agentleFS
aizechaizech/clinical-skills/.agents/skills/ai-report-assist/SKILL.md

ai-report-assist

Skill4 starsChanged 7 months agoMITView on GitHub

Guidance for AI-assisted structured reporting tools. Also use when the user mentions AI reporting, automated templating, speech-to-report, or wants to configure or optimize AI-assisted radiology reporting systems (RadAI, Abba, DeepRad).

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Start with standardized templates: - Ensure consistency

    line 279

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AI Report Assistance

You are an expert in AI-assisted radiology reporting. Your role is to help users configure, integrate, and optimize AI reporting tools.

Supported Platforms

PlatformFocusModality
RadAIStructured reporting automationCT, X-ray
AbbaSpeech recognition + structured reportingCT, MRI
DeepRadMulti-modality structured reportingCT, MRI, X-ray
DeepScribeAmbient AI documentationAll
ScribeAnywhereVoice-powered reportingAll

Key Concepts

AI Reporting Workflow

Image → AI Analysis → Finding Detection → Template Population → Radiologist Review → Signed Report

Structured Reporting Benefits

  • Consistent terminology
  • Complete documentation
  • Data extraction for analytics
  • Quality metrics
  • Research queries

RadAI Integration

API Configuration

import requests

RADAI_API = "https://api.radai.ai/v1"

def configure_radai(api_key, modality="ct"):
    """Configure RadAI API connection."""
    return {
        "base_url": RADAI_API,
        "headers": {
            "Authorization": f"Bearer {api_key}",
            "Content-Type": "application/json"
        },
        "default_modality": modality
    }

def submit_study_for_ai_report(config, study_uid, modality="ct"):
    """Submit study for AI-assisted reporting."""
    response = requests.post(
        f"{config['base_url']}/studies",
        headers=config["headers"],
        json={
            "study_uid": study_uid,
            "modality": modality,
            "report_type": "structured"
        }
    )
    return response.json()

Template Configuration

def configure_template(config, template_type="default"):
    """Configure reporting template."""
    templates = {
        "ct_chest": {
            "sections": ["lungs", "mediastinum", "pleura", "bones", "impression"],
            "required_fields": ["lungs.findings", "impression"],
            "measurement_fields": ["size", "attenuation", "volume"]
        },
        "ct_abdomen": {
            "sections": ["liver", "gallbladder", "pancreas", "spleen", "kidneys", "bowel", "impression"]
        },
        "ct_head": {
            "sections": ["brain", "ventricles", "basal_ganglia", "vessels", "bones", "impression"]
        }
    }
    return templates.get(template_type, templates["ct_chest"])

Retrieve AI Suggestions

def get_ai_suggestions(config, study_id):
    """Get AI-generated report suggestions."""
    response = requests.get(
        f"{config['base_url']}/studies/{study_id}/suggestions",
        headers=config["headers"]
    )
    return response.json()

# Response structure
{
    "study_id": "123",
    "findings": [
        {
            "anatomy": "right_upper_lobe",
            "finding": "nodule",
            "size_mm": 12,
            "location_detail": "RUL",
            "characteristics": {
                "margins": "spiculated",
                "attenuation": "solid"
            }
        }
    ],
    "impression_suggestion": "12mm spiculated nodule in right upper lobe, suspicious for malignancy.",
    "confidence": 0.89
}

Abba Integration

Speech Recognition Setup

def configure_abba(api_key, specialty="radiology"):
    """Configure Abba speech recognition."""
    return {
        "base_url": "https://api.abba.ai",
        "headers": {
            "Authorization": f"Bearer {api_key}"
        },
        "specialty": specialty,
        "format": "structured"
    }

def transcribe_dictation(config, audio_file):
    """Transcribe dictation with structured output."""
    with open(audio_file, "rb") as f:
        files = {"audio": f}
        response = requests.post(
            f"{config['base_url']}/transcribe",
            headers=config["headers"],
            files=files,
            data={"specialty": config["specialty"]}
        )
    return response.json()

DeepRad Integration

Multi-Modality Configuration

def configure_deeprad(api_key):
    """Configure DeepRad for multi-modality."""
    return {
        "base_url": "https://api.deeprad.ai",
        "api_key": api_key,
        "modalities": ["ct", "mri", "xray", "pet"]
    }

def get_structured_report(config, study_data, modality):
    """Get structured report for any modality."""
    response = requests.post(
        f"{config['base_url']}/report/{modality}",
        headers={"Authorization": f"Bearer {config['api_key']}"},
        json=study_data
    )
    return response.json()

Template Types

By Modality

ModalityTemplate TypeKey Elements
CT ChestLung-RADSNodule tracking, comparison
CT AbdomenLI-RADSLiver lesion assessment
CT HeadNo specificHemorrhage, stroke
MRI ProstatePI-RADSPI-RADS scoring
MRI LiverLI-RADSLI-RADS scoring
MammographyBI-RADSAssessment categories
X-ray ChestNo specificCritical findings

Template Structure

STANDARD_TEMPLATE = {
    "header": {
        "patient_id": "required",
        "study_date": "required",
        "accession": "required",
        "modality": "required",
        "clinical_history": "required"
    },
    "findings": {
        "anatomy": "free_text",
        "finding": "structured",
        "size": "measurement",
        "location": "structured",
        "characteristics": "structured"
    },
    "impression": {
        "primary": "required",
        "secondary": "optional",
        "recommendations": "optional"
    }
}

Integration with PACS

Workflow Integration

def setup_pacs_integration(pacs_url, ai_platform="radai"):
    """Set up PACS integration for AI reporting."""
    integration = {
        "pacs": {
            "url": pacs_url,
            "auto_submit": True,
            "receive_results": True
        },
        "ai_platform": ai_platform,
        "workflow": {
            "auto_populate": True,
            "require_review": True,
            "sign_immediately": False
        }
    }
    return integration

Auto-Populate Configuration

def configure_auto_populate(settings):
    """Configure auto-population behavior."""
    return {
        "populate_findings": settings.get("findings", True),
        "populate_impression": settings.get("impression", True),
        "populate_measurements": settings.get("measurements", True),
        "highlight_changes": settings.get("highlight_changes", True),
        "require_acknowledgment": settings.get("require_ack", True)
    }

Optimization Strategies

High Volume Practice

HIGH_VOLUME_CONFIG = {
    "auto_accept_normal": True,  # Accept normal AI reports
    "auto_populate": True,
    "require_review_abnormal": True,
    "batch_processing": True,
    "templates": "standardized"
}

Quality Focus

QUALITY_FOCUSED_CONFIG = {
    "auto_accept_normal": False,
    "auto_populate": True,
    "require_review_all": True,
    "double_read_option": True,
    "templates": "comprehensive"
}

Best Practices

  1. Start with standardized templates - Ensure consistency
  2. Enable auto-population gradually - Train radiologists on workflow
  3. Monitor accuracy - Track AI vs final report differences
  4. Customize templates - Adapt to your practice patterns
  5. Regular review - QA AI suggestions periodically

Troubleshooting

IssueSolution
AI not submittingCheck PACS integration
Slow responsesEnable caching
Incorrect findingsRetrain with local data
Template mismatchUpdate template mapping

Related Skills

  • structured-reporting: For report template details
  • pacs-workflow: For PACS integration
  • ai-quality-review: For AI output QA
  • radiology-report-analysis: For report analysis

Examples

Example 1: Enable AI Reporting

Enable AI-assisted reporting for CT chest studies using RadAI

Configuration:

config = configure_radai(api_key="your-key", modality="ct")
template = configure_template(config, "ct_chest")

Example 2: Review AI Suggestions

Review AI suggestions for study ACC123
suggestions = get_ai_suggestions(config, "ACC123")
# Present to radiologist for review
# Accept or modify suggestions

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.