BiCycle

Summary

• BiCycle is a mature, clinically validated automated brachytherapy treatment planning system, significantly reducing planning time while maintaining or improving plan quality.
• Initially developed for prostate HDR brachytherapy, it is now used in clinical practice for image-guided adaptive cervical brachytherapy, meeting EMBRACE II standards and in compliance with MDR regulations.
• It is part of a larger push towards AI-driven, automated treatment pipeline for hyperfast plan preparation for brachytherapy treatments.

About our project

What is BiCycle?

• It is a fully automated, multicriteria optimization treatment planning for high‑dose‑rate (HDR) brachytherapy, in-house developed based on the original Erasmus‑iCycle.
• BiCycle extends this technology specifically to prostate and cervical cancer brachytherapy, translating clinical planning objectives (“wish‑lists”) into precise dose distributions by optimizing dwell times.
• It is designed to conform to EMBRACE II guidelines for cervical cancer planning and has been validated both internally (Erasmus MC) and externally (Tata Memorial Hospital in India).

Key Achievements & Clinical Results

1. Rapid Planning
   For cervical HDR brachytherapy, BiCycle produces autoplan in 9.4 min, compared to ~45 minutes manually, including plan evaluation and fine tuning from the treating physician.
2. Plan Quality Superior or Comparable to Manual
   - Blind comparisons showed that autoplan quality is equal or better than manual plans.
   - BiCycle autoplan consistently scored higher across clinician-rated domains versus manual plans, especially for overall quality, CTV and OAR, with comparable loading pattern.
3. Clinical Adoption
   Since August 2023, BiCycle has been used in routine clinical workflows at Erasmus MC for cervical brachytherapy treatments.

Background: Erasmus‑iCycle Roots

BiCycle builds on Erasmus‑iCycle — an in‑house developed, multicriteria optimizer dating back to the early 2010s.

• Originally for external beam IMRT/VMAT, covering sites like prostate, head & neck, and lung with automatic beam angle and fluence optimization.
• Later extended to CyberKnife, proton radiotherapy, MRI-linac and brachytherapy (BiCycle) with the same philosophy: use a defined objective priority list to automatically generate Pareto-optimal plans.

Our research focus is to improve plan quality, as well as the patient and employee experience for brachytherapy treatments.

By automating the treatment planning process, we aim to:
• Shorten time-to-treatment
• Reduce waiting time for patient, reducing pain and anxiety
• Improve plan quality
• Reduce workload
• Reduce inter-operator variability
• Improve capacity

We focus on translating our research into clinically usable tools. BiCycle is currently in routine use for cervical cancer patients at Erasmus MC and has been successfully evaluated at external sites, demonstrating generalizability and robustness across institutions.

Our work lies at the intersection of medical physics, computer science, and clinical oncology. We collaborate closely with radiation oncologists, technicians, medical physicists, developers and AI researchers to ensure our tools are not only technically sound but also clinically meaningful.

Tata Memorial Hospital, Mumbai, India

1. Breedveld S, Bennan ABA, Aluwini S, Schaart DR, Kolkman-Deurloo IK, Heijmen BJM. Fast automated multi-criteria planning for HDR brachytherapy explored for prostate cancer. Phys Med Biol. 2019;64:205002.
2. Oud M, Kolkman-Deurloo IK, Mens JW, Lathouwers D, Perko Z, Heijmen B, et al. Fast and fully-automated multi-criterial treatment planning for adaptive HDR brachytherapy for locally advanced cervical cancer. Radiother Oncol. 2020;148:143-50.
3. Fully automated adaptive multi-criteria planning for EMBRACE II cervical cancer brachytherapy. Radiotherapy and OncologyVol. 170SupplementS390Published in issue: May, 2022 R. Bijman L. Rossi S. Breedveld I. Kolkman-Deurloo J. Mens M. Christianen M. Huge ∙ B. Heijmen ∙ R. Nout. 10.1016/S0167-8140(22)02582-8. ESTRO 2022
4. L. Rossi, S. Chopra, R. Bijman, P. Mittal, S. Panda, S. Breedveld, R. Nout, B. Heijmen. External validation of automated adaptive planning for EMBRACE II cervical cancer brachytherapy. Radiotherapy and Oncology. May 2023 Volume 182, Supplement 1. https://doi.org/10.1016/S0167-8140(23)08544-4 ESTRO 2023
5. Automated MCO to assess dosimetric price-to-pay for control on loading pattern in cervical cancer BT. L. Rossi ∙ N. Panico ∙ S. Breedveld ∙ N. Lanconelli ∙ R. Nout ∙ B. Heijmen Radiotherapy and Oncology Volume 182, Supplement 1S1940-S1942May 2023 ESTRO 2023
6. Clinician preferred, fast autoplanning in cervical cancer brachytherapy using BiCycle. Linda Rossi, Rik Bijman, Henrike Westerveld, Miranda Christianen, Lorne Luthart, Michèle Huge, Inger-Karine olkman - Deurloo, Jan Willem Mens, Huda Abusaris, Raymond de Boer, Sebastiaan Breedveld, Ben Heijmen, Remi Nout. Radiotherapy and Oncology Volume 194, Supplement 1, May 2024, Pages S215-S217. https://doi.org/10.1016/S0167-8140(24)00985-X ESTRO 2024
7. Rule-based AI automated adaptive treatment planning for image guided cervical cancer brachytherapy. Linda Rossi, Rik Bijman, Supriya Chopra, Prachi Mittal, Subhajit Panda, Henrike Westerveld, Miranda Christianen, Inger-Karine Kolkman-Deurloo, Sebastiaan Breedveld, Remi Nout, Ben Heijmen. Brachytherapy Under Press
8. Prospective evaluation of AI-based BiCycle autoplanning for advanced cervical cancer brachytherapy Linda Rossi, Rik Bijman, Henrike Westerveld, Miranda Christianen, Lorne Luthart, Michèle Huge, Inger-Karine Kolkman-Deurloo, Jan Willem Mens, Huda Abusaris, Raymond de Boer, Sebastiaan Breedveld, Ben Heijmen, Remi Nout. Radiother Oncol. 2025 210 (2025) 111029

Linda Rossi
Henrike Westerveld
Ben Heijmen
Miranda Christianen
Inger-Karine Kolkman-Deurloo
Lorne Luthart
Michele Huge
Sebastiaan Breedveld
Andras Zolnay
Remi Nout