{
  "schema_version": 1,
  "record_url": "https://ramtin-mojtahedi.github.io/publications/spectral-adapters-sam-crlm-2026/",
  "last_verified": "2026-09-13",
  "profile": {
    "name": "Ramtin Mojtahedi",
    "orcid": "https://orcid.org/0000-0002-3953-3256",
    "google_scholar": "https://scholar.google.com/citations?user=KjUrlGUAAAAJ&hl=en",
    "website": "https://ramtin-mojtahedi.github.io/"
  },
  "publication": {
    "id": "spectral-adapters-sam-crlm-2026",
    "title": "Spectral Adapters for Segment Anything Model-based Segmentation of Colorectal Liver Metastases in Computed Tomography",
    "year": 2026,
    "publication_date": "2026-09-10",
    "group": "2026",
    "status": "Preprint · arXiv · Not peer reviewed",
    "authors": [
      "Ramtin Mojtahedi",
      "Mohammad Hamghalam",
      "Jacob J. Peoples",
      "Natalie Gangai",
      "Mithat Gonen",
      "Yun Shin Chun",
      "HyunSeon Christine Kang",
      "Richard K. G. Do",
      "Amber L. Simpson"
    ],
    "authors_html": "<strong>R. Mojtahedi</strong>, M. Hamghalam, J. J. Peoples, N. Gangai, M. Gonen, Y. S. Chun, H. C. Kang, R. K. G. Do, and A. L. Simpson",
    "citation": "arXiv:2609.11703 [cs.CV] (2026). Version 1, submitted 10 September 2026.",
    "citation_plain": "arXiv:2609.11703 [cs.CV] (2026). Version 1, submitted 10 September 2026.",
    "doi": "10.48550/arXiv.2609.11703",
    "url": "https://arxiv.org/abs/2609.11703",
    "link_label": "arXiv ↗",
    "manual": true,
    "source": "Public arXiv record, verified 13 September 2026",
    "last_verified": "2026-09-13",
    "title_aliases": [],
    "venue": "arXiv",
    "volume": "",
    "issue": "",
    "pages": "",
    "article_number": "",
    "type": "preprint",
    "peer_reviewed": false,
    "publisher": "arXiv",
    "keywords": [
      "colorectal liver metastases",
      "computed tomography",
      "Segment Anything Model",
      "spectral adapters",
      "DiSECT",
      "SiGA",
      "parameter-efficient fine-tuning",
      "medical image segmentation"
    ],
    "note": "Public preprint; not peer reviewed. 13 pages, 1 figure, 4 tables.",
    "detail_url": "/publications/spectral-adapters-sam-crlm-2026/",
    "open_access_url": "https://arxiv.org/pdf/2609.11703",
    "open_access_label": "Preprint PDF",
    "open_access_license": "",
    "preprint_url": "https://arxiv.org/abs/2609.11703",
    "code_url": "",
    "code_label": "",
    "pubmed_url": "",
    "pmc_url": "",
    "citation_pdf_url": "https://arxiv.org/pdf/2609.11703",
    "bibtex_type": "misc",
    "bibtex_key": "Mojtahedi2026SpectralAdapters",
    "eprint": "2609.11703",
    "primary_class": "cs.CV",
    "abstract": "Accurate segmentation of colorectal liver metastases (CRLM) in contrast-enhanced computed tomography (CT) is important for response assessment, surgical planning, and follow-up. We propose two parameter-efficient spectral adapters for the Segment Anything Model (SAM): the Directional Spectral Adapter (DiSECT) and Spectral Instance-Guided Adapter (SiGA). DiSECT uses singular value decomposition of frozen weights to constrain residual updates to leading spectral directions, while SiGA adds global and input-conditioned gating through a multilayer perceptron. We evaluate these methods on 446 contrast-enhanced CT volumes (355 training, 91 testing) and compare them with LoRA, QLoRA, convolutional adapters (CAD), and a 3D nnU-Net baseline. Experiments consider single-point, three-point, bounding-box, and no-prompt regimes. SiGA achieves the best single-point performance with a Dice score of 0.77, IoU of 0.69, and HD95 of 35.39 mm. Under no-prompt inference, SiGA reaches 0.76 Dice, 0.68 IoU, and 46.76 mm HD95, comparable to the nnU-Net baseline (0.758 Dice). DiSECT uses only 0.14 million trainable parameters. These results show that spectral adapters can efficiently adapt SAM for CRLM segmentation while retaining strong accuracy with limited trainable parameters."
  },
  "citation": {
    "plain_text": "Ramtin Mojtahedi, Mohammad Hamghalam, Jacob J. Peoples, Natalie Gangai, Mithat Gonen, Yun Shin Chun, HyunSeon Christine Kang, Richard K. G. Do, and Amber L. Simpson. “Spectral Adapters for Segment Anything Model-based Segmentation of Colorectal Liver Metastases in Computed Tomography.” arXiv:2609.11703 [cs.CV] (2026). Version 1, submitted 10 September 2026. https://doi.org/10.48550/arXiv.2609.11703",
    "bibtex": "@misc{Mojtahedi2026SpectralAdapters,\n  author = {Ramtin Mojtahedi and Mohammad Hamghalam and Jacob J. Peoples and Natalie Gangai and Mithat Gonen and Yun Shin Chun and HyunSeon Christine Kang and Richard K. G. Do and Amber L. Simpson},\n  title = {{Spectral Adapters for Segment Anything Model-based Segmentation of Colorectal Liver Metastases in Computed Tomography}},\n  year = {2026},\n  howpublished = {arXiv},\n  doi = {10.48550/arXiv.2609.11703},\n  url = {https://arxiv.org/abs/2609.11703},\n  note = {Public preprint; not peer reviewed. 13 pages, 1 figure, 4 tables.},\n  eprint = {2609.11703},\n  archivePrefix = {arXiv},\n  primaryClass = {cs.CV}\n}",
    "ris": "TY  - GEN\nTI  - Spectral Adapters for Segment Anything Model-based Segmentation of Colorectal Liver Metastases in Computed Tomography\nAU  - Ramtin Mojtahedi\nAU  - Mohammad Hamghalam\nAU  - Jacob J. Peoples\nAU  - Natalie Gangai\nAU  - Mithat Gonen\nAU  - Yun Shin Chun\nAU  - HyunSeon Christine Kang\nAU  - Richard K. G. Do\nAU  - Amber L. Simpson\nPY  - 2026\nDA  - 2026-09-10\nT2  - arXiv\nDO  - 10.48550/arXiv.2609.11703\nUR  - https://arxiv.org/abs/2609.11703\nKW  - colorectal liver metastases\nKW  - computed tomography\nKW  - Segment Anything Model\nKW  - spectral adapters\nKW  - DiSECT\nKW  - SiGA\nKW  - parameter-efficient fine-tuning\nKW  - medical image segmentation\nN1  - Public preprint; not peer reviewed. 13 pages, 1 figure, 4 tables.\nER  -\n"
  },
  "related_records": [
    {
      "title": "Parameter-Efficient Fine-Tuning of Foundation Models for Liver Tumor Segmentation in CT",
      "year": 2026,
      "url": "https://ramtin-mojtahedi.github.io/publications/foundation-models-liver-tumor-segmentation-2026/"
    },
    {
      "title": "Radiomic Signatures from Baseline CT Predict Chemotherapy Response in Unresectable Colorectal Liver Metastases",
      "year": 2026,
      "url": "https://ramtin-mojtahedi.github.io/publications/radiomic-signatures-chemotherapy-response-2026/"
    },
    {
      "title": "The Search for CT Imaging Subtypes of Colorectal Liver Metastases and the Impacts of Slice Thickness",
      "year": 2025,
      "url": "https://ramtin-mojtahedi.github.io/publications/ct-imaging-subtypes-slice-thickness-2025/"
    }
  ]
}