Advancing Microwave-Based Imaging Techniques for Medical Applications in the Wake of the 5G Revolution - Emerald

Page created by Susan Henderson
 
CONTINUE READING
Advancing Microwave-Based Imaging Techniques for Medical Applications in the Wake of the 5G Revolution - Emerald
Advancing Microwave-Based Imaging
 Techniques for Medical Applications
  in the Wake of the 5G Revolution
                   Jean-Charles Bolomey
                 University Paris Sud, France
                  jch.bolomey@gmail.com

                        Special Session
CS33 Horizon 2020 research and innovation session (EMERALD):
  ElectroMagnetic imaging for a novel genERation of medicAL
                           Devices
                (http://www.msca-emerald.eu/)
Advancing Microwave-Based Imaging Techniques for Medical Applications in the Wake of the 5G Revolution - Emerald
OUTLINE

• FEW QUESTIONS ON MICROWAVE-BASED IMAGING
  FOR MEDICAL APPLICATIONS
• PART I: WHERE ARE WE ARRIVED ?
• PART II: WHAT COULD BE TRIED/DONE ?
• CONCLUSIONS

                                   Special Session
            CS33 Horizon 2020 research and innovation session (EMERALD):
              ElectroMagnetic imaging for a novel genERation of medicAL
                                       Devices

J.Ch. Bolomey                   13th EuCAP, Krakow, March 2019             2
Advancing Microwave-Based Imaging Techniques for Medical Applications in the Wake of the 5G Revolution - Emerald
MICROWAVE-BASED MEDICAL IMAGING
              FACTS AND QUESTIONS …

                        •    SINCE EARLY 80’s… “MICROWAVES ARE
                             CLAIMED TO OFFER PROMISES AS IMAGING
                             MODALITY”
                             L.E. Larsen and J.H. Jacobi
                             (Diagnostic Imaging in Clinical Medicine, 11, 44-47,1982)

                        •    WHAT EXACTLY MICROWAVES ARE
                             PROMISING OF ? AND BY WHEN ?

                        •    INDEED, ALMOST 40 YEARS LATER,
                             MICROWAVES ARE STILL CLAIMING TO
                             OFFER PROMISES… BUT ONLY GAINED A
                             MODEST CLINICAL ACCEPTANCE

                        •    IN FACT, THE JOURNEY WAS MORE
                             COMPLICATED THAN EXPECTED BUT THE
                             RESULTS COLLECTED FROM RECENT
                             CLINICAL TRIALS SEEMS SUGGESTING A
                             POSSIBLE EXIT OF THE TUNNEL…

J.Ch. Bolomey      13th EuCAP, Krakow, March 2019                                  3
Advancing Microwave-Based Imaging Techniques for Medical Applications in the Wake of the 5G Revolution - Emerald
SIMPLIFIED CLASSIFICATION CHART
            OF MICROWAVE-BASED APPLICATIONS

                                           ACTIVE/PASSIVE
Industrial
                            SHORT RANGE                        LONG RANGE
Scientific
                          DIELECTOMETRY                           RADAR
 Medical

                                              I S M

                     COMMUNICATIONS                               TRANSFER/HARVESTING

                  MICROWAVE ISM APPLICATIONS AT THE CROSS-ROAD
                  OF SIGNALING/SENSING/POWERING TECHNOLOGIES

  J.Ch. Bolomey                       13th EuCAP, Krakow, March 2019                    4
Advancing Microwave-Based Imaging Techniques for Medical Applications in the Wake of the 5G Revolution - Emerald
VARIOUS ASPECTS OF MICROWAVES
                  FOR MEDICAL APPLICATIONS

             « Low »                                                                                        « High »
              Power                                       Medical                                            Power
                     [1]                                                                                           [2]
                                 DIAGNOSIS

                                                                                                  THERAPY
                                                BIO-                      HYPER-
                                              SENSING                    THERMIA

                                               IMAGING                  ABLATION

[1] M.F. Iskander and C.H. Durney, “Electromagnetic Techniques         [2] A.W. Guy, “History of Biological Effects and Medical Applications
for Medical Diagnosis: a Review”, Proc. IEEE, 68, 226-132, Jan. 1980   Of Microwave Energy”, IEEE Trans. MTT-32, 226-132,Sept. 1984

                               MICROWAVE-BASED MEDICAL IMAGING
                               IS THE “LAST COMER” ISM APPLICATION
J.Ch. Bolomey                                        13th EuCAP, Krakow, March 2019                                                            5
TIME LINE OF MAJOR IMAGING MODALITIES
        MICROWAVES AS “LAST COMER”
 1940             1950              1960              1970              1980             1990              2000               2010

                                                                           DIGITAL                   BREAST
                                  MAMMO-                CT                                                                      X-rays
RADIOGRAPHY                       GRAPHY             SCANNER            RADIOGRAPHY                   TOMO-
                                                                                                    SYNTHESIS                        [1]

SONAR                 NDT                           US                                                                               US
                                                    SCANNER                                                                           [2]

MAGNETIC                                                                   MRI                                                     MRI
RESONANCE                                                                  SCANNER                                                    [3]

RADIO-              NUCLEAR                  PET                                                               PET-CT              PET
NUCLEIDES           MEDECINE                 SCANNER                                                                                  [4]

                                                            1ST DIAG.                   EARLY IMAGING                              MW
                                                              APPL.                    INVESTIGATIONS                                 [5]
    RADAR                                      NDT          ORGAN MVTS.        BIOLOGICAL TARGETS       PATHOLOGY ORIENTED
                                                            LUNG DISEASE     ISOLATED ANIMAL ORGANS (BREAST, BRAIN, LUNG, BONES…)
                                                                 …           HUMAN IN-VIVO (FOREARM)
                                                                                                       HYPERTHERMIA CONTROL

    [1] R. Ciernak, “X-Ray Computed Tomography in Biomedical Engineering”, Springer-Verlag London Limited, 2011
    [2] J. Woo, “A short history of the Real-time ultrasound scanner”, http://www.ob-ultrasound.net/history-realtime.html
    [3] T. Geva, “Magnetic Resonance Imaging: Historical Perspective”, Journ. Cardiovascular Magnetic Resonance, 8, 573-580, 2006
    [4] G. Bernal, “History of PET scanners”, http://large.stanford.edu/courses/2014/ph241/bernal1/
    [5] J.Ch. Bolomey, “Crossed Viewpoints on Microwave-based Imaging for Medical Diagnosis: From Genesis to Earliest Clinical outcomes ”,
    in The World of Applied Electromagnetics, A. Lakhtakia, C.M. Furse (eds), Springer 2018

J.Ch. Bolomey                                     13th EuCAP, Krakow, March 2019                                                           6
GLOBAL VIEW OF MEDICAL IMAGING MARKET
                     WHAT PLACE FOR MICROWAVES ?
• SALIENT MARKET FEATURES:                                                              MEDICAL IMAGING
   –    Global Market size worth 29.48 billions and
                                                                 NUCLEAR   Advanced
                                                                           Modalities   EQUIPMENT MARKET
        projected to reach 46.18 by 2024,                                               (Transparency Market Research,
                                                                  MRI
   –    Market growth supported by the rise of geriatric                   49%
                                                                                        transparencymarketresearch.com)
        population and the increased prevalence of CV
        diseases,                                                 CT
   –    Expected Growth Rate: 6.56%, over the forecast
                                                                           Basic
        period 2016 - 2024                                        US       Modalities
                                                                                            GLOBAL MEDICAL
   –    X-ray first; fastest growth expected for CT/PET                                    IMAGING MARKET
        (earlier disease detection),                                       51%                (Pictures of the Future,
                                                                                        Medical Imaging:Facts and Forecasts,
   –    Larger growth expected for emerging markets.             X-RAY                             siemens.com)
   –    Global market shared by only a few big companies.

• KEY GROWTH DRIVERS:                                A PRIORI
   –    Miniaturization and portability, ✔         MICROWAVES                              Emerging
                                                                                           Markets         Developped
   –    Digitization of measured data, ✔           LOOK ALMOST                             36%             Markets
   –    Hybrid imaging systems, ✔                    PERFECT                                               64%
   –    Use of non-ionizing modalities, ✔
   –    Non-invasive and easy-to-use equipment, ✔
   –    Inexpensive, energy saving, ergonomic medical ✔                               BUT, A
        equipment with low maintenance requirements …                              POSTERIORI…
                                                                                  NOT SO SIMPLE !
  J.Ch. Bolomey                           13th EuCAP, Krakow, March 2019                                              7
sensitivity for breast cancer detection, especially for cases where the use of mammography
and ultrasound has been equivocal. However, MRI is almost always considered as an addition
to, rather than a replacement for, digital mammography in Europe today. Constraints of cost
and availability for MRI mean it is reserved only for high-risk patients; only in Austria and Italy
          MICROWAVES AS “CHALLENGER” TECH ?
is MRI more commonly offered for regular screening in addition to digital mammography. Given
MRI resource constraints, it is hard to see MRI realistically becoming a significant challenger to
digital mammography anytime soon.

                   Comparison of 'challenger' breast screening technologies
                                                                                             'Challenger'
                       Measure                           DBT                  ABUS     MRI
                                                                                                tech

            From: Stephen Holloway, AuntMinnie.com contributing writer March 6, 2019
            “DBT is well-positioned as successor in European breast screening”
            European Congress of Radiology, Vienna, Feb. 27 – March 3, 2019

   J.Ch. Bolomey                                    13th EuCAP, Krakow, March 2019                          8
5. Plateau of Productivity
Mainstream adoption starts to take off. Criter ia for assessing
    HYPE CYCLE FOR EMERGING TECHNOLOGIES
legitim ate provider s are m or e clearly defined. The technology’s
APPLICATION TO MICROWAVE-BASED MEDICAL IMAGING
broad m arket applications and relevance are under stood.

                                   NUMERICAL
                                   & PHANTOM
                                     MODELS
                                                                                                              ?
                                                                                                            CLINICAL
                                                                                                            ADOPTION AND
                                                                                                            DISSEMINATION
 EXPECTATION

                                                                                    CLINICAL                ▪ Acceptance
                                                                                    TRIALS                    threshold
                                                       FUNCTIONAL
                                                       PROTOTYPES
                                                                                ▪ prototype-to-product
                  PHANTOMS,                                                       conversion,
                   ISOLATED                                                     ▪ regulatory approval,
                    ORGANS                                                      ▪ commercialization
                 EXPERIMENTS                       ▪ instrument development,                             …all in the face
                                                   ▪ preclinical testing,
                                                   ▪ clinical prototyping                                of competing
                      RF                                                                                 technologies !
                  DOSIMETRY
                   TRIGGER

                                                                                  Slope of
                                                                                                             TIME
                                Peak of             Through
                                                MATURITY      of
               Innovation       Inflated          Disilliusonment              Enlightenment
                 Trigger      Expectations
Machines are becom ing better at under standing hum ans and the
               (Adapted from https://www.gartner.com/en/research/methodologies/gartner-hype-cycle)
environm ent — for exam ple, recognizing the em otion in a
person’s
 J.Ch. Bolomey
               voice — and hum ans  are becom ing better at
                                13th EuCAP, Krakow, March 2019                                                             9
DEVELOPMENT
          MICROWAVES           FLOW-CHART
                     FOR BIOMEDICAL APPLICATIONS
            FOR MICROWAVE-BASED      IMAGING
                 SIMPLIFIELD FLOW-CHART
 “INSIDE” MICROWAVE                                                                       Encouraging results…
     COMMUNITY
MICROWAVE EXPERTISE                               Good candidate…                           EARLY PROOF
(HARD/SOFTWARE)                                                                             OF CONCEPT
                                                   IDENTIFIED                             NUMERICAL SIMULATION
                &                                   CLINICAL                             ON-PHANTOM EXPERIMENT
 SUPPOSED SPECIFIC                                   TARGET
 ADVANTAGES OF
 MICROWAVES DUE TO
 LIVING TISSUE DIELECTRIC                                  TARGET                            PROTOTYPE
                                                                               PROTO
 PROPERTIES                                              ADJUSTMENT           REVISION         TEST CASES
   “Technology” pull                                                                          PRE-CLINICAL
                                                                                              ASSESSMENT

                            CLINICAL NEEDS

                                                                                 ?
                                                                        NO
      « MAJOR »                                                                                CLINICAL
                                EXISTING                          PERF/COST                     TRIALS
        HEALTH                 MODALITIES                          BENEFIT
        ISSUES                                                          YES                 INTER-COMPARISON
                                                                                             MULTI-MODALITY
                                                                                          LARGE SCALE VALIDATION
                                        “Market” push

                                         ACCESS TO                 CLINICAL
                                                                                           “Technology” transfer
 “OUTSIDE” MICROWAVE                      MARKET                   PRACTICE
     COMMUNITY
 J.Ch. Bolomey                         13th EuCAP, Krakow, March 2019                                         10
MICROWAVE SCANNERS FOR EARLY IN-VIVO IMAGING
                 SOME EXAMPLES…
                                  Carolinas Heart Inst (USA).         Keele Univ. (UK)
   INVERSE       F=2.45GHz                    F=1GHz                 F=1-2.3GHz
 SCATTERING      Nt=Nr=32                     N=32                   N=24                             Univ. of Calgary (CA)
 (TOMOGRAPHY)                                                                                    F=50MHz-15GHz
       T                                                                                         N=1

   RADAR
     MULTI-
  MU VIEW

  MO MONO-                                            whole-body     functional animal                    MO    breast imaging
       VIEW          T                         T   animal imaging             imaging       T                       case study

1990                     1995                   2000                  2005                   2010                     2015

 human forearm               NIT on animal                    breast imaging on volunteers and patients
                                                   F=0.3-3GHz               F=4-8GHz                                  F=2-4GHz
                                                   N=16                     N=60                                      N=16

F=2.3GHz     T           T                          T                                               MU                     MU
N=64
UPC Barcelona (SP)             Thayer School of Engng. Dartmouth (USA)            Univ. of Bristol (UK)        McGill Univ (CA).
J.Ch. Bolomey                                            COST MiMed, 20-22 June, 2016                                        11
MICROWAVE IMAGING SYSTEMS
                       ENGAGED IN (PRE)CLINICAL TRIALS
     MANY DATA ACQUISITION AND PROCESS
                     ACQUISITION TIME  MODALITIES
       NUMBER OF                                    25000
       CHANNELS
                                                              MARIA                                       MICRIMARIA (2006)
                                                             1.5-8 GHz                                    MARIA
           11
       MIST (1995)
                                              MIST
                                                                                                          Multi-static, 60 slot antennas
                                              0.5-3                                                       1/2 spherical              RADAR
       DARTMOUTH Coll.                        GHz
       Multi-static,16 monopoles                                      SUST
       Cylindrical, Variable height,                                4-8.5 GHz                              SUST-SHENZEN
                 TOMOGRAPHY                                                                                Multi-static, , 2 antennas
                                                                                                           5 rings, 360 positions/ring,
                                                                                                           mech./scan, Cylindrical RADAR
       Mc Gill Univ.
       Multi-static, 16 antennas
       Wearable,          RADAR
                                                                                                           TSAR UoCALGARY
       MEDFIELDS DIAG. (2005)                         MGU                                                  Monostatic, 2 vivaldi antennas
                                                      2-4                  TSAR                            Cylindrical/Conformal,
       Head adjustable, 8 slot antennas                             1.5-9 GHz                              mech./scan              RADAR
                        CLASSIFICATION                GHz

                     EM-Tensor (2012)
                     Cylindrical
                     5x32=160 slot antennas
                              TOMOGRAPHY                 FREQUENCY (GHz)
Adapted from: “Microwave Breast Imaging: Clinical Advances and Remaining Challenges”, D. O’Loughlin, M. O’Halloran, B.M. Moloney et al.;
IEEE Transactions on Biomedical Engineering ( Volume: 65 , Issue: 11 , Nov. 2018 )
     J.Ch. Bolomey                                  13th EuCAP, Krakow, March 2019                                                 12
WHAT HAS ALREADY EMERGED ?

  OUTCOMES                                      A FEW OPERATIONAL SYSTEMS
                                                ENGAGED IN CLINICAL TRIALS :
                                                * BREAST, BRAIN, BONES, LUNG
                                                * ENCOURAGING” SENSITIVITY
                                                * SPECIFICITY TO BE VALIDATED

                                            -   SIGNIFICANT ACHIEVEMENTS
                                                * EXTENSIVE NUMERICAL AND
                                                  EXPERIMENTAL MODELING,
                                                * RECONSTRUCTION ALGORITHMS
                                                  (TOMOGRAPHY, RADAR, HOLOGRAPHY),
                                                * TISSUE CHARACTERIZATION,
                                                * PROTOTYPE DEVELOPMENT,
                                                * PHANTOM DEVELOPMENT/EXPERIMENT,
                                                *ETC…

                                        -   BUT…
                                            SOME PERSISTENTLY “OPEN” ISSUES:
                                            * LACK OF SYNTHESIS AND BENCHMARKING
                                            * A PRIORI INFORMATION
                                            * DIELECTRIC CHARACTERIZATION
                                            * PATIENT INTERFACE OPTIMIZATION
  EFFORTS                                   * IMAGE RELEVANCE (SPECIFICITY)

                           WHAT COULD BE TRIED / DONE ?
                LOOK AT EMERAD PROGRAM… AND THREE SUGGESTIONS
J.Ch. Bolomey               13th EuCAP, Krakow, March 2019                      13
SUGGESTION 1

      “BETTER” DATA ACQUISITION TECHNIQUES

  •    EXPLOITING ALREADY EXISTING OR SUPPOSEDLY RAPIDLY
       AVAILABLE MICROWAVE TECHNOLOGIES FOR OBTAINING “BETTER”
       DATA ACQUISITION TECHNIQUES:

         – INTRODUCING THE BEST OF WIRELESS TECHNOLOGIES USED FOR
           EXISTING COMMUNICATION, RADAR, SENSOR GRID SYSTEMS
           (MODULATION/DEMODULATION SCHEMES, OFDN, MIMO PROCESSING,…)

         – APPLICATION-DEDICATED AND/OR PATIENT ADAPTIVE OPTIMIZATION OF
           THE PROBE ARRAY

         – INCREASING DYNAMIC RANGE TO REDUCE MULTI-PATH CORRUPTION

         – TAKING PROFIT OF 5G, IoT OR IoE DEVELOPMENTS
           (MINIATURIZATION, ARRAYS OF INTEGRATED PROBES, …)

         – REDUCING FABRICATION/EXPLOITATION/MAINTENANCE COSTS BY
           MEANS OF APPROPRIATE ARCHITECTURES

J.Ch. Bolomey                13th EuCAP, Krakow, March 2019           14
MICROWAVES MM-WAVES SPECTRUM
MEDICAL IMAGING VS 5G FREQUENCY BANDS
                                      VISIBLE                      (Source: Qualcomm)

                                                                   FREQUENCY (GHz)

0.6 GHz            6 GHz     24 GHz                                     100 GHz

       BRAIN    TOMOGRAPHY

                                                              MEDICAL
                 BREAST      UWB RADAR                        IMAGING

J.Ch. Bolomey                13th EuCAP, Krakow, March 2019                       15
SUGGESTION 2
                    “BETTER” DATA PROCESSING
                        AND ALGORITHMS

       • EXPLOITING/ANTICIPATING GROWING COMPUTING POWER
                • RADAR: TOWARD REAL-TIME, FAST PRE/POST PROCESSING
                • INVERSE-SCATTERING IMAGING: DECREASING MODEL NOISE, FULL
                  3D ACCURATE MODELING (INCLUDING INTERACTIONS AND MUTUAL
                  COUPLING…)
                • INCREASING AMOUNT OF DATA
                • IMPROVING SPATIAL RESOLUTION
                • QUANTUM COMPUTERS ? BEYOND MOORE’s LAW ?

       – USING OTHER IMAGING MODALITIES
                • A PRIORI INFORMATION (WHEN NEEDED)
                • MICROWAVE DATA FUSION WITH OTHER MODALITIES FOR AI/ML/DL
                  ALGORITHMS

       – CONSIDERING “NON IMAGING-BASED” MICROWAVE SENSING
         PROTOCOLS (SENSOR NETWORKS GRIDS, ARTIFICIAL/AUGMENTED
         INTELLIGENCE, MACHINE LEARNING, CLASSIFICATION, NEURAL
         NETWORKS, ETC.)

J.Ch. Bolomey                    13th EuCAP, Krakow, March 2019              16
MOORE’s LAW EXTRAPOLATIONS
                 OF COMPUTATIONAL POWER

J.Ch. Bolomey           13th EuCAP, Krakow, March 2019   17
SUGGESTION 3
         INCREASING INTERACTIONS WITH BIOLOGISTS
            AND “END-USER” MEDICAL COMMUNITY
•    RETURNING TO BASIC MW/BIO INVESTIGATIONS
       ▪ TISSUE DIELECTRIC CHARACTERIZATION (LOCAL PROBE, MRI,
         SCANNING MICROSCOPY,…)
       ▪ PHYSICAL OR HEURISTIC MODELS FOR FREQUENCY DEPENDENCE
       ▪ FOCUSING ON DIELECTRIC SPECIFICITY AND SENSITIVITY
         ASSESSMENTS
       ▪ QUANTIFYING DIELECTRIC IMAGES IN TERMS OF BIO-PHYSIOLOGICAL
         FACTORS
•     INCREASING INTERACTIONS WITH “END USER” MEDICAL
      COMMUNITY
       ▪ IDENTIFYING REAL NEEDS AND LOOKING FOR THE MOST
         MICROWAVE-FRIENDLY AND CLINICALLY RELEVANT SCENARIOS
       ▪ CASE BY CASE VALIDATION THANKS TO LARGE SCALE /MULTIMODALITY
         CLINICAL ASSESSMENT CAMPAIGNS
       ▪ GUIDING TECHNICAL ADVANCES BY CLINICAL RETURNS
       ▪ HYBRIDIZING WITH OTHER MODALITIES
       ▪ FOLLOWING UP IMPROVEMENT OF OTHER IMAGING MODALITIES:
         MAN-PORTABLE MRI, DIGITAL TOMOSYNTHESIS, LOW-DOSE PHASE
         CONTRAST X-RAYS, ETC.
       ▪ NOT UNDERESTIMATING THE CHANGES IN THE MEDICAL PRACTICE
         INDUCED BY AI / ML DEVELOPMENTS
       ▪ ATTENDING MEDICAL MEETINGS, LOOKING AT MEDICAL JOURNALS, ETC.
J.Ch. Bolomey               13th EuCAP, Krakow, March 2019             18
FUTURE MEDICAL IMAGING FACING REVOLUTIONARY
           AI AND ML CHALLENGES (EXAMPLES)
▪ “Look Ahead: The Future of Medical Imaging”∑
  James H. Thrall, radiology imaging, August 1, 2015
▪ “How Artificial Intelligence Will Change Medical Imaging”
  D. Fornel, artificial intelligence, February 2017
▪ “What is the Future of Medical Imaging Equipment ?”
  M. Taschetta-millane, radiology imaging, July 2018
▪ “Technologies to Watch in Breast Imaging”
  J. Zagoudis, radioimaging, July 2018
▪ “Artificial Intelligence in Radiology: Hype or Hope ?”
  M.B. Massat, pp. xx-yy, March, 2018
▪ “A Promising future for AI in breast cancer screening”
  M. B. Massat, pp. 22-25, September 2018
                                            ∑
▪ “Demystifying Artificial Intelligence”
   Part I, “Simplifying AI and Machine Learning”, E. Siegel, pp. 26-28, May, 2018
   Part II, “An Imaging Tool Ready to Explode”, L.N. Tanenbaum,pp. Search26-27,
                                                                          the site... July 2018
   Part III, “Going Beyond Escape”, R.B. Shrestha, pp. 8-11, November 2018
▪ “Machine (Deep) Learning Methods for Image Processing                 IEEE Trans. Radiation and
                                                                         Plasma Medical Sciences
  and Radiomics”, M. H. Latim, C. Parmar, J. QiI, I. el Naqa
  Vol. 3,pp.104-107, March 2019             ∑
   SPECIAL ISSUE ON MACHINE LEARNING FOR IMAGE PROCESSING AND RADIOMICS

   J.Ch. Bolomey                     13th EuCAP, Krakow, March 2019                          19
CS33 Horizon 2020 research and innovation session (EMERALD):
        ElectroMagnetic imaging for a novel genERation of medicAL Devices

                            CONCLUSIONS
• AFTER A LONG MATURATION PERIOD, MICROWAVE-BASED
  IMAGING SYSTEMS ARE BEGINNING TO PROVIDE CLINICAL
  RESULTS

• FUTURE DEVELOPMENTS MUST:
      – TAKING PROFIT OF 5G TECHNOLOGY AND DEEP LEARNING ALGORITHMS
      – ACCOUNTING FOR THE EVOLUTION OF THE MEDICAL PRACTICE, MORE
        PARTICULARLY IMAGING MODALITIESUNDER THE IMPACT OF ARTIFICIAL
        INTELLIGENCE

• DUE TO THEIR NECESSARY SPECIFICITY, NEW SYSTEMS
  MUST BE FOCUSED ON REAL CLINICAL NEEDS FOR:
      – COMPLETING/SUPPLEMENTING OTHER IMAGING MODALITIES
      – ADDRESSING SPECIFIC REQUIREMENTS FOR DEVELOPPING COUNTRIES
• A

J.Ch. Bolomey                  13th EuCAP, Krakow, March 2019               20
You can also read