arXiv subjects

Benjamin Movsas

Publications and source records attributed to Benjamin Movsas.

8 recordsLinked to original sources

arXiv2025

Automated stereotactic radiosurgery planning using a human-in-the-loop reasoning large language model agent

Stereotactic radiosurgery (SRS) demands precise dose shaping around critical structures, yet black-box AI systems have limited clinical adoption due to opacity concerns. We tested whether chain-of-thought reasoning improves agentic planning in a retrospective cohort of 41 patients with brain metastases treated with 18 Gy single-fraction SRS. We developed SAGE (Secure Agent for Generative Dose Expertise), an LLM-based planning agent for automated SRS treatment planning. Two variants generated plans for each case: one using a non-reasoning model, one using a reasoning model. The reasoning variant showed comparable plan dosimetry relative to human planners on primary endpoints (PTV coverage, maximum dose, conformity index, gradient index; all p > 0.21) while reducing cochlear dose below human baselines (p = 0.022). When prompted to improve conformity, the reasoning model demonstrated systematic planning behaviors including prospective constraint verification (457 instances) and trade-off deliberation (609 instances), while the standard model exhibited none of these deliberative processes (0 and 7 instances, respectively). Content analysis revealed that constraint verification and causal explanation concentrated in the reasoning agent. The optimization traces serve as auditable logs, offering a path toward transparent automated planning.

cs.AI↗

arXiv2025

Characterization and ex vivo application of flexible 2D scintillating coatings in ultra-high dose rate electron beams for FLASH radiotherapy

The increasing interest in FLASH-RT has lead to the conversion of linear accelerators to enable ultra-high dose rate (UHDR) beams for preclinical research. Dosimetry hereof remains challenging with several crucial aspects missing. This work shows the challenges for real-time 2D UHDR dosimetry, and aims to present a solution in the context of preclinical irradiations in non-homogeneous UHDR electron beams. An experimental camera-scintillation sheet combination, was used to investigate the spatial dose distribution of a converted UHDR Varian Trilogy. The dosimetric system was characterized by variation of the number of pulses and source to surface distance (SSD) and its application was investigated by variation of bolus thickness and ambient light intensity. The challenges of prelcinical real time 2D dosimetry with scintillating coatings were assessed by ex vivo irradiations of a rat brain, mouse hindlimb and whole body mouse. Radiochromic EBT XD film was used as passive reference dosimeter. The coating showed a linear response with the number of pulses in the absence and presence of transparent bolus, up to 3 cm thick, and with the inverse squared SSD. The presence of ambient light reduces the signal-background ratio. The sheet showed to have sufficient flexibility to be molded on the subjects' surface, following its curvatures. Linearity with number of pulses was preserved in a preclinical setting. For small field sizes the light output became too low, resulting in noisy dose maps. The system showed robust within 5% for camera set up differences. Calibration of the system was complicated due to set up variations and the inhomogeneity of the beam. We showed the need for 2D real-time dosimetry to determine beam characteristics in non-homogeneous UHDR beams using a preclinical setting. We presented one solution to meet this need with scintillating based dosimetry.

physics.med-ph↗

arXiv2025

Autonomous Radiotherapy Treatment Planning Using DOLA: A Privacy-Preserving, LLM-Based Optimization Agent

Radiotherapy treatment planning is a complex and time-intensive process, often impacted by inter-planner variability and subjective decision-making. To address these challenges, we introduce Dose Optimization Language Agent (DOLA), an autonomous large language model (LLM)-based agent designed for optimizing radiotherapy treatment plans while rigorously protecting patient privacy. DOLA integrates the LLaMa3.1 LLM directly with a commercial treatment planning system, utilizing chain-of-thought prompting, retrieval-augmented generation (RAG), and reinforcement learning (RL). Operating entirely within secure local infrastructure, this agent eliminates external data sharing. We evaluated DOLA using a retrospective cohort of 18 prostate cancer patients prescribed 60 Gy in 20 fractions, comparing model sizes (8 billion vs. 70 billion parameters) and optimization strategies (No-RAG, RAG, and RAG+RL) over 10 planning iterations. The 70B model demonstrated significantly improved performance, achieving approximately 16.4% higher final scores than the 8B model. The RAG approach outperformed the No-RAG baseline by 19.8%, and incorporating RL accelerated convergence, highlighting the synergy of retrieval-based memory and reinforcement learning. Optimal temperature hyperparameter analysis identified 0.4 as providing the best balance between exploration and exploitation. This proof of concept study represents the first successful deployment of locally hosted LLM agents for autonomous optimization of treatment plans within a commercial radiotherapy planning system. By extending human-machine interaction through interpretable natural language reasoning, DOLA offers a scalable and privacy-conscious framework, with significant potential for clinical implementation and workflow improvement.

physics.med-ph↗

arXiv2024

Hybrid Student-Teacher Large Language Model Refinement for Cancer Toxicity Symptom Extraction

Large Language Models (LLMs) offer significant potential for clinical symptom extraction, but their deployment in healthcare settings is constrained by privacy concerns, computational limitations, and operational costs. This study investigates the optimization of compact LLMs for cancer toxicity symptom extraction using a novel iterative refinement approach. We employ a student-teacher architecture, utilizing Zephyr-7b-beta and Phi3-mini-128 as student models and GPT-4o as the teacher, to dynamically select between prompt refinement, Retrieval-Augmented Generation (RAG), and fine-tuning strategies. Our experiments on 294 clinical notes covering 12 post-radiotherapy toxicity symptoms demonstrate the effectiveness of this approach. The RAG method proved most efficient, improving average accuracy scores from 0.32 to 0.73 for Zephyr-7b-beta and from 0.40 to 0.87 for Phi3-mini-128 during refinement. In the test set, both models showed an approximate 0.20 increase in accuracy across symptoms. Notably, this improvement was achieved at a cost 45 times lower than GPT-4o for Zephyr and 79 times lower for Phi-3. These results highlight the potential of iterative refinement techniques in enhancing the capabilities of compact LLMs for clinical applications, offering a balance between performance, cost-effectiveness, and privacy preservation in healthcare settings.

cs.CL↗

arXiv2024

Iterative Prompt Refinement for Radiation Oncology Symptom Extraction Using Teacher-Student Large Language Models

This study introduces a novel teacher-student architecture utilizing Large Language Models (LLMs) to improve prostate cancer radiotherapy symptom extraction from clinical notes. Mixtral, the student model, initially extracts symptoms, followed by GPT-4, the teacher model, which refines prompts based on Mixtral's performance. This iterative process involved 294 single symptom clinical notes across 12 symptoms, with up to 16 rounds of refinement per epoch. Results showed significant improvements in extracting symptoms from both single and multi-symptom notes. For 59 single symptom notes, accuracy increased from 0.51 to 0.71, precision from 0.52 to 0.82, recall from 0.52 to 0.72, and F1 score from 0.49 to 0.73. In 375 multi-symptom notes, accuracy rose from 0.24 to 0.43, precision from 0.6 to 0.76, recall from 0.24 to 0.43, and F1 score from 0.20 to 0.44. These results demonstrate the effectiveness of advanced prompt engineering in LLMs for radiation oncology use.

cs.CL↗

arXiv2023

An Introduction to Natural Language Processing Techniques and Framework for Clinical Implementation in Radiation Oncology

Natural Language Processing (NLP) is a key technique for developing Medical Artificial Intelligence (AI) systems that leverage Electronic Health Record (EHR) data to build diagnostic and prognostic models. NLP enables the conversion of unstructured clinical text into structured data that can be fed into AI algorithms. The emergence of the transformer architecture and large language models (LLMs) has led to remarkable advances in NLP for various healthcare tasks, such as entity recognition, relation extraction, sentence similarity, text summarization, and question answering. In this article, we review the major technical innovations that underpin modern NLP models and present state-of-the-art NLP applications that employ LLMs in radiation oncology research. However, these LLMs are prone to many errors such as hallucinations, biases, and ethical violations, which necessitate rigorous evaluation and validation before clinical deployment. As such, we propose a comprehensive framework for assessing the NLP models based on their purpose and clinical fit, technical performance, bias and trust, legal and ethical implications, and quality assurance, prior to implementation in clinical radiation oncology. Our article aims to provide guidance and insights for researchers and clinicians who are interested in developing and using NLP models in clinical radiation oncology.

cs.CL↗

arXiv2022

Prostate Cancer Malignancy Detection and localization from mpMRI using auto-Deep Learning: One Step Closer to Clinical Utilization

Automatic diagnosis of malignant prostate cancer patients from mpMRI has been studied heavily in the past years. Model interpretation and domain drift have been the main road blocks for clinical utilization. As an extension from our previous work where we trained a customized convolutional neural network on a public cohort with 201 patients and the cropped 2D patches around the region of interest were used as the input, the cropped 2.5D slices of the prostate glands were used as the input, and the optimal model were searched in the model space using autoKeras. Something different was peripheral zone (PZ) and central gland (CG) were trained and tested separately, the PZ detector and CG detector were demonstrated effectively in highlighting the most suspicious slices out of a sequence, hopefully to greatly ease the workload for the physicians.

eess.IV↗

arXiv2020

A Deep Dive into Understanding Tumor Foci Classification using Multiparametric MRI Based on Convolutional Neural Network

Deep learning models have had a great success in disease classifications using large data pools of skin cancer images or lung X-rays. However, data scarcity has been the roadblock of applying deep learning models directly on prostate multiparametric MRI (mpMRI). Although model interpretation has been heavily studied for natural images for the past few years, there has been a lack of interpretation of deep learning models trained on medical images. This work designs a customized workflow for the small and imbalanced data set of prostate mpMRI where features were extracted from a deep learning model and then analyzed by a traditional machine learning classifier. In addition, this work contributes to revealing how deep learning models interpret mpMRI for prostate cancer patients stratification.

cs.CV↗