Evaluating Conversational Image Segmentation for Medicine: Performance, Failure Modes, and a Fairness Audit Across Seven Modalities

Introduction Medical-image segmentation underpins quantitative diagnostics and research, yet state-of-the-art models remain task-specific and data-hungry. The recent emergence of powerful, multimodal large language models (LLMs) presents a generalizable option; however, their efficacy in the specialized medical domain remains largely unquantified. We aim to benchmark the foundational Gemini 2.5 Flash and Gemini 2.5 Flash-Lite models for zero-shot medical image segmentation and evaluate them for potential bias in performance.

The models were tested on 13,086 medical images spanning seven distinct imaging modalities (endoscopy, fundoscopy, dermoscopy, laparoscopy, ultrasound, radiography, and CT) and 10 clinically relevant segmentation targets (e.g., colorectal polyps, skin lesions, liver tumors). Prompts followed a standard template (“Segment the … in this image.”). Per-image Dice and Intersection-over-Union (IoU) were computed against publicly released expert masks. Bias was assessed on 4,057 dermoscopy images split by Individual-Typology-Angle (ITA) into Light (> 28°, n = 3,499) and Dark (≤ 28°, n = 558) groups.

Flash achieved a mean Dice=0.766 (IoU=0.680) for colorectal polyp segmentation, Dice=0.761 (IoU=0.672) for skin lesion segmentation, Dice=0.824 (IoU=0.736) for optic disc segmentation, and Dice=0.718 (IoU=0.616) for surgical tool segmentation, outperforming Flash-Lite by approximately 0.1 Dice points. Accuracy declined on low-contrast radiological tasks (Liver Mass CT Dice=0.071). In the fairness audit, Flash produced a successful mask for 2,839/3,499 light-tone images (81.2%) versus 378/558 dark-tone images (67.7%); χ² = 51.8, p < 0.001. Using all images, mean IoU=0.686 for light tones and IoU=0.591 for dark tones (Kruskal–Wallis H = 62.6, p < 0.001); Cliff’s δ = –0.208 (95% CI – 0.259 to –0.159).

Gemini 2.5 Flash delivers competitive accuracy on high-contrast photographic datasets at negligible cost. Performance is weaker on radiographic modalities (ultrasound, CT, chest radiography), and in dermoscopy, we observe lower accuracy on darker ITA skin-tone groups. This study informs the field where foundational LLMs are deployment-ready for medical image segmentation and where targeted debiasing or domain adaptation is required. Competing Interest Statement Nidhir Guggilla is an employee of OpenAI and holds stock/stock options in the company. Vivaswat Suresh is a prospective employee of CompoundEye. The employers had no role in the study design, data collection, analysis, manuscript preparation, or the decision to submit the manuscript for publication. Funding Statement This work was not supported by any sources of funding. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study used ONLY openly available human data that were originally located at: Kvasir-Seg: https://github.com/DebeshJha/Kvasir-SEG REFUGE2: https://www.kaggle.com/datasets/victorlemosml/refuge2 ISIC 2016, 2017, and 2018 Part 1: https://challenge.isic-archive.com/data/ AutoLaparo: https://autolaparo.github.io/ BUSI: https://www.kaggle.com/datasets/sabahesaraki/breast-ultrasound-images-dataset SIIM-ACR: https://www.kaggle.com/datasets/jesperdramsch/siim-acr-pneumothorax-segmentation-data LiTS 2017: https://www.kaggle.com/datasets/andrewmvd/lits-png I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability All data produced in the present study are available upon reasonable request to the authors https://github.com/DebeshJha/Kvasir-SEG https://www.kaggle.com/datasets/victorlemosml/refuge2 https://challenge.isic-archive.com/data/ https://www.kaggle.com/datasets/sabahesaraki/breast-ultrasound-images-dataset https://www.kaggle.com/datasets/jesperdramsch/siim-acr-pneumothorax-segmentation-data