tissues. This gene induces catecholamine activation and leads to lipid mobilization in adipose tissue. Clinical Case A 25-year-old male patient applied to our outpatient clinic with obesity, sexual dysfunction. The patient had been overweight since childhood and did not use any medication at admission. On physical examination, his body mass index:50.8 kg/cm2.and has only hair on the genital area and axilla. His parents are consanguineous, and his mother, father, and brother are also obese. He had diabetes mellitus, hypolipidemia, hypogonadotropic hypogonadism, and GH deficiency in his initial tests (Table-1). The other pituitary hormones were usual. There was not any pathology in the pituitary and brain magnetic resonance images. LHRH test result was interpreted as hypothalamic dysfunction. The insulin hypoglycemia test was performed due to growth hormone deficiency to accurate to GH deficiency. Clinic exome sequencing was performed on the patient, and a de novo heterozygous mutation in the ADRB2 gene was detected, which has a role in lipolysis, and thermogenesis and hence is important for obesity. Two more mutations were identified that related to thyroid functions. The DUOX2 gene variant has an association with thyroid dyshormonogenesis, and the IRS4 gene variant has a role in hypothyroidism. Further studies are needed to identify the pathogenicity of variants. Molecular identification of mutations has important implications for personalized medicine, including genetic counseling and the development of specific treatment protocols. We started the patient with exenatide 10 mg/day, metformin 2000 mg/day, testosterone propionate, and a diet program. Conclusion Studies show that low ADRB2 expression increases the risk of obesity. Obesity is increasing rapidly over time globally, and it is necessary to know the underlying pathophysiologies and genetic etiologies and to develop new treatment algorithms for this in order to reduce mortality and morbidity. Therefore, gene mutations underlying obesity, the metabolic changes it causes, environmental factors, and diet should be considered. Considering the scarcity of available anti-obesity drugs, understanding the genetic makeup may be influential in developing new therapeutic targets.
Introduction Rathke cleft cysts (RCCs) are commonly believed to be cysts formed from remnants of Rathke's pouch. It is usually asymptomatic, < 3mm in diameter, and found in 13–33% of normal pituitary glands. The coexistence of pituitary adenoma (PA) and RCC is rare (incidence rate of 0.51%-1.7%), and growth hormone hypersecretion has also been observed in a limited number of cases in these patients. Therefore, we aim to present a rare case of growth hormone- (GH-) producing PA with concomitant RCC. Clinical Case A 29-year-old female patient, due to primary amenorrhea, was diagnosed as having 46XY chromosomal Swyer syndrome from her previous visit, and she had bilateral salphengoopherectomy afterward. In clinical follow-ups, due to the presence of growth complaints in the hands and feet, insulin-like growth factor-1 (IGF-1) was found to be high. The basal serum GH value was 8.89 ng/mL, and IGF-1 was 369 ng/mL. The 75 gr oral glucose tolerance test (OGTT) result showed that the serum GH level was not suppressed below 1.0 μg/L, and the test was repeated twice. Other anterior pituitary hormone axes are normal. In the pituitary magnetic resonance imaging revealed a low-signal 4 mm Rathke cleft cyst is observed in the T1A image in the midsection of the pituitary gland and a high-signal intra-cystic nodule is observed within the cyst on the T1A image (Figure 1A). The fluid-fluid level is observed in the cyst on the T2A image (Figure 1B). Conclusion The relationship between RCC and PA is unclear and theoretical. The theory is “transitional pituitary tumor,” defined as the Rathke pouch forming the anterior lobe of the pituitary gland with a proliferation of the anterior wall. Moreover, PAs originate from the anterior lobe of the pituitary gland via clonal changes. This shows that RCCs originated from the remnant of the Rathke pouch, and PA has a common embryonic origin. Nevertheless, they rarely occur at the same time. Only 14 patients with GH-producing PA and concomitant RCC have been reported in the literature. Preoperative correct diagnosis of PA and RCC coexistence is difficult, because RCCs’ imaging characteristics (size, position, and intensity) are highly variable, and cystic degeneration occurs in 5% to 18% of PA cases. RCCs are often surrounded by the normal pituitary gland and are located in the nearby midline of the sellar region. MRI characteristics of sellar lesions for PAs existing concomitant with RCCs may present with variable intensity on T1-weighted imaging. If radiographic imaging shows that typical RCC and PA are not obvious at first glance, the possibility of concomitant PA still needs to be considered. If the PA in this patient had been nonfunctional, correct diagnosis of the concomitant PA would have been much more difficult. This case offers a reminder that although rare, RCC can accompany PA, and performing contrast-enhanced MRI is useful to detect RCC.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.