Mathematical model for preoperative differential diagnosis for the parathyroid neoplasms

Krupinova, J.; Elfimova, A.; Kim, E.; Eremkina, A.; Maganeva, I.; Kovaleva, E.; Mokrysheva, N.

Авторы: Krupinova Julia, Elfimova Alina, Kim Ekaterina, Eremkina Anna, Maganeva Irina, Kovaleva Elena, Mokrysheva Natalia
Сборник: ECE 2023 Abstract Book
Тезисы
Год издания: 2023
Место издания: Bioscientifica
DOI: 10.1530/endoabs.90.ep579
Аннотация: BackgroundsPreoperative diagnosis of parathyroid carcinoma (PC) is critical for thedetermination of the scope of surgical intervention. Nowadays, specific diagnosticmarkers for differentiation of PC and benign tumors are unknown, and less thanhalf of patients with PC undergo necessary en bloc surgery.AimsTo develop the instrument for preoperative diagnosis of PC.Materials and methodsA multi-center retrospective study included 242 patients with primaryhyperparathyroidism: 50 patients with PC, 30 with ?typical adenoma (AA), and162 with adenoma of the parathyroid glands. We compared laboratory,instrumental and clinical characteristics. Comparison of two independent groupsfor quantitative data was performed using the Mann–Whitney test (U-test).Comparison of three independent groups for quantitative data was conductedusing the Kruskal–Wallis ANOVA. The frequencies of binary variables werecompared using the two-tailed Fisher exact test and Freeman–Halton test. Thecritical level of statistical significance for statistical hypotheses testing was takenas 0.05. CatBoosting algorithm was conducted to construct mathematical modelsfor the differential diagnosis of PTG neoplasm types. Twelve factors were chosenby expert method for the model construction. At the first step, we reduced thedimension of feature space using sklearn.feature_selection, module ExtraTreesClassifier and SelectFromModel. Then we constructed the first model to differPA and (PC or AA). At the second step, we also reduced feature space andconstructed the second model to differentiate PC and AA.ResultsPatients with PC and AA had higher levels of PTH, ionized and albumincorrected calcium, alkaline phosphatase, volume and the largest diameter ofneoplasm, and the higher frequency of glomerular filtration rate decrease less than60 ml/min per 1.73 m2 compared to patients with adenoma. The frequency oflow-energy fractures was higher in the carcinoma group versus the adenomagroup (32% vs 8%). Heterogeneous structure and indefinite contour of glandsdetected by ultrasound were more typical for PC than for AA and adenomas. Themathematical model was developed using CatBoost gradient boosting algorithmfor the noninvasive preoperative differential diagnosis of PC, AA, and adenoma.ConclusionsModel can predict adenoma with PPV 100% and PC with PPV 81–92%. Usingmodel clinicians could plan extended en bloc resection for PC and selectiveparathyroidectomy for adenoma. If AA is predicted, he has to make a decision onthe choice of the necessary volume of parathyroidectomy based on his experience,because AA are the zone of uncertainty
Добавил в систему: Kim Ekaterina Igorevna

	ИСТИНА	Войти в систему Регистрация
	ИСТИНА ЦЭМИ РАН
	Главная Поиск Статистика О проекте Помощь

ИСТИНА

ИСТИНА ЦЭМИ РАН

Mathematical model for preoperative differential diagnosis for the parathyroid neoplasmsтезисы доклада

Прикрепленные файлы