An algorithm for cognitive fusion targeted tumor puncture based on 3-D mathematical modelling.

ABSTRACT

Background: Percutaneous puncture is an important means of tumor diagnosis and treatment. At present, most puncture operations are still based on imaging location and clinical experience, and quantitative and accurate targeted puncture cannot be achieved. How to improve the accuracy of percutaneous tumor puncture, avoid errors to the greatest extent, reduce the occurrence of complications, and improve the overall clinical diagnosis and treatment quality and curative effect, are scientific problems worthy of further study. Method: In the present study, mathematical modeling was first used to construct the tumor puncture path, determine the needle entry angle, and define the relevant limited parameters and the substitution formula. Secondly, relevant parameters were extracted from CT and other imaging data and substituted into formulas, the deviation angle and puncture path were determined, and the personalized tumor puncture scheme was carried out. Third, targeted puncture was precisely implemented under the guidance of B-ultrasound. Compared with the traditional empirical puncture, our model improved the accuracy, decreased the puncture time, and reduced the pain of diagnosis and treatment for patients. Results: A tumor-targeted puncture model was established based on mathematical theory and imaging data. By extracting clinical data, such as tumor radius, projection distance of tumor center and projection distance from puncture point to body surface, the optimal puncture deviation angle was modeled and calculated and a personalized puncture scheme was established. Compared with the conventional method, our model markedly increased the puncture accuracy rate by ∼30%. The puncture number was decreased by ∼50% using our model. Furthermore, our model shortened the operation time by 20% to ease pain of patients and guarantee greater security for patients. Doctor satisfaction and patient discomfort scores were examined. Our model improved doctor satisfaction by ∼20% and reduced subjective discomfort of patients by ∼25%. These data revealed that the model could markedly improve the accuracy and efficiency of puncture, clinical efficacy and accuracy of tumor diagnosis. Additionally, the confidence of doctors in the operation was greatly enhanced and patient discomfort was greatly reduced. Conclusion: The present study analyzed in detail how to find the best puncture path using a mathematical model. Based on the mathematical model of cognitive fusion puncture, combined with clinical personalized data and mathematical calculation analysis, accurate puncture was effectively realized. It not only greatly improved the effectiveness of puncture, but also ensured the safety of clinical patients and reduced injury, which means it may be worthy of clinical application.