Use of an artificial neural network to predict Graves’ disease outcome within 2 years of drug withdrawal

Orunesu, Eva; Bagnasco, Marcello; Salmaso, C.; Altrinetti, Vania; Bernasconi, D.; Monte, P. Del; Marugo, M; Gs, Mela

doi:10.1111/j.1365-2362.2004.01318.x

Cited by 24 publications

(17 citation statements)

References 43 publications

(65 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As far as AITD are concerned, compelling evidence has been provided for a significant role of cigarette smoking as an independent risk factor for GD, and notably Graves' ophthalmopathy. Cigarette smoking also negatively affects the long-term effectiveness of antithyroid drug therapy [6,7] .…”

Section: Aitd Environment and Stressmentioning

confidence: 99%

Stress and Autoimmune Thyroid Diseases

Bagnasco¹,

Bossert²,

Pesce³

2006

Neuroimmunomodulation

View full text Add to dashboard Cite

Autoimmune thyroid diseases (AITD) are the far most common autoimmune disorders, their prevalence in Western countries exceeding 5% of the general population. In the large majority of individual cases the clinical impact of AITD is not severe, however, their widespread diffusion renders them a significant health problem. AITD are heterogeneous in their clinical presentation: the two main forms are autoimmune thyroiditis (AT) and Graves’ disease (GD). Although they probably share, at least in part, a common genetic background and may occur in the same family as well as in the same individual, they are definitely two distinct diseases both in their clinical presentation and their pathophysiology. In fact, AT causes structural thyroid damage (mainly via cell-mediated immune destruction of thyroid follicular cells) which results, as a rule, in functional impairment (hypothyroidism); however, depending on clinical variants, evolution towards hypothyroidism may be very low, or thyroid function impairment occurs after an initial phase of mild thyrotoxicosis due to relatively rapid gland destruction. GD patients have hyperthyroidism, often severe, due to autoantibody-mediated thyrotropin receptor stimulation, with thyroid cell hyperplasia and hyperfunction. Such a functional heterogeneity is a key feature for the clinical management: as a matter of fact, therapy of AITD is mainly therapy of thyroid dysfunction. Moreover, since hyperthyroidism is quite early perceived by the patient as a cause of discomfort, the timing of the natural history of GD is relatively well defined; on the other hand, AT may be asymptomatic for a long time and defining its natural history in a single patient may be difficult. In some AITD patients (mainly, but not exclusively, with GD), clinical features not directly related to thyroid dysfunction, such as orbitopathy, are present. Graves’ orbitopathy is probably related, at least in part, to autoantibodies directed to thyrotropin receptor; it may be, in a minority of patients, severe and sight-threatening, and represents an independent clinical problem.

show abstract

Section: Aitd Environment and Stressmentioning

confidence: 99%

Stress and Autoimmune Thyroid Diseases

Bagnasco¹,

Bossert²,

Pesce³

2006

Neuroimmunomodulation

View full text Add to dashboard Cite

show abstract

“…The major advantage is that with sufficient data, ANN can be trained to learn the relationship between the inputs and outputs, even the mechanism of the relationship is unknown or unclear [14]. This ensures the flexibility of ANN in computeraided diagnosis [2,10,13,16,17]. Accordingly, in present study an ANN model based on back propagation (BP) net is utilized.…”

Section: The Structure and Algorithm Of The Modelmentioning

confidence: 99%

Artificial neural network based model for cardiovascular risk stratification in hypertension

Ning

Ying-qi

et al. 2006

Med Bio Eng Comput

View full text Add to dashboard Cite

This study was to develop an objective method to stratify cardiovascular risk in hypertension. Stratification for cardiovascular risk is crucial in deciding treatment strategy for hypertension but has yielded undesirable results in clinic due to its low accuracy which is caused by physicians' subjective experience and the uncertainty of patients' statements. Our model proposed herein overcomes these disadvantages by applying artificial neural network based on a classic back propagation net. The model input is derived from the clinical investigation. The target output is the stratification level of total cardiovascular risk, which is learned from the guidelines of hypertension treatment. Study in 348 normotensive and hypertensive subjects showed that the results of model stratification are consistent with the standard stratification suggested by hypertension guidelines in 81.61% cases. The results confirm the accuracy of the model and demonstrate its ability in risk evaluation for hypertension.

show abstract

“…Искусственные нейронные сети, позволяющие решать проблемы распознавания состояний [25], способны обобщать данные только из примеров определенного диапазона, тогда как вне этого преде-ла результаты могут не соответствовать действитель-ности [24,26]. Так, искусственная нейронная сеть типа Perceptron использовалась для прогнозирования клинического течения болезни Грейвса в отношении ремиссии или рецидива после отмены метимазо-ла [27]. Были рассмотрены 27 переменных, получен-ных при постановке диагноза или во время лечения.…”

Section: персонализированная педиатрияunclassified

Personalized Medicine: Genome, Electronic Health and Intelligent Systems. Part 2. Molecular Genetics and Methods of Intellectual Analysis

Кобринский¹

2017

Ross. vestn. perinatol. pediatr.

View full text Add to dashboard Cite

Use of an artificial neural network to predict Graves’ disease outcome within 2 years of drug withdrawal

Abstract: This study reveals that perceptron-like ANN is potentially a useful approach for GD-management in choosing the most appropriate therapy schedule at the time of diagnosis.

Cited by 24 publications

References 43 publications

Stress and Autoimmune Thyroid Diseases

Stress and Autoimmune Thyroid Diseases

Artificial neural network based model for cardiovascular risk stratification in hypertension

Personalized Medicine: Genome, Electronic Health and Intelligent Systems. Part 2. Molecular Genetics and Methods of Intellectual Analysis

Contact Info

Product

Resources

About