Artificial neural networks in diagnosis of thyroid function from in vitro laboratory tests

Abstract

We studied the potential benefit of using artificial neural networks (ANNs) for the diagnosis of thyroid function. We examined two types of ANN architecture and assessed their robustness in the face of diagnostic noise. The thyroid function data we used had previously been studied by multivariate statistical methods and a variety of pattern-recognition techniques. The total data set comprised 392 cases that had been classified according to both thyroid function and 19 clinical categories. All cases had a complete set of results of six laboratory tests (total thyroxine, free thyroxine, triiodothyronine, triiodothyronine uptake test, thyrotropin, and thyroxine-binding globulin). This data set was divided into subsets used for training the networks and for testing their performance; the test subsets contained various proportions of cases with diagnostic noise to mimic real-life diagnostic situations. The networks studied were a multilayer perceptron trained by back-propagation, and a learning vector quantization network. The training data subsets were selected according to two strategies: either training data based on cases with extreme values for the laboratory tests with randomly selected nonextreme cases added, or training cases from very pure functional groups. Both network architectures were efficient irrespective of the type of training data. The correct allocation of cases in test data subsets was 96.4-99.7% when extreme values were used for training and 92.7-98.8% when only pure cases were used.

Clinical validation of an artificial neural network trained to identify acute allograft rejection in liver transplant recipients

“…ANNs have been studied in a wide variety of medical applications, including cancer diagnosis, 11 myocardial infarction diagnosis, [12][13][14] cross-match prediction in kidney transplant patients, 15 thyroid function diagnosis, 16 and prediction of blood glucose levels from the concentration of counterregulatory hormones. 17 Neural networks have been designed to diagnose CMV disease 18 and acute rejection using data obtained from posttransplantation renal biopsies 19 in the kidney transplant population.…”

“…Artificial neural networks (ANNs), a data-modeling method, have been studied in a wide variety of medical applications. [11][12][13][14][15][16][17] In the field of transplantation, ANNs have been designed to diagnose cytomegalovirus (CMV) disease 18 and acute rejection using data obtained from posttransplantation renal biopsies 19 after kidney transplantation. In addition, the use of ANNs was investigated in the prediction of graft failure, 20 prediction of liver transplantation outcome, 21 selection of patients for liver transplantation, 22 and prediction of tacrolimus blood levels.…”

Applicability of artificial neural networks in modelling endocrine disorders and growth processes

“…Chociaż sieci neuronowe uważane są obecnie za nowoczesne narzędzie statystyczne, pierwsze publikacje dotyczące ich zastosowania w medycynie, w tym również w endokrynologii, pochodzą sprzed ponad 20 lat. W 1993 r. ukazała się praca autorów brytyjskich [3], w której podjęto próbę wykorzystania ANN w diagnostyce czynności tarczycy. Początkowo w procesie uczenia sieci wprowadzano wyniki sześciu testów diagnostycznych: stężenia tyroksyny całkowitej (T4) i wolnej (FT4), stężenie trijodotyroniny (T3), test wychwytu trijodotyroniny (T3U), stężenie tyreotropiny (TSH) metodą radioimmunologiczną (tj.…”

Prototipo de sistema experto en diagnóstico médico basado en síntomas de los pacientes. Caso de estudio: esclerosis múltiple

“…Otros estudios evaluaron la aplicación de sistemas ND para el diagnóstico de infarto del miocardio (Buarque, 2000), (Bizios, Heij, & Bengtsson, 2007), artritis celular (Astion, Wener, Thomas, Hunder, & Bloch, 1994), (Borges, Moura, & Steiner, 2010), tipos de cáncer (Bourdes, Bonnevay, Bachelot, & Perol, 2007), pronosticar la duración de pacientes en la unidad de cuidados intensivos (Cook, 2005), evolución de enfermedades después de realizarse una operación de trasplante (Doyle, y otros, 1994), (Gil, Johnsson, Garcia, & Soriano, 2009). Otros estudios que usan RNA están relacionados con enfermedades cardio-pulmonares y grado de sobreviviencia después del trauma (Ebell, 1993), (Harrison & Kennedy, 2005), interpretación de pruebas de diagnóstico relacionadas con el páncreas (Kasmierczak, Catrou, & Van Lente, 1998), (Bartosh, Andersson, & Nilsson, 2008) y tiroideos (Sharpe, Solberg, Rootwelt, & Yearworth, 1993), (Gannous & Elhaddad, 2011).…”