AI System Detects Prostate Cancer During Routine CT Scans
|
By MedImaging International staff writers Posted on 15 Nov 2021 |

Image: CT slices of a patient with prostate cancer (a), and one without (b) (Photo courtesy of SVHM/ RMIT)
A new study describes an artificial intelligence (AI) based framework that can rapidly spot incidental prostate tumors during abdominal or pelvic scans.
Developed at RMIT University (RMIT; Melbourne, Australia) and St. Vincent's Hospital (SVHM; Melbourne, Australia), the new convolutional neural network (CNN) is designed for incidental computer aided detection (CADe) of clinically significant prostate cancer in patients undergoing a computerized tomography (CT) scan of the abdomen or pelvis for other reasons. The dataset used to develop the CNN consisted of 139 clinically significant prostate cancer patients and 432 controls.
The results showed that the proposed CNN pipeline achieved an area under the receiver operating characteristic curve (ROC-AUC) of 0.88 on CT, significantly higher than that of two radiologists (0.61 and 0.70) set on the same task. In addition, the results confirmed that the screening capabilities of CT-based pipelines, when combined with deep learning CNNs, are comparable to those of magnetic resonance imaging (MRI)-based diagnostic pipelines. The study was published in the November 2021 issue of Scientific Reports.
“Australia doesn’t have a screening program for prostate cancer, but armed with this technology, we hope to catch [such] cases early in patients who are scanned for other reasons,” said study co-author radiologist Mark Page, MD, of SVHM. “For example, emergency patients who have CT scans could be simultaneously screened for prostate cancer. If we can detect it earlier and refer them to specialist care faster, this could make a significant difference to their prognosis.”
In Australia, prostate cancer is responsible for approximately 12% of all male cancer deaths, as the slowly-growing tumors often go unnoticed for years. It is typically difficult to spot prostate cancer in CT images, and the radiation doses required make CT unsuitable as a screening modality. However, if men need to undergo an abdominal or pelvic scan for other reasons, CADe could help spot prostate cancer and let clinicians initiate early treatment.
Related Links:
RMIT University
St. Vincent's Hospital
Developed at RMIT University (RMIT; Melbourne, Australia) and St. Vincent's Hospital (SVHM; Melbourne, Australia), the new convolutional neural network (CNN) is designed for incidental computer aided detection (CADe) of clinically significant prostate cancer in patients undergoing a computerized tomography (CT) scan of the abdomen or pelvis for other reasons. The dataset used to develop the CNN consisted of 139 clinically significant prostate cancer patients and 432 controls.
The results showed that the proposed CNN pipeline achieved an area under the receiver operating characteristic curve (ROC-AUC) of 0.88 on CT, significantly higher than that of two radiologists (0.61 and 0.70) set on the same task. In addition, the results confirmed that the screening capabilities of CT-based pipelines, when combined with deep learning CNNs, are comparable to those of magnetic resonance imaging (MRI)-based diagnostic pipelines. The study was published in the November 2021 issue of Scientific Reports.
“Australia doesn’t have a screening program for prostate cancer, but armed with this technology, we hope to catch [such] cases early in patients who are scanned for other reasons,” said study co-author radiologist Mark Page, MD, of SVHM. “For example, emergency patients who have CT scans could be simultaneously screened for prostate cancer. If we can detect it earlier and refer them to specialist care faster, this could make a significant difference to their prognosis.”
In Australia, prostate cancer is responsible for approximately 12% of all male cancer deaths, as the slowly-growing tumors often go unnoticed for years. It is typically difficult to spot prostate cancer in CT images, and the radiation doses required make CT unsuitable as a screening modality. However, if men need to undergo an abdominal or pelvic scan for other reasons, CADe could help spot prostate cancer and let clinicians initiate early treatment.
Related Links:
RMIT University
St. Vincent's Hospital
Latest Radiography News
- Rapid X-Ray Test Quantifies Pulmonary Regurgitation After Tetralogy of Fallot Repair
- AI Tool Flags Osteoporosis Risk from Routine Chest X-Rays
- Simple Chest X-Ray Measure Predicts Survival After Lung Cancer Surgery
- AI Detection Tool Improves Identification of Lobular Breast Cancer
- New Contrast Agent Enables Low-Dose X-Ray Joint Imaging
- AI Boosts Breast Cancer Detection and Cuts Screening Workload
- AI Tool Predicts Breast Cancer Risk Years Ahead Using Routine Mammograms
- Routine Mammograms Could Predict Future Cardiovascular Disease in Women
- AI Detects Early Signs of Aging from Chest X-Rays
- X-Ray Breakthrough Captures Three Image-Contrast Types in Single Shot
- AI Generates Future Knee X-Rays to Predict Osteoarthritis Progression Risk
- AI Algorithm Uses Mammograms to Accurately Predict Cardiovascular Risk in Women
- AI Hybrid Strategy Improves Mammogram Interpretation
- AI Technology Predicts Personalized Five-Year Risk of Developing Breast Cancer
- RSNA AI Challenge Models Can Independently Interpret Mammograms
- New Technique Combines X-Ray Imaging and Radar for Safer Cancer Diagnosis
Channels
MRI
view channel
AI Approach Could Shorten Advanced Brain MRI Scans by Up to 90%
Long acquisition times for advanced brain magnetic resonance imaging (MRI) can limit access, extend waiting lists, and disrupt clinical workflows. Reducing data requirements without sacrificing image fidelity... Read more
Cardiac MRI Measure Improves Risk Prediction in Tricuspid Regurgitation
Tricuspid regurgitation, in which blood flows back from the right ventricle into the right atrium, can lead to progressive right-sided heart failure. Clinicians need reliable ways to gauge severity and... Read moreUltrasound
view channelAI Robotic Ultrasound System Automates Echocardiography and Improves Consistency
Echocardiography, an ultrasound examination of the heart, is central to diagnosing and managing cardiovascular disease. Many services struggle with limited availability of skilled sonographers, variable... Read more
Whole Cross-Section Ultrasound System Enables Operator-Independent Imaging
Conventional ultrasound is central to bedside imaging but is limited by a narrow field of view and operator variability. Comprehensive cross-sectional assessment typically requires computed tomography... Read moreNuclear Medicine
view channel
Targeted PET Platform Guides Osteosarcoma Resection and Margin Verification
Osteosarcoma, an aggressive primary bone cancer that mainly affects children and adolescents, demands wide excision to prevent local recurrence. Surgeons must achieve negative margins while preserving... Read more
Portable PET System Enables Real-Time Bedside Guidance for Biopsies and Ablations
Interventional radiology procedures typically rely on ultrasound, X-ray fluoroscopy, or computed tomography for image guidance. These modalities visualize anatomy but offer limited molecular information,... Read moreGeneral/Advanced Imaging
view channel
AI Tool Enables Real-Time Diffuse Optical Tomography for Brain Lesion Detection
Diffuse optical tomography is a noninvasive imaging technique that uses near-infrared light to detect internal abnormalities such as cerebral hemorrhage and tumors. Its clinical utility for real-time ... Read moreNew SPECT/CT Method Differentiates Inflammation from Fibrosis in Interstitial Lung Disease
Interstitial lung disease (ILD) encompasses more than 200 disorders that inflame or scar the lung interstitium and can lead to progressive respiratory failure. Determining whether active inflammation is... Read moreImaging IT
view channel
Interactive AI Tool Supports Explainable Lung Nodule Assessment
Lung cancer is a leading cause of cancer mortality, and timely characterization of pulmonary nodules on chest computed tomography (CT) is essential for directing care. Interpreting nodule morphology demands... Read more
Breast Imaging Software Enhances Visualization and Tissue Characterization in Challenging Cases
Breast imaging can be particularly challenging in cases involving small breasts or implants, where image reconstruction and tissue characterization may be limited. Clinicians also need reproducible analysis... Read more
New Google Cloud Medical Imaging Suite Makes Imaging Healthcare Data More Accessible
Medical imaging is a critical tool used to diagnose patients, and there are billions of medical images scanned globally each year. Imaging data accounts for about 90% of all healthcare data1 and, until... Read more
Global AI in Medical Diagnostics Market to Be Driven by Demand for Image Recognition in Radiology
The global artificial intelligence (AI) in medical diagnostics market is expanding with early disease detection being one of its key applications and image recognition becoming a compelling consumer proposition... Read moreIndustry News
view channel
GE HealthCare Showcases AI-Enabled Nuclear Medicine Portfolio at SNMMI 2026
Nuclear medicine is expanding rapidly as health systems adopt theranostics and broaden access to radiopharmaceuticals, increasing demand for scalable operations and consistent diagnostic confidence.... Read more
GE HealthCare Highlights AI-Supported Radiation Therapy Tools at ESTRO 2026
At the European Society for Radiotherapy and Oncology (ESTRO) 2026 Congress in Stockholm, GE HealthCare is highlighting Intelligent Radiation Therapy (iRT), MIM Software innovations, and BK Medical surgical... Read more





 Guided Devices.jpg)

