Clinical and radiological parallels in primary lung cancer diagnosis using mathematical modeling technologies
Year: 2024
Type: article
Abstract: BACKGROUND: Lung cancer is becoming increasingly relevant healthcare issue every year. According to the Altai Regional Oncology Center cancer registry, the incidence of lung cancer in 2019 was 114.8/100 000 in men, 19.3/100 000 in women; in 2020 — 96.8/100 000 and 16.8/100 000, respectively. In 2021, the incidence for both genders was 108.9/100 000. Diagnostic rates in 2022 among patients with respiratory cancer were discouraging: at the time of diagnosis 42.2% had stage IV, 27.9% stage III, 16.3% stage I and 12.9% stage II, and in 0.7% of cases the stage was not established. Distribution of lung cancer patients of different age groups depending on the tumor histotype, showed that the majority are adenocarcinoma and squamous cell lung cancer — 85%. AIM: To assess the possibility to identify morphological forms of lung cancer (adenocarcinoma, squamous cell and small cell cancer) by using artificial intelligence, based on the results of multispiral computed tomography and additional parameters. MATERIALS AND METHODS: We used multispiral computed tomography data of patients with lung cancer, analyzed with the “Radiologist+” program (Russia, Barnaul), which allows direct sampling of average pixel densities in tabular form in selected areas of interest from DICOM files for subsequent analysis and statistical processing. The resulting densitometric indicators were incorporated in the artificial neural network. RESULTS: Data from 485 patients with lung cancer aged 45 to 80 years was analyzed, taking into account nine parameters. CONCLUSION: Mathematical model for differential diagnosis of histological lung cancer forms, taking into account the presence or absence of tobacco smoking, showed 85.8% sensitivity, 85.0% specificity and 85.4% accuracy.
Source: Russian Journal of Oncology
DOI: 10.17816/onco630357
Authors: Olga V. Borisenko, Alexander F. Lazarev, Konstantin S. Titov
Language: English
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: diamond
OpenAlex ID: W4402496515