Prostate cancer treatment is evolving quickly, with new technologies helping doctors improve accuracy, reduce treatment burden, and support better patient recovery. Two recent clinical reports highlight major progress in the use of MRI-guided Transurethral Ultrasound Ablation (TULSA) for localized prostate cancer.

The first study examined how artificial intelligence (AI) can automate prostate segmentation during treatment planning. The second reported early outcomes from the CAPTAIN randomized controlled trial, comparing TULSA with robotic prostatectomy.

Together, these findings suggest that TULSA, especially when combined with AI, may help streamline workflows while offering less invasive treatment options.

What Is TULSA?

TULSA is a minimally invasive prostate cancer treatment that uses directional ultrasound energy delivered through the urethra under real-time MRI guidance. Physicians can precisely heat and destroy targeted prostate tissue while monitoring temperature and protecting surrounding structures.

Unlike traditional surgery, TULSA does not require incisions and may preserve urinary continence and sexual function in selected patients.

Why Prostate Segmentation Matters in Treatment Planning

Before TULSA can begin, physicians must outline the prostate gland on MRI scans. This process is known as prostate segmentation or contouring.

Traditionally, this is done manually by radiologists or urologists. However, manual contouring has several limitations:

• Time-consuming workflow
• Repetitive image review
• Reader-to-reader variability
• Potential delays in treatment planning

AI-assisted contouring aims to solve these problems by automatically generating accurate prostate outlines that clinicians can review and adjust if needed.

AI Segmentation Study: Faster Planning Without Sacrificing Accuracy

Researchers validated two deep learning models integrated into TULSA planning software:

• First-generation model using SEResNet34
• Updated model using SAM2-UNet

These systems were trained using T2-weighted MRI scans from public datasets and previous TULSA treatment cases.

How the Study Was Conducted

The validation study included:

• 51 historical TULSA cases
• 3 urologists
• 5 radiologists

Each physician segmented prostate images manually and then with AI assistance after a four-week washout period.

Accuracy was measured using the Dice similarity score, which compares AI contours with expert reference contours.

Key Results of the AI Planning Study

Urologists Improved Speed and Accuracy

With AI support:

• Dice score improved from 0.909 to 0.919
• Planning time dropped from 5.3 minutes to 3.6 minutes

Radiologists Maintained Accuracy and Worked Faster

With AI support:

• Accuracy remained equivalent
• Planning time improved from 3.6 minutes to 2.5 minutes

Updated AI Model Performed Extremely Well

The second-generation SAM2-UNet model achieved:

• Dice score of 0.921
• Superior performance compared with urologists
• Non-inferior performance compared with radiologists

Real-World Clinical Impact

After regulatory clearance, researchers compared:

• 153 AI-assisted TULSA planning cases
• 962 prior non-AI cases at the same institutions

Result:

• 11% reduction in treatment planning time

Physicians reported that in treatment-naïve prostates, AI contouring was accurate enough to reduce workload and effort significantly.

What This Means for Patients and Clinics

AI in TULSA planning may lead to:

• Faster treatment preparation
• More standardized prostate contouring
• Reduced physician workload
• Improved scheduling efficiency
• Consistent planning quality across centers

As imaging volumes rise globally, automation tools like this could become essential in modern prostate cancer programs.

CAPTAIN Trial: TULSA vs Robotic Prostatectomy

A second major report came from the CAPTAIN trial (NCT05027477), a randomized controlled study comparing TULSA with robotic prostatectomy for intermediate-risk localized prostate cancer.

Historically, trials comparing focal ablation with surgery have struggled to recruit participants, making this study particularly important.

Study Design

Patients were randomized in a 2:1 ratio:

• 148 received TULSA
• 64 received robotic prostatectomy

Total treated patients:

• 212 men across 23 sites

This exceeded the planned enrollment target of 201 patients.

Early Perioperative Results

Blood Loss

TULSA patients had:

• 0 mL median blood loss

Robotic prostatectomy patients had:

• 100 mL median blood loss

Hospital Stay

TULSA:

• 0.29 days

Robotic surgery:

• 1.24 days

This suggests many TULSA patients were discharged the same day.

Recovery and Pain

Patients reported:

• Less decline in overall health after TULSA
• Reduced post-procedure pain
• Faster return to normal activities over 30 days

Catheter Duration

One tradeoff noted:

• TULSA catheterization lasted 13 days
• Robotic surgery catheterization lasted 8 days

Why These Findings Matter

For many men with intermediate-risk prostate cancer, treatment decisions often balance:

• Cancer control
• Side effects
• Recovery time
• Lifestyle impact
• Urinary and sexual function preservation

These early CAPTAIN data indicate TULSA may offer a less invasive pathway with faster short-term recovery compared with robotic prostatectomy.

Long-term cancer control and functional outcomes are still being evaluated.

Future of Prostate Cancer Treatment

These two reports show a powerful trend in prostate care:

1. AI Is Improving Workflow

Automation reduces repetitive tasks and speeds treatment planning.

2. Minimally Invasive Therapies Are Expanding

TULSA offers image-guided precision treatment without surgery.

3. Personalized Treatment Is Becoming Reality

Subtotal, hemi-gland, and whole-gland approaches may allow physicians to tailor therapy based on tumor location and patient priorities.

Final Thoughts

The combination of AI-assisted planning and MRI-guided TULSA treatment may reshape how localized prostate cancer is managed. Faster workflows, reduced recovery burden, and highly targeted treatment are attractive benefits for both patients and clinicians.

While more long-term data are needed, these early results are promising and support continued adoption of advanced imaging, AI, and minimally invasive technologies in urology.

Source

1. Woodrum D, Mynderse L, Raman S ...Abstract No. 50 Randomized Controlled Trial of MRI-Guided Transurethral Ultrasound Ablation (TULSA) Versus Robotic Prostatectomy for Intermediate-Risk Prostate Cancer: Ablation Parameters and Initial Perioperative Outcomes. Journal of Vascular and Interventional Radiology, 37

Disclaimer

This article is for educational and informational purposes only and does not constitute medical advice, diagnosis, or treatment recommendations. Patients should consult a qualified urologist, oncologist, or healthcare professional before making treatment decisions regarding prostate cancer.