基于人工神经网络影像智能分析的前列腺癌精准诊断模型建立的关键技术研究

前列腺癌是危害老年男性健康最重要的问题之一，近年来我国前列腺癌的发病率呈快速上升趋势，因此，早期诊断前列腺癌具有重要临床价值。在我们前期研究基于人工神经网络超声智能分析技术（人工智能超声CT）提高前列腺癌检出率的基础上，本项目将继续探索基于人工神经网络影像智能分析的前列腺癌精准诊断模型建立的关键技术，重点研究基于人工神经网络的MRI智能分析诊断模型的建立及人工智能超声CT的优化。本项目通过高通量测序及生物信息学分析筛选与前列腺癌发生密切相关的“明星”分子标记物，并引入前列腺硬度信息，优化人工智能超声CT在前列腺癌诊断中的价值，进一步在其基础上建立基于人工神经网络的MRI智能分析诊断模型，完善前列腺癌诊断的影像智能分析体系。本项目将为基于影像智能分析的前列腺癌早期诊断提供新的策略。

Prostate cancer becomes one of the most critical health issues in male population worldwide. The incidence of prostate cancer increased rapidly in China in recent years. Therefore, early detection of prostate cancer is of great importance. Our preliminary results indicate that Artificial Neural Network Analysis / Computerized Transrectal Ultrasound (ANNAcTRUS, Ultrasound CT) increase the detection rate of prostate cancer. Based on these data, we will continue to explore the crucial techniques of intelligent imaging analysis based on artificial neural network in diagnosis of prostate cancer and we will focus on establishing a diagnostic model of MRI with artificial intelligence based on ANN and optimization of ultrasound CT. We use high-throughput sequencing and bioinformatics analysis to screen “star” biomarkers related to prostate cancer to optimize ultrasound CT in prostate cancer diagnosis together with prostate hardness. We further establish a diagnostic model of MRI with artificial intelligence based on ANN to improve intelligent imaging analysis in prostate cancer detection. Successful completion of these experiments may provide novel strategies for early detection of prostate cancer based on intelligent imaging analysis.