26
COLLAGE Project Intelligent System for Analysis and Diagnosis of Collagen-Based Artefacts www.collage.com.ro collag e Elena Badea Project Director Mid term results

Intelligent System for Analysis and Diagnosis of Collagen ...uefiscdi.gov.ro/userfiles/file/PARTENERIATE/Competite...COLLAGE Project Intelligent System for Analysis and Diagnosis of

  • Upload
    others

  • View
    11

  • Download
    0

Embed Size (px)

Citation preview

  • COLLAGE Project

    Intelligent System for Analysis and Diagnosis of Collagen-Based Artefacts

    www.collage.com.ro

    collage

    Elena Badea Project Director

    Mid term results

  • COLLAGE Consortium

    collage

    Coordinator Institution - National Research & Development Institute for

    Textile and Leather (INCDTP), ICPI Division, Bucharest Team Leader: Lucretia Miu; Team members: Claudiu Sendrea, Elena Poiana,

    Gabriela Neagulov, Marius Purcea

    P1 - National Institute for Research & Development in Electrical

    Engineering (ICPE-CA), Bucharest

    Team Leader: Petru Budrugeac; Team members: Andrei Cucos, Laura Chiose,

    Carmen Stafanescu

    P2 - National Museum of Romanian History, Centre of Research &

    Scientific Investigation, Bucharest

    Team Leader: Irina Petroviciu; Team members: Cristina Carsote

    P3 - Faculty of Electronics, Telecommunications & Information

    Technology, University Politehnica of Bucharest

    Team Leader: Ovidiu Grigore; Team members: Valentin Velican, Ionut Bornoiu

    P4 - Mira Telecom s.r.l., Bucharest

    Team Leader: Silviu Ciobanu; Team members: Oana Miu

  • THE PREMISES

    Historical and cultural objects made of collagen include; parchment, in the form of scrolls,

    documents and codices; leather, in a huge variety of items such as bindings, wall tapestry,

    storage vessels, musical instruments, luggage, objects of personal use such as shoes and

    garments, and mummified skins of animals or humans.

    Documents made of parchment are the most important bearers of our recorded history from

    early times such as the Dead Sea Scrolls (408 BC - 318 AD) and the Magna Carta (1215)

    until the modern U.S. Declaration, Constitution, and Bill of Rights (1787), and present important

    state documents and parliamentary records.

    Leather objects are an infinite source of information of historical and cultural interest; they

    illustrate the evolution of social customs, habits,

    aesthetics and technology, but also the

    perpetuation of popular and religious traditions.

    On the other hand, collagen is an integral part of the human

    body and the study of mummies can supply new informed

    directions for anthropological, paleopathological, genetic and

    medical research.

    All these objects and artefacts represent repositories of tangible-intangible cultural heritage

    which can offer answers on the state of health and beliefs of past civilisations.

    co

    lla

    ge

  • THE CHALLENGE OF THE COLLAGE PROJECT FILLING THE GAP BETWEEN RESEARCH AND ITS APPLICATIONS FOR

    CULTURAL HERITAGE PRESERVATION

    co

    lla

    ge

    COLLAGE MAIN OBJECTIVE aims to develop a unified kit of assessment tools for conservators, conservation

    scientists and scientists to evaluate the level of damage in collagen-based heritage and

    archaeological objects using a multi-disciplinary and multi-scale approach ranging from

    molecular, nano and mesoscale to macro scale visual assessment.

    Additionally, the project

    (i) takes advantage of previous results from national, international and EU projects and

    develop a parchment and leather damage database

    (ii) critically reviews practices in the preparation of archival material and their

    conservation and restoration treatments in order to trace differences in traditions in

    different geographical areas and establish markers for identification and anti-

    counterfeiting purposes .

    2002 – 2005 EC FP5 Framework

    2006 – 2009 CIPE - Italy

    2006 - 2008 2008 - 2011 2007 – 2009

  • THE CHALLENGE COLLAGE – SPECIFIC OBJECTIVES

    co

    lla

    ge

    1. Development of a new measurement system for the automatic

    detection of shrinkage activity of collagen fibres in laboratory

    and in situ: imageMHT system

    2. Development of an intelligent system for assisted damage

    diagnosis CLEAR diagnostic software

    3. Development of a comprehensive parchment and leather

    damage database

    4. Establishment of a unified calibrated measurement protocol

    enabling inter-laboratory comparison

    5. Identification of analytical markers for early damage

    identification and anti-counterfeiting purposes

    6. Definition of quantitative protocols for damage quantification

  • THE APPROACH c

    oll

    ag

    e

    Accelerated

    Ageing / Exposure

    temperature

    humidity

    UV and Vis light

    air pollutants

    (NOx, SO2)

    Parchment

    - New calf and goat parchment

    - Historical parchment

    Leather

    - New calf and sheep leather - (tannins: mimosa bark, quebracho wood,

    chestnut wood)

    - Historical leather

    1-32 weeks 8 - 64 days

  • THE APPROACH c

    oll

    ag

    e

    Molecular level Multiscale analyses: Fibre level

    fibre bundles

    100-500 m

    SEM, DSC ATR/FTIR NMR MHT method

    fibres morphology

    thermal denaturation

    Badea E., Della Gatta G. et al. Study of deterioration of historical parchments by various thermal analysis techniques, complemented by

    SEM, FTIR, UV-Vis-NIR and unilateral NMR investigations, J. Therm. Anal. Calorim. 91, 17-27 (2008).

  • THE APPROACH Translate analytical data into practically relevant measures

    30

    35

    40

    45

    50

    55

    60

    65

    70

    TO

    10

    2-1

    TO

    10

    2 -

    2

    TO

    10

    4 -

    1

    TO

    10

    4 -

    2

    TO

    10

    6-1

    TO

    10

    6-2

    TO

    10

    7-1

    TO

    10

    7-2

    TO

    10

    8-1

    TO

    10

    8-2

    TO

    10

    9-1

    TO

    10

    9-2

    TO

    11

    0 -

    1

    TO

    11

    0 -

    2

    TO

    11

    1 -

    1

    TO

    11

    1 -

    2

    Te

    mp

    era

    ture

    / °

    C

    Historical samples

    Shrinkage activity of collagen fibres

    Parchment bindings, Ordinati Collection (1325 to

    TO109_1 TO106_1

    TO108_1 TO102_1 TO102_2

    TO104_2 TO108_2

    TO102_2 TO109_1

    TO107_2 TO102_2 TO111_2 TO108_1

    TO106_1 TO104_2

    TO106_2 TO108_2

    TO111_1 TO109_2

    TO102_1

    TO107_1 TO102_2

    TO102_1 TO111_2

    TO107_1 TO104_1

    TO106_2 TO107_2

    TO111_1

    TO111_2 TO108_1

    0 10 20 30 40 50 60 70

    T1/ms

    His

    toro

    cal s

    am

    ple

    s

    NMR longitudinal relaxion time T1 of historical

    parchments bindings, Ordinati Collection (1325 to

    1499) Historical Archives of the City of Turin

    DSC denaturation of collagen fibrils in

    parchment samples exposed to 80 °C and

    40% and 80% RH in alternate days

    1800 1600 1400 1200 1000

    0,0

    0,1

    0,2

    0,3

    0,4

    A

    Wavenumber / cm-1

    new parchment

    CR14: NO2 2 weeks

    CR15: NO2 4 weeks

    CR16: NO2 8 weeks

    CR17: NO2 16 weeks

    ATR-IR spectra of parchments exposed to

    NOx(50 ppm) for 2 to 16 weeks

    microscopic

    scale

    mesoscopic

    scale

    nanoscopic

    scale

    co

    lla

    ge

    molecular

    scale

  • co

    lla

    ge

    Damage assessment: from macro to nanoscale

    Damage quantifying Early damage detection Fake identification

    gelatine

    ordered triple helix

    Collagen denaturation & gelatinisation

    Molecular manifestation

    disordered

    triple helix

    3D packing

    loose

    peptide chains

    conformational change

    fragmentation

    Differential Scanning Calorimetry

    (micro DSC)

    ordered triple helix

    disordered

    triple helix

    gelatine

    microDSC peak of a historical parchment binding

    The State Archives of Florence, Italy

    Total Damage Ranking Equation

    Stot = 0.2 S[Td] + 0.5 S[DH] + 0.3 S[IS]

    E. Badea, G. Della Gatta, T. Usacheva, Effects of temperature and relative humidity on fibrillar collagen within parchment: a micro

    Differential Scanning Calorimetry (micro DSC) study, Polymer Degradation and Stability, 97, 346-353 (2012).

  • co

    lla

    ge

    Damage assessment: from macro to nanoscale

    Damage quantifying by DSC and SEM

    Ranking of a group of archival bookbindings in 4 damage categories:

    (1) no damage; (2) minor damage; (3) medium damage and (4) heavy damage

    Sample no. 4 comprehensively displays increasing damage of its flap (F4), cover (C4) and spine (S4)

    E. Badea, G. Della Gatta, L. Miu, C. Carşote, I. Petroviciu, R. Larsen, M. Odlyha, A. Braghieri, S. Benedetto, Damage assessment

    of historical parchments: a protocol for preventive conservation. Restitutio, Bilingual Conservation-restoration bulletin, No. 5-6, p.

    61-74, Muzeul Satului “Dimitrie Gusti”, Bucureşti (2012), ISSN 2065-2992.

  • co

    lla

    ge

    Damage assessment: from macro to nanoscale

    Damage quantifying Early damage detection Fake identification

    deteriorated sample

    I(AI)/I(AII) ~ 1 new parchment / leather

    deteriorated material

    Δν~ 90 cm-1 new parchment / leather

    deteriorated material

    I(AI)/I(AII) ~ 1 new parchment / leather

    E. Badea, W. Vetter, I. Petroviciu, C. Carsote, L. Miu, M. Schreiner, G. Della Gatta, How parchment responds to temperature and

    relative humidity: a combined micro DSC, MHT, SEM and ATR-FTIR study, in Proceedings Book of ICAMS 2012 (SCOPUS indexed

    volume), Certex, ISSN 2068-0783, Bucharest, p. 487-492 (2012).

  • co

    lla

    ge

    Damage assessment: from macro to nanoscale

    Damage quantifying Early damage detection Fake identification

    40 80 120 160 200 240

    -0.8

    -0.6

    -0.4

    -0.2

    0.0

    T/0C

    he

    at

    flo

    wW

    .g-1

    Dehydration

    Denaturation

    ENDO

    New vegetable leather

    30 60 90 120 150 180 210 240

    -0.8

    -0.6

    -0.4

    -0.2

    0.0

    Gospel - Bucharest - XVIII century

    he

    at

    flo

    wW

    .g-1

    T/0C

    Dehydration

    Denaturation 1

    Denaturation 2

    ENDO

    1 2 3 4

    200

    210

    220

    230

    240

    250

    1- new parchments

    2- old parchments

    3 - new leathers

    4 - old leathers

    Td/0

    C

    Heavily damaged old leathers: Td< 227

    0C Damaged old leathers:

    2270C ≤ Td ≤2340C

    New and artifcially aged Td > 234

    0C.

    P. Budrugeac, A. Cucos, L. Miu, Use of thermal analysis methods to

    asses the damage in the bookbindings of some religious books from

    XVIII century, stored in Romanian libraries, J Therm Anal Calorim

    (2014) 116:141–149.

  • co

    lla

    ge

    Interdisciplinary approach

    Analytical outcomes (chemistry, technology)

    Image algorithms (ICT)

    Software Hardware Con serva tion

    Sus taina ble

    Unification of all these assessment methods to create a reference toolkit

    for conservators and scientists alike to assess the healthiness of objects /

    artefacts, the impact of past and future conservation treatments, and of

    climate change.

    Applications Development

    Image MHT system

    • image capturing device (ICD)

    • precisely controlled heating block (HP)

    • software commanding both heating

    element and image capturing

    CLEAR diagnostic system • analytical database

    • software tool for evaluation and

    ranking damage, identifying main

    causes of deterioration, fingerprint the

    studied object

  • co

    lla

    ge

    Image MHT system vs classic MHT method

    MHT method: thermal microscopy technique that measures the shrinkage activity of collagen fibres

    Shrinkage is the macroscopic manifestation of thermal denaturation

    triple helix random coil

    In the first two intervals, A1 and B1,

    shrinkage discretely occurs in

    individual fibers and generally display

    higher activity in B1 interval. Fiber

    mass cooperatively shrinks in the C

    interval.

    The starting temperature of this

    interval is the shrinkage temperature,

    Ts. Generally, shrinkage activity levels

    off through B2 and A2 intervals.

    New collagen-based samples undergo the following changes as a function of temperature:

    no activityA1B1CB2A2complete shrinkage

    E. Badea, D.V. Poulsen Sommer, K. Mühlen Axelsson, R. Larsen , A. Kurysheva, L. Miu, G. Della Gatta, Damage ranking in historical

    parchments: from microscopic study of fibres structure to collagen denaturation assessment by micro DSC, e-Preservation Science,

    9, 97-109 (2012).

  • co

    lla

    ge

    Image MHT system vs classic MHT method

    MHT method: a simple test for evaluating collagen deterioration

    Shrinkage is visualised through a stereo microscope using reflected light as the

    motion of the fibres dispersed in water under constant heating rate (2 °C min-1).

    MHT measurement system

    1: Caloris heating plate

    1a: window slot for inserting the sample

    placed on a microscope slide with a

    concavity. Fibres are preliminary thoroughly

    wetted and separated in demineralised water

    for 10 min.

    1b: window for viewing under the microscope

    2: FP90 Central Processer for temperature

    programming

    3: Steromicroscope equipped with a digital

    camera

    4: PC

    1b

    1a

    2

    3 4

  • co

    lla

    ge

    Image MHT system and classic MHT method

    Mid Term Results

    - Hardware: prototype of the precisely controlled heating block

    - Software: experimental model for collagen fibres tracking and shrinkage intervals

    automatic detection

    - Testing of the experimental model

    imageMHT method: automated system that measures the shrinkage activity of collagen fibres using image processing techniques

    Advantages

    - More reliable results secured by:

    - ALGORITHMS QUALITY, higher number of observed fibres, lower heating rate

    which enables to better discriminate between various motion types

    - AVOID HUMAN ERRORS ( poor experience, fatigue, sight variability/problems,

    bad mood, etc.)

    - Dramatic reduction of the operator time

    - Portability and in-situ measurements

    - Patentability

    imageMHT – mainstream analytical technique for the primary evaluation of collagen-based artefacts, historical and archaelogical objects

  • co

    lla

    ge

    Image MHT system

    Heating block prototype

    Improved features

    •Heating/cooling interval: DT = 15 – 120 °C

    (Mettler Toledo heating plate: DT = 25 – 100 °C, but the system is

    stable only above 30 °C)

    • Variable temperature increment: 1, 2, 5 and 10 °Cmin-1

    (Mettler Toledo heating plate: 2 Cmin-1 )

    New functionalities

    • Quick cooling to room temperature

    • Miniaturized low-cost system

  • co

    lla

    ge

    Image MHT system

    Image experimental model

    Optical flow algorithms

    Step 1: Finding the “good features to track”

    As fibers are almost transparent and have thus the color of the background the

    best features to be tracked are their edges.

    Step 2: Tracking the interest points in each sequence using the optical flow tracker.

    Automatic detection of shrinkage intervals using

    image algorithms for movement detection

    Optical flow is the distribution of apparent velocities of movement of brightness

    patterns in an image.

    E. Badea, O.A. Miu, V. Velican, C. Sendrea, Intelligent system for damage diagnosis of collagen based artefacts. Presentation at the

    11th Mediterranean Conference on Calorimetry and Themal Analysis, Athens, Greece, 12-15 June 2013, not yet published.

  • co

    lla

    ge

    ImageMHT test results Ts detected by optical flow algorithms (orange)

    compared with Ts detected by the human eye (brown)

    The mean absolute error of the estimated Ts value was 1.1 °C for a series of 38 historical leather samples,

    less than the experimental accuracy of 2 °C determined for the classic MHT method by R. Larsen

    0

    10

    20

    30

    40

    50

    60

    70

    80

    90

    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

    Te

    mp

    era

    ture

    , 0C

    Sample number

    E. Badea, C. Carsote, O.A. Miu, V. Velican, G. Della Gatta, Damage assessment approaches for collagen-based artefacts: simple

    tests and sophisticated analyses. Invted lecture at the 2nd Conference “Men and Books: From Microorganisms to Mega-

    organisms” 28 April - 1 May 2014. Published in Cnference Proceedings Book, Verlag Berger, Horn, Austria (in print, expected for

    the second half of 2014).

  • co

    lla

    ge

    CLEAR diagnostic system

    Analytical database

    Features

    • component of the diagnosis software

    • elaborate and organise a huge amount of data (analytical and descriptive) on collagen

    artefacts of the Romanian patrimony with a common standard

    • group the sub-samples from the same object and facilitate data correlation

    • identify various deterioration patterns and suggest the most appropriate technique and

    procedure to detect such a pattern

    • accounts for the animal species and age of each object

    Leather Object

    New/Historical

    Sample 1

    Reference

    Analysis

    FTIR1

    Results

    MicroDSC1

    Results

    MHT1

    Results

    DSC1

    Results

    Sample 2

    Artificially Aged

    Artificial Aging Procedures

    UV

    32 hours

    Temperature

    70°C, 2 weeks

    Analysis

    FTIR2

    Results

    MicroDSC2

    Results

    MHT2

    Results

    DSC2

    Results

    Advantage: support for monitoring and managing of the museum/library/archive environment

    and prioritization, and selection of the most appropriate methods for conservation to obtain the

    longest life for collection with the lowest resource employment

  • Related research: bilateral collaborations

    • Assessment and mitigation of climate impact on library and archival heritage:

    experience, research, innovation (LIBER)

    Romania – Hungary bilateral collaboration PN II 671/2013

    • Quantitative assessment of environmental impact on collagen-based materials for low-

    energy climate control in archives and museums (ENVICOLL) Romania – France, bilateral

    collaboration PN II 713/2013)

    • Advanced techniques and interdisciplinary studies for improved assessment of

    historical parchment documents (ParIS)

    Romania – Italy, bilateral collaboration PN II 638/2013

    co

    lla

    ge

  • co

    lla

    ge

    Young researchers career

    Enrolment for PhD programmes related to the project topic

    Thesis on Damage assessment of historical parchment and leather

    Thesis on Collagen based materials analysis using unilateral Nuclear Magnetic Resonance

    Participation in International Training Schools and Workshops

    Summer School on Thermal Analysis and Calorimetry, 11th Mediterranean Conference on Calorimetry

    and Thermal Analysis (MEDICTA 2013), Athens, Greece, 12 – 15 June 2013

    Ampere Summer School on Analyzing Cultural Heritage using Portable Magnetic Resonance, organised

    by the team of Prof Bernhard Blümich (Aachen University) which has been pioneering the field of mobile

    NMR and has developed MOUSE, Volterra, Italy, 12-15 June 2013

    • Oana Miu

    Granted participation in the COST Action TD 1201 (COSCH) Training School on Automated 3D

    documentation of CH artefacts with robotized structured light system, Faculty of Mechatronics, Warsaw

    University of Technology, Warsaw, 25-27 November 2013

    Research stages within the Bilateral Collaboration Projects

    Hungarian Academy of Sciences (August 2013), University of Turin (October 2013 and April 2014)

    University of Turin (April 2014)

  • co

    lla

    ge

    Publications (2012-2014)

    • ISI articles: 5 published, 1 accepted, 2 submitted

    • Peer-reviewed articles : 14 (2 indexed journals; 1 open-access journal;

    2 indexed proceedings; 6 ISSN/ISBN conference proceedings; 3 conservation-

    restoration bulletins)

    • Books: 1

    • Patents applications: 1

    • Studies, 1; software, 2; informatic services, 1

    Dissemination

    Conference participation (2012-2014)

    • Invited lectures: 8

    • Oral presentations: 21

    • Poster presentations: 20

  • co

    lla

    ge

    Dissemination & Training Seminar and Workshop organisation

    1st International Seminar and Workshop Cultural Heritage–Emerging Technology

    and Innovation: Preservation of Parchment, Leather and Textiles, 24–26 Sep 2012, National Museum of Romanian History, Bucharest

    Media coverage

    TV: Omul si Timpul (6 oct 2012)

    Radio Cultural (25 sept 2012)

    Press release

    http://www.e-conservationline.com/content/view/1093

  • co

    lla

    ge

    Seminar and Workshop organisation

    2nd International Seminar and Workshop Cultural Heritage–Emerging Technology

    and Innovation: Preventive Conservation: Collections of Parchment, Leather and

    Textiles and Their Environment, 3–6 Sep 2013, The Academy Library, Bucharest

    Dissemination & Training

  • co

    lla

    ge

    Seminar and Workshop organisation

    3rd International Seminar and Workshop Cultural Heritage–Emerging Technology

    and Innovation: Advanced Technology for Diagnosis, Preservation and Management of Organic Cultural Heritage, 8 – 10 Oct 2014, CePCoR, ASTRA Museum, Sibiu

    You are warmly invited to attend this highly interesting event. Participation is free.

    Contact: elena.badea@unito it

    Dissemination & Training