Endocrine Abstracts

Introduction: Finding the optimal levothyroxine (LT4) dose regime to ameliorate the abnormally low levels of natural thyroid hormones, especially for thyroidectomized patients, is still challenging. Many researchers have studied various LT4 dosage regimen clinically; and ultimately, they proposed multiple variables affecting LT4 requirements including age, gender, body weight, body mass index (BMI), and body surface area (BSA). However, prescribing the most appropriate LT4 dose regime for different patients remains ambiguous.

Method: We attempt to adjust an appropriate LT4 dose regime for a total thyroidectomized virtual-patient by means of fuzzy logic system (FLS) methodology, an applicable artificial intelligence technique. THYROSIM 3.0^© a free web application developed by UCLA Biocybernetics Laboratory, was utilized as a model of feedback control of hypothalamus-pituitary-thyroid axis. In order to evaluate patient responses to LT4 monotherapy, we simulated our total thyroidectomized virtual-patient by setting T3 and T4 secretion parameters at 1% while receiving dynamic oral LT4 dosages post-surgery. In addition, with an assumption that no supplement was administrated, the absorption rate of oral LT4 was set to 88%. Fuzzy logic controller was developed using MATLAB software ver. 2019. The discrepancies of TSH value at day n and one-step time back TSH value (at day n-1) in regard with the TSH set point were considered as the controlled variables while LT4 daily dosage was considered as the manipulated variable.

Results: According to our proposed algorithm, our developed FLS recommends a LT4 monotherapy dose regime for the assumed total thyroidectomized virtual-patient on a daily basis as presented in the following table. The resulting doses provided by FLS are indicated as “Precise FLS LT4 dose” while available doses are presented by rounding the precise dose considering 25 μg intervals in respect with the smallest increment between LT4 dosing strengths.

Conclusion: The FLS method could precisely predict the appropriate LT4 daily doses for our total thyroidectomized virtual-patient. This proposed method would dramatically declines the numbers of days in which our virtual-patient experiences thyroid hormone levels out of normal ranges while eliminating the rigorous fluctuations of the plasma levels of thyroid hormones.