breastscreeningai.github.io/evidence.html at main · BreastScreeningAI/breastscreeningai.github.io · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
<!DOCTYPE html>
<html class="no-js" lang="en">
  <head>
    <meta charset="utf-8" />
    <meta http-equiv="x-ua-compatible" content="ie=edge" />
    <title>Evidence | BreastScreening-AI</title>
    <meta name="description" content="Overall clinical evidence for BreastScreening-AI across pilots, publications and future validation studies." />
    <meta name="viewport" content="width=device-width, initial-scale=1" />
    <meta property="og:title" content="Evidence | BreastScreening-AI" />
    <meta property="og:description" content="Overall clinical evidence for BreastScreening-AI across pilots, publications and future validation studies." />
    <meta property="og:url" content="https://breastscreeningai.github.io/evidence.html" />
    <meta property="og:type" content="website" />
    <link rel="shortcut icon" type="image/x-icon" href="assets/images/favicon.svg" />
    <link rel="stylesheet" href="assets/css/bootstrap.min.css" />
    <link rel="stylesheet" href="assets/css/icons.css" />
    <link rel="stylesheet" href="assets/css/main.css" />
    <link rel="stylesheet" href="assets/css/evidence.css" />
    <link rel="stylesheet" href="assets/css/solutions.css" />
  </head>
  <body class="solution-page">
    <div class="preloader" aria-hidden="true"><div class="preloader-inner"><div class="preloader-icon"><span></span><span></span></div></div></div>
    <header class="header navbar-area"><div class="container"><div class="row align-items-center"><div class="col-lg-12"><div class="nav-inner">
      <nav class="navbar navbar-expand-lg">
        <a class="navbar-brand" href="index.html"><img src="assets/images/logo/logo-bs-main.svg" alt="BreastScreening-AI" /></a>
        <button class="navbar-toggler mobile-menu-btn" type="button" data-bs-toggle="collapse" data-bs-target="#navbarSupportedContent" aria-controls="navbarSupportedContent" aria-expanded="false" aria-label="Toggle navigation"><span class="toggler-icon"></span><span class="toggler-icon"></span><span class="toggler-icon"></span></button>
        <div class="collapse navbar-collapse sub-menu-bar" id="navbarSupportedContent"><ul id="nav" class="navbar-nav ms-auto"><li class="nav-item"><a href="#portfolio">Portfolio</a></li><li class="nav-item"><a href="#evidence-layers">Evidence</a></li><li class="nav-item"><a href="#journey">Validation</a></li><li class="nav-item"><a href="#priorities">Priorities</a></li></ul></div>
        <div class="button add-list-button"><a href="index.html" class="btn"><i class="lni lni-arrow-left"></i> Main Site</a></div>
      </nav>
    </div></div></div></div></header>

    <main>
      <section class="solution-hero" aria-labelledby="evidence-title"><div class="container">
        <p class="solution-kicker">Evidence</p>
        <h1 id="evidence-title">Overall evidence across research and clinical pilots</h1>
        <p class="solution-lead">A machine-learning and clinical evidence view of BreastScreening-AI across model development, human-AI design, controlled evaluation and exploratory clinical integration. Measured findings are separated from scale-up estimates.</p>
        <div class="solution-tags" aria-label="Evidence portfolio"><span>Curated Outputs</span><span>Study Participations</span><span>Hospitals Engaged</span><span>Research Timeline</span></div>
      </div></section>

      <nav class="solution-subnav" aria-label="Evidence sections"><div class="container"><a href="#portfolio">Research portfolio</a><a href="#evidence-layers">Evidence layers</a><a href="#journey">Validation timeline</a><a href="#priorities">Next priorities</a><a href="#voucher">Voucher pilot</a><a href="#boundaries">Boundaries</a></div></nav>

      <section id="portfolio" class="solution-section soft-blue"><div class="container">
        <div class="solution-heading"><p class="solution-label">Research portfolio</p><h2>Project-Related Evidence</h2><p>The portfolio combines scholarly outputs, protected intellectual property, human-AI studies and hospital engagement. Study participations may overlap across publications and are not presented as unique clinicians.</p></div>
        <div class="metric-grid evidence-portfolio-grid">
          <article class="metric-card"><span class="metric-value">14</span><h3>Curated outputs</h3><p>Four journal articles, five conference papers, one scientific exhibit, two theses and two patent families.</p><a href="publications.html">Review the catalogue <i class="lni lni-arrow-right"></i></a></article>
          <article class="metric-card"><span class="metric-value">12</span><h3>Scholarly outputs</h3><p>Peer-reviewed papers, scientific exhibit and academic theses connected to the project's research lineage.</p><a href="publications.html">Explore publications <i class="lni lni-arrow-right"></i></a></article>
          <article class="metric-card"><span class="metric-value">135</span><h3>Study participations</h3><p>Reported participation across human-centred, adoption, clinician-AI and exploratory pilot studies. Cohorts may overlap.</p><a href="publications.html">Trace the research lineage <i class="lni lni-arrow-right"></i></a></article>
          <article class="metric-card"><span class="metric-value">25+</span><h3>Hospitals engaged</h3><p>Internal business-development traction across hospital conversations, research engagement and collaboration activity; not 25 clinical validation sites.</p><a href="voucher.html">Review documented pilot activity <i class="lni lni-arrow-right"></i></a></article>
        </div>
      </div></section>

      <section id="evidence-layers" class="solution-section white"><div class="container">
        <div class="solution-heading"><p class="solution-label">Evidence layers</p><h2>One programme, several forms of validation</h2><p>The Evidence page now focuses on how the research components fit together. Detailed performance figures remain on the Platform, Workflow, Publications and Voucher pages where their study context is easier to preserve.</p></div>
        <div class="solution-grid">
          <article class="solution-card"><i class="lni lni-layers"></i><h3>Model evidence</h3><p>Multimodal fusion, weak supervision, lesion detection and external validation establish the technical research foundation.</p><a href="platform.html">Explore platform evidence <i class="lni lni-arrow-right"></i></a></article>
          <article class="solution-card"><i class="lni lni-users"></i><h3>Human-AI evidence</h3><p>Interaction, adoption, explanation and personalization studies examine how clinicians understand and use AI assistance.</p><a href="publications.html">Explore publications <i class="lni lni-arrow-right"></i></a></article>
          <article class="solution-card"><i class="lni lni-hospital"></i><h3>Clinical integration evidence</h3><p>Exploratory activities investigate workflow integration, usability, structured reporting and evidence-generation readiness.</p><a href="voucher.html">Startup Voucher Report <i class="lni lni-arrow-right"></i></a></article>
        </div>
      </div></section>

      <section id="journey" class="solution-section soft-green"><div class="container">
        <div class="solution-heading"><p class="solution-label">ML validation journey</p><h2>Model and design evidence developed together</h2><p>The assumed timeline maps published work to five validation tracks: interface design, human factors, clinical evaluation, model validation and clinical integration. It is a synthesis of the research record, not a formal regulatory development chronology.</p></div>
        <div class="evidence-chart-card evidence-timeline-card"><div id="validation-timeline-chart" class="evidence-plot evidence-timeline-plot" role="img" aria-label="Timeline of model validation, interface design, human factors and clinical integration from 2017 to 2026"></div></div>
        <div class="evidence-comparison evidence-journey">
          <article><span class="evidence-year">2017-2021</span><h3>Foundational multimodality research</h3><p>Early work established the human-centred, multimodal and intelligent-agent foundations for breast imaging decision support.</p><a href="publications.html">Explore publications <i class="lni lni-arrow-right"></i></a></article>
          <article><span class="evidence-year">2022</span><h3>Controlled clinician-AI evaluation</h3><p>A peer-reviewed study involved 45 clinicians from nine institutions and reported differences in errors, task time and clinician response.</p><a href="https://doi.org/10.1016/j.artmed.2022.102285" target="_blank" rel="noopener noreferrer">Read the study <i class="lni lni-arrow-top-right"></i></a></article>
          <article><span class="evidence-year">2023</span><h3>Personalized AI communication</h3><p>A CHI study evaluated assertiveness-based communication and reported faster mean task completion with a statistically significant result.</p><a href="https://doi.org/10.1145/3544548.3580682" target="_blank" rel="noopener noreferrer">Read the study <i class="lni lni-arrow-top-right"></i></a></article>
          <article><span class="evidence-year">2025-2026</span><h3>Clinical integration activities</h3><p>Hospital da Luz and CHTMAD activities examined integration, usability, triage support and structured clinical reporting in relevant environments.</p><a href="voucher.html">Open Voucher report <i class="lni lni-arrow-right"></i></a></article>
        </div>
        <div class="ml-validation-flow" aria-label="Machine-learning validation lifecycle">
          <article><span>01</span><h3>Define</h3><p>Intended users, modalities, decision points, endpoints and failure modes.</p></article>
          <article><span>02</span><h3>Develop</h3><p>Model, fusion, interaction and explanation components with versioned data.</p></article>
          <article><span>03</span><h3>Verify</h3><p>Technical performance, calibration, robustness, subgroup and error analysis.</p></article>
          <article><span>04</span><h3>Validate</h3><p>Clinician comparison, usability, workflow and prospective clinical endpoints.</p></article>
          <article><span>05</span><h3>Monitor</h3><p>Drift, overrides, safety signals, equity and post-deployment performance.</p></article>
        </div>
      </div></section>

      <section id="priorities" class="solution-section dark"><div class="container">
        <div class="solution-heading"><p class="solution-label">Validation priorities</p><h2>Turn the research portfolio into decision-grade evidence</h2><p>The next evidence phase should focus less on extrapolation and more on prospective, reproducible and institution-specific validation.</p></div>
        <div class="principle-grid">
          <article><h3>Lock intended use</h3><p>Define target users, modalities, patient populations, outputs and clinical decision points.</p></article>
          <article><h3>Freeze evaluation plans</h3><p>Prespecify endpoints, denominators, subgroup analyses, missing-data rules and statistical tests.</p></article>
          <article><h3>Validate externally</h3><p>Test across independent hospitals, scanner vendors, populations and workflow conditions.</p></article>
          <article><h3>Measure workflow</h3><p>Capture reading time, overrides, recalls, follow-up, workload and integration reliability.</p></article>
          <article><h3>Build economic evidence</h3><p>Use observed local outcomes, costs and implementation resources rather than headline extrapolations.</p></article>
          <article><h3>Monitor continuously</h3><p>Plan drift, safety, equity, cybersecurity and post-deployment performance surveillance.</p></article>
        </div>
      </div></section>

      <section id="voucher" class="solution-section soft-green"><div class="container">
        <div class="status-panel platform-status"><div><h3>Startup Voucher clinical activities</h3><p>The Voucher report documents the operation, technology-readiness context, Hospital da Luz exploratory results, CHTMAD fieldwork and specialist support. It provides the project-level context behind the newest evidence shown here.</p><a href="voucher.html" class="doi-link">Open the complete Voucher report <i class="lni lni-arrow-right"></i></a></div><div class="status-fact"><strong>TRL 5</strong><span>Project-level position</span><p>Progressing toward TRL 6 based on integration and usability activities. This is not an external certification, regulatory authorization or clinical deployment approval.</p></div></div>
      </div></section>

      <section id="boundaries" class="solution-section white"><div class="container">
        <div class="solution-heading"><p class="solution-label">Evidence boundaries</p><h2>What remains to be demonstrated</h2><p>A credible evidence strategy must make the gaps as visible as the positive results.</p></div>
        <div class="principle-grid">
          <article><h3>Prospective performance</h3><p>Larger, prespecified and prospective multicentre studies are needed to test diagnostic performance in intended-use populations.</p></article>
          <article><h3>Patient outcomes</h3><p>The available studies do not yet demonstrate reduced interval cancer, morbidity, mortality or unnecessary procedures.</p></article>
          <article><h3>Generalizability</h3><p>Performance requires validation across sites, modalities, scanner vendors, demographics and clinically relevant subgroups.</p></article>
          <article><h3>Workflow value</h3><p>Reading time, recall, throughput and workload effects need local baseline measurement and prospective comparison.</p></article>
          <article><h3>Health economics</h3><p>Cost-effectiveness, budget impact and realized return on investment have not yet been established.</p></article>
          <article><h3>Regulatory readiness</h3><p>Intended use, risk management, quality systems, cybersecurity and post-market monitoring require formal development.</p></article>
        </div>
      </div></section>

      <section class="solution-cta"><div class="container"><h2>Review the newest project evidence</h2><p>See the Startup Voucher operation for detailed pilot context, activities, limitations and supporting results.</p><a href="voucher.html" class="btn">Startup Voucher Report</a> <a href="publications.html" class="btn btn-outline-light business-cta-secondary">Explore publications</a></div></section>
    </main>

    <footer class="secondary-footer"><div class="container"><div class="secondary-footer-grid"><div class="secondary-footer-brand"><a href="index.html"><img src="assets/images/logo/logo-bs-ai-2.svg" alt="BreastScreening-AI" /></a><p>AI should strengthen, not replace, clinical judgement.</p><ul class="secondary-footer-social"><li><a href="https://www.linkedin.com/company/breastscreeningai" target="_blank" rel="noopener noreferrer" aria-label="BreastScreening-AI on LinkedIn"><i class="lni lni-linkedin-original"></i></a></li><li><a href="https://breastscreeningai.substack.com/feed" target="_blank" rel="noopener noreferrer" aria-label="BreastScreening-AI RSS feed"><i class="lni lni-rss-feed"></i></a></li></ul></div><nav class="secondary-footer-nav" aria-label="Business"><h2>Business</h2><ul><li><a href="platform.html">Platform</a></li><li><a href="workflow.html">Workflow</a></li><li><a href="https://www.crunchbase.com/organization/breastscreening-ai" target="_blank" rel="noopener noreferrer">Analytics</a></li><li><a href="https://www.f6s.com/breastscreeningai" target="_blank" rel="noopener noreferrer">Insights</a></li></ul></nav><nav class="secondary-footer-nav" aria-label="Support"><h2>Support</h2><ul><li><a href="https://github.com/BreastScreeningAI/meta/wiki" target="_blank" rel="noopener noreferrer">Documentation</a></li><li><a href="index.html#live-support">Live</a></li><li><a href="mailto:info@breastscreeningai.com">Contact</a></li><li><a href="faq.html">FAQs</a></li></ul></nav><nav class="secondary-footer-nav" aria-label="Company"><h2>Company</h2><ul><li><a href="https://github.com/BreastScreeningAI/meta" target="_blank" rel="noopener noreferrer">About</a></li><li><a href="story.html">Story</a></li><li><a href="https://wellfound.com/company/breastscreening-ai/jobs" target="_blank" rel="noopener noreferrer">Careers</a></li><li><a href="reports.html">Reports</a></li></ul></nav><nav class="secondary-footer-nav" aria-label="Legal"><h2>Legal</h2><ul><li><a href="media/legal/breastscreeningai_terms_and_conditions.pdf" target="_blank" rel="noopener noreferrer">Terms</a></li><li><a href="media/legal/breastscreeningai_privacy_policy.pdf" target="_blank" rel="noopener noreferrer">Privacy</a></li><li><a href="media/legal/breastscreeningai_cookie_policy.pdf" target="_blank" rel="noopener noreferrer">Cookies</a></li><li><a href="media/legal/breastscreeningai_disclaimer.pdf" target="_blank" rel="noopener noreferrer">Disclaimer</a></li></ul></nav><nav class="secondary-footer-nav" aria-label="Explore"><h2>Explore</h2><ul><li><a href="evidence.html">Evidence</a></li><li><a href="publications.html">Publications</a></li><li><a href="news.html">News</a></li><li><a href="media.html">Media</a></li></ul></nav></div><div class="secondary-footer-support"><div class="secondary-footer-funding"><a href="https://recuperarportugal.gov.pt/" target="_blank" rel="noopener noreferrer"><img src="assets/recognitions/startup_voucher_2025.png" alt="Startup Voucher 2025" /></a><a href="https://www.fct.pt/" target="_blank" rel="noopener noreferrer"><img src="assets/recognitions/2022_FCT_Logo_A_horizontal_branco_cropped.png" alt="Fundacao para a Ciencia e a Tecnologia" /></a></div><p class="secondary-footer-disclaimer">Research software and information only. BreastScreening-AI does not provide medical advice, diagnosis, or emergency services.</p></div></div></footer>
    <a href="#" class="scroll-top" aria-label="Back to top"><i class="lni lni-chevron-up"></i></a>
    <script src="assets/js/bootstrap.min.js"></script><script src="assets/js/wow.min.js"></script><script src="assets/js/main.js"></script><script src="https://cdn.plot.ly/plotly-2.35.2.min.js" charset="utf-8"></script><script src="assets/js/evidence.js"></script>
  </body>
</html>