Machine Learning for Tomographic Imaging. Professor Ge Wang. Читать онлайн. Newlib. NEWLIB.NET

Автор: Professor Ge Wang
Издательство: Ingram
Серия:
Жанр произведения: Медицина
Год издания: 0
isbn: 9780750322164
Скачать книгу
rel="nofollow" href="#litres_trial_promo">References

       Part IV Others

       8 Modalities and integration

       8.1 Nuclear emission tomography

       8.1.1 Emission data models

       8.1.2 Network-based emission tomography

       8.2 Ultrasound imaging

       8.2.1 Ultrasound scans

       8.2.2 Network-based ultrasound imaging

       8.3 Optical imaging

       8.3.1 Interferometric and diffusive imaging

       8.3.2 Network-based optical imaging

       8.4 Integrated imaging

       8.5 Final remarks

       References

       9 Image quality assessment

       9.1 General measures

       9.1.1 Classical distances

       9.1.2 Structural similarity

       9.1.3 Information measures

       9.2 System-specific indices

       9.3 Task-specific performance

       9.4 Network-based observers*

       9.5 Final remarks*

       References

       10 Quantum computing*

       10.1 Wave–particle duality

       10.2 Quantum gates

       10.3 Quantum algorithms

       10.4 Quantum machine learning

       10.5 Final remarks

       References

       Appendices

       A Math and statistics basics

       B Hands-on networks

      IPEM–IOP Series in Physics and Engineering in Medicine and Biology

       Editorial Advisory Board Members

       Frank Verhaegen Maastro Clinic, the Netherlands

       Carmel Caruana University of Malta, Malta

       Penelope Allisy-Roberts formerly of BIPM, Sèvres, France

       Rory Cooper University of Pittsburgh, USA

       Alicia El Haj University of Birmingham, UK

       Ng Kwan Hoong University of Malaya, Malaysia

       John Hossack University of Virginia, USA

       Tingting Zhu University of Oxford, UK

       Dennis Schaart TU Delft, the Netherlands

       Indra J Das New York University, USA

      About the Series Series in Physics and Engineering in Medicine and Biology will allow IPEM to enhance its mission to ‘advance physics and engineering applied to medicine and biology for the public good.’

      Focusing on key areas including, but not limited to:

       clinical engineering

       diagnostic radiology

       informatics and computing

       magnetic resonance imaging

       nuclear medicine

       physiological measurement

       radiation protection

       radiotherapy

       rehabilitation engineering

       ultrasound and non-ionising radiation.

      A number of IPEM–IOP titles are published as part of the EUTEMPE Network Series for Medical Physics Experts.

      For my wife Ying Liu, whose support has been extraordinary, my academic mentors Michael Vannier et al, and all other family members and collaborators.

      —Ge Wang

      For my wife and daughter, who have been always fully supporting my academic career development.

      —Yi Zhang

      For my family.

      —Xiaojing Ye

      For my family.

      —Xuanqin Mou

      We are currently witnessing a revolution in science and engineering: in only a few years, machine learning has become the basis for almost every single algorithm development. I still remember a college course on non-linear systems in the early 1990s: except for a few zealots, my peers and I used to think of neural networks as an exotic and entirely impractical field of science. Little did we realize how wrong we were. Increases in computing power, access to large amounts of data, and the creativity of many scientists have entirely changed that view. Today, neural networks have become more pervasive than the Fourier transform.

      Some of the most popular applications of machine learning are in image analysis. You probably have some powerful deep learning networks in your pocket: just type ‘dog’ or ‘hat’ on your smart phone and it will instantaneously find all pictures containing your targets of interest. The application of machine learning may be less obvious in other areas, such as image generation. By now, machine learning has been at least considered for almost every imaginable algorithmic challenge. In the field of medical imaging, machine learning techniques