A5046341694- Radiomics and Machine Learning in Medical Imaging
- Prostate Cancer Diagnosis and Treatment
- Advanced Neural Network Applications
- Prostate Cancer Treatment and Research
- Medical Imaging and Analysis
- Domain Adaptation and Few-Shot Learning
- MRI in cancer diagnosis
- AI in cancer detection
- Urological Disorders and Treatments
- Colorectal Cancer Surgical Treatments
- Pediatric Urology and Nephrology Studies
- Bladder and Urothelial Cancer Treatments
- Urologic and reproductive health conditions
- Cancer-related molecular mechanisms research
- Breast Cancer Treatment Studies
- COVID-19 diagnosis using AI
- Renal cell carcinoma treatment
- Cancer Diagnosis and Treatment
- Medical Imaging Techniques and Applications
- Anatomy and Medical Technology
- Medical Image Segmentation Techniques
- Lung Cancer Diagnosis and Treatment
- Genital Health and Disease
- Cell Image Analysis Techniques
- Colorectal and Anal Carcinomas
Peking University
2020-2024
Peking University First Hospital
2020-2024
Abstract Background Accurate segmentation of pelvic bones is an initial step to achieve accurate detection and localisation bone metastases. This study presents a deep learning-based approach for automated normal bony structures in multiparametric magnetic resonance imaging (mpMRI) using 3D convolutional neural network (CNN). Methods retrospective included 264 mpMRI data obtained between 2018 2019. The manual annotations (which lumbar vertebra, sacrococcyx, ilium, acetabulum, femoral head,...
Graph-based approaches are successful for histology image classification tasks but still face many challenges, such as: 1) the lack of nuclei-level labels and significant variations between images make it extremely difficult to extract discriminative high-level nuclei features like type, texture micro-environment; 2) graph-based cannot handle large-scale cell graph nodes typically contained in images; 3) neural networks (GNNs) struggle learn long-range dependency graphs. To address above we...
Abstract Background Preoperative pelvic lymph node metastasis (PLNM) prediction can help clinicians determine whether to perform dissection (PLND). The purpose of this research is explore the feasibility diffusion-weighted imaging (DWI)-based radiomics for preoperative PLNM in PCa patients at nodal level. Methods MR images 1116 pathologically confirmed nodes (LNs) from 84 were enrolled. subjects divided into a primary cohort (67 with 192 positive and 716 negative LNs) held-out (17 43 165 4:1...
Cognitive fusion MRI-guided targeted biopsy (cTB) has been widely used in the diagnosis of prostate cancer (PCa). However, cTB relies heavily on operator's experience and confidence MRI readings. Our objective was to compare detection rates artificial intelligence-guided (AI-cTB) routine explore added value using AI for guidance cTB. This a prospective, single-institution randomized controlled trial (RCT) comparing clinically significant PCa (csPCa) between AI-cTB A total 380 eligible...
Pure ground-glass nodules (pGGNs) on chest CT representing invasive adenocarcinoma (IAC) warrant lobectomy with lymph node resection. For pGGNs other entities, close follow-up or sublobar resection without dissection may be appropriate.
Background: The aim of this study was to evaluate the effectiveness and safety real-time surgical navigation by three-dimensional (3D) virtual reconstruction models in robot-assisted laparoscopic pyeloplasty (RALP). Methods: Between November 2018 January 2020, 38 patients with ureteropelvic junction obstruction (UPJO) who underwent RALP were retrospectively enrolled. operations assisted real time 3D 16 patients, while 22 surgery without navigation. Based on whether had a prior intervention...
Prostatic stromal tumours of uncertain malignant potential (STUMPs) are rare prostate tumours. The purpose this study was to investigate the magnetic resonance imaging features STUMPs.A total 12 patients with STUMP confirmed pathology who underwent MRI from 2012 2020 were retrospectively reviewed. Pathological characteristics including histopathology and immunohistochemistry also recorded.Among STUMPs, detected in peripheral zone (41.7%[n = 5]) transitional (58.3% [n 7]) prostate. 8 cases...
Purpose To evaluate the efficacy of high b-value diffusion-weighted imaging (DWI) with a continuous-time random-walk (CTRW) diffusion model in determining pathological grade and variant histology (VH) bladder cancer (BCa). Methods A total 81 patients (median age, 70 years; range, 35-92 18 females; 66 grades; 30 VH) pathologically confirmed urothelial carcinoma were retrospectively enrolled underwent MRI on 3.0T scanner. Multi-b-value DWI was performed using 11 b-values. Three CTRW parameters...
The purpose of this study was to evaluate the nature ultrasound characteristics during mpMRI/TRUS cognitive fusion targeted biopsy (cTB). From 2023 2024, data from 502 lesions in 426 men who underwent combined systematic were analyzed. All had a Prostate Imaging Reporting and Data System (PI-RADS) score ≥ 3. primary endpoint detection rate prostate cancer (PCa) according PI-RADS score/ultrasound characteristics, categorized as benign or invisible (Bi), hypoechoic only (Ho), with...
Contouring prostate tumor in magnetic resonance images is a prerequisite for diagnosis. Automatically segmenting blurred lesion regions challenging and requires fully leveraging multi-parameter MR images. This paper proposes MFSL-Net, an end-to-end network that cascades two novel sub-networks: 1) modality fusion selectively fuses information of MRI modalities by expanding dual-stream CNN with spatial channel attention modules; 2) shape learning integrates context to recognize the edge while...
To evaluate the detection rates of prostate cancer (PCa) and clinically significant (CSPCa) via target biopsy (TB), systematic (SB), combined (CB) in patients with PI-RADS 5 lesions.Patients at least one lesion were retrospectively enrolled a prospectively collected database. The underwent multiparametric magnetic resonance imaging (mpMRI) followed by transrectal TB lesions SB. PCa CSPCa cores SB compared those CB.In 585 patients, revealed 560 cases (95.73%) 549 (93.85%). was detected T2...
Objective: The current study aimed to evaluate the predictive performances of anatomic staging system (AS) and prognostic (PS) proposed in 8th edition American Joint Committee on Cancer (AJCC) manual patients with pure mucinous breast cancer (PMBC). Methods: Clinicopathologic features follow-up information were collected from a total 3628 PMBC. Breast cancer-specific survival (BCSS) compared among different stage groups. Likelihood ratio (LR) χ(2), Akaike criterion (AIC) Harrell's...
Objective To evaluate the effect of neoadjuvant androgen deprivation therapy (ADT)for prostate cancer on diffusion weighted imaging base pathological results after radical prostatectomy. Methods Medical records 33 patients diagnosed with and treated prostatectomy between January 2016 September 2019 at Peking University First Hospital were retrospectively reviewed. Average age was 67.7 (49-81) years old. All underwent MRI examination before ADT. Results Mean volume ADT is 28.5...
As the most commonly used radiological examination for prostate disorder diagnosis, Magnetic Resonance Imaging (MRI) acquisition is very time-consuming. To accelerate MRI reconstruction while maintaining high quality, this paper provides MPTGAN, a multimodal prior-based triple-branch network. MPTGAN guides of time-consuming modality by utilizing time-efficient as prior knowledge, thus massive loss low and high-frequency information in under-sampled data could be replenished. In particular,...
Abstract Background Preoperative pelvic lymph node metastasis (PLNM) prediction can help clinicians determine whether to perform dissection (PLND). The purpose of this research is explore the feasibility diffusion-weighted imaging (DWI)-based radiomics for preoperative PLNM in PCa patients at nodal level. Methods MR images 1116 pathologically confirmed nodes (LNs) from 84 were enrolled. subjects divided into a primary cohort (67 with 192 positive and 716 negative LNs) held-out (17 43 165 4:1...
High-performance deep learning models require large amounts of data with high quality annotations for model training, while the labeling work usually takes a lot time experts. Meanwhile, inter-observer variability al-ways exist between from different experts and distribution shift acquired medical institutions. To address these challenges, we propose an end-to-end domain adaptive collaborative network multi-institutional prostate MRI segmentation. Specifically, introduce unpaired image...