{
  "docId": "019dd923-5e88-73ef-bd5c-f6fdd8f57895",
  "docSlug": "3a892393f72c1ff5",
  "documentTitle": "2019 Air Street Capital The State of AI Report 2019",
  "authorId": "AirStreetCapital",
  "authorName": "Nathan Benaich and Ian Hogarth",
  "documentKindSlug": "consulting-deck",
  "documentKindLabel": "Consulting deck",
  "sourceTypeSlug": "vc_research",
  "sourceTypeLabel": "VC research",
  "presentationDate": null,
  "orientation": "landscape",
  "aspectRatio": 1.777,
  "pageNumber": 30,
  "pageCount": 136,
  "prevPage": 29,
  "nextPage": 31,
  "slideType": "industry_trends",
  "function": "analyze_data",
  "density": "dense",
  "nDataPoints": 5,
  "notes": "The slide uses ROC curves to compare model performance against human radiologists across five medical conditions.",
  "elementsJson": [
    "headline_text",
    "action_title",
    "bullet_list",
    "scatter_plot",
    "logo_grid"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5c-f6fdd8f57895/30",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5c-f6fdd8f57895#slide-30",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": ">600k chest x-rays have been published to boost model performance, but dataset issues remain.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-41ab-72d9-bbc3-e6f1b31ed2e2",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.3,
        "w": 0.9,
        "x": 0.05,
        "y": 0.28
      },
      "kind": "chart",
      "text": "ROC curves for five conditions: Atelectasis, Cardiomegaly, Consolidation, Edema, Pleural Effusion",
      "attrs": null,
      "subkind": "line",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "4bf2de63-4130-4be3-889a-52efd3b963ed",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.3,
        "w": 0.95,
        "x": 0.02,
        "y": 0.63
      },
      "kind": "list",
      "text": "Deep learning models for imaging diagnostics fit datasets well, but they have difficulties generalising to new data distributions. Despite improved documentation to this new dataset, label definitions are shallow.\nThere are challenges with extracting labels using NLP from doctors notes: Its error-prone and suffers from the lack of information contained in radiology reports, with 5-15% error rates in most label categories.\nSignificant number of repeat scans, with 70% of the scans coming from 30% of the patients. This reduces the effective size of the dataset and its diversity, which will impact the generalisability of trained models.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "6f25c7ab-2a1b-4411-9024-3d5d52c0a03c",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "metric",
      "text": "AUC: 0.97",
      "attrs": null,
      "subkind": "primary",
      "toolName": "Quantification",
      "toolSlug": "quantification",
      "confidence": null,
      "componentId": "019dd952-41ab-72d9-bbc3-ea8542966d7d",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.03,
        "w": 0.8,
        "x": 0.02,
        "y": 0.22
      },
      "kind": "title",
      "text": ">600k chest x-rays have been published to boost model performance, but dataset issues remain.",
      "attrs": null,
      "subkind": "action-title",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "bf7080e8-5527-40a0-87d9-1b1175991715",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.05,
        "w": 0.6,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "Deep learning in medicine: The bigger the dataset, the better the model?",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7b08187c-5fed-4dad-862a-5ff99adfc0ac",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Action Titles",
      "slug": "action-titles",
      "agent": "Architect",
      "layer": "slide",
      "matchId": "019dd95a-0fd5-7148-8ecc-e935e2ab4418",
      "evidence": "Title poses a hypothesis question.",
      "confidence": 70
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 92,
      "from": 5,
      "beatId": "019dd95a-0682-776c-8e34-ce57e5bbe574",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Definitions through Research/Talent/Industry — fact-dense case studies and benchmarks.",
      "position": 1,
      "confidence": 70,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 92,
      "from": 9,
      "beatId": "019dd95a-0682-776c-8e34-ddbccb391af1",
      "arcName": "Voyage and Return",
      "arcSlug": "voyage-return",
      "beatName": "The Unknown",
      "beatSlug": "voyage-return-the-unknown",
      "evidence": "Research/Talent/Industry frontiers explored.",
      "position": 2,
      "confidence": 45,
      "parentBeatName": "Complication",
      "parentBeatSlug": "complication"
    }
  ],
  "loops": [
    {
      "to": 33,
      "from": 28,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3c-a49628ab8218",
      "evidence": "Six contiguous medical AI case studies with quantitative outcomes (0.97 AUC, 73%, 80%).",
      "position": 5,
      "objective": "Compile clinical-grade ML wins across imaging, cardiology, neuro and chemistry",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 75,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}