Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis
Abstract Medical imaging is an important tool for accurate medical diagnosis, while state-of-the-art image reconstruction algorithms raise critical challenges in massive data processing for high-speed and high-quality imaging. Here, we present a memristive image reconstructor (MIR) to greatly accele...
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-04-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-023-38021-7 |
| _version_ | 1797840976771809280 |
|---|---|
| author | Han Zhao Zhengwu Liu Jianshi Tang Bin Gao Qi Qin Jiaming Li Ying Zhou Peng Yao Yue Xi Yudeng Lin He Qian Huaqiang Wu |
| author_facet | Han Zhao Zhengwu Liu Jianshi Tang Bin Gao Qi Qin Jiaming Li Ying Zhou Peng Yao Yue Xi Yudeng Lin He Qian Huaqiang Wu |
| author_sort | Han Zhao |
| collection | DOAJ |
| description | Abstract Medical imaging is an important tool for accurate medical diagnosis, while state-of-the-art image reconstruction algorithms raise critical challenges in massive data processing for high-speed and high-quality imaging. Here, we present a memristive image reconstructor (MIR) to greatly accelerate image reconstruction with discrete Fourier transformation (DFT) by computing-in-memory (CIM) with memristor arrays. A high-accuracy quasi-analogue mapping (QAM) method and generic complex matrix transfer (CMT) scheme was proposed to improve the mapping precision and transfer efficiency, respectively. High-fidelity magnetic resonance imaging (MRI) and computed tomography (CT) image reconstructions were demonstrated, achieving software-equivalent qualities and DICE scores after segmentation with nnU-Net algorithm. Remarkably, our MIR exhibited 153× and 79× improvements in energy efficiency and normalized image reconstruction speed, respectively, compared to graphics processing unit (GPU). This work demonstrates MIR as a promising high-fidelity image reconstruction platform for future medical diagnosis, and also largely extends the application of memristor-based CIM beyond artificial neural networks. |
| first_indexed | 2024-04-09T16:23:27Z |
| format | Article |
| id | doaj.art-f60b54853cf34c23b0573ba8d7917bdc |
| institution | Directory Open Access Journal |
| issn | 2041-1723 |
| language | English |
| last_indexed | 2024-04-09T16:23:27Z |
| publishDate | 2023-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj.art-f60b54853cf34c23b0573ba8d7917bdc2023-04-23T11:21:59ZengNature PortfolioNature Communications2041-17232023-04-0114111010.1038/s41467-023-38021-7Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosisHan Zhao0Zhengwu Liu1Jianshi Tang2Bin Gao3Qi Qin4Jiaming Li5Ying Zhou6Peng Yao7Yue Xi8Yudeng Lin9He Qian10Huaqiang Wu11School of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversitySchool of Integrated Circuits, Beijing National Research Center for Information Science and Technology (BNRist), Tsinghua UniversityAbstract Medical imaging is an important tool for accurate medical diagnosis, while state-of-the-art image reconstruction algorithms raise critical challenges in massive data processing for high-speed and high-quality imaging. Here, we present a memristive image reconstructor (MIR) to greatly accelerate image reconstruction with discrete Fourier transformation (DFT) by computing-in-memory (CIM) with memristor arrays. A high-accuracy quasi-analogue mapping (QAM) method and generic complex matrix transfer (CMT) scheme was proposed to improve the mapping precision and transfer efficiency, respectively. High-fidelity magnetic resonance imaging (MRI) and computed tomography (CT) image reconstructions were demonstrated, achieving software-equivalent qualities and DICE scores after segmentation with nnU-Net algorithm. Remarkably, our MIR exhibited 153× and 79× improvements in energy efficiency and normalized image reconstruction speed, respectively, compared to graphics processing unit (GPU). This work demonstrates MIR as a promising high-fidelity image reconstruction platform for future medical diagnosis, and also largely extends the application of memristor-based CIM beyond artificial neural networks.https://doi.org/10.1038/s41467-023-38021-7 |
| spellingShingle | Han Zhao Zhengwu Liu Jianshi Tang Bin Gao Qi Qin Jiaming Li Ying Zhou Peng Yao Yue Xi Yudeng Lin He Qian Huaqiang Wu Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis Nature Communications |
| title | Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis |
| title_full | Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis |
| title_fullStr | Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis |
| title_full_unstemmed | Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis |
| title_short | Energy-efficient high-fidelity image reconstruction with memristor arrays for medical diagnosis |
| title_sort | energy efficient high fidelity image reconstruction with memristor arrays for medical diagnosis |
| url | https://doi.org/10.1038/s41467-023-38021-7 |
| work_keys_str_mv | AT hanzhao energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT zhengwuliu energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT jianshitang energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT bingao energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT qiqin energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT jiamingli energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT yingzhou energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT pengyao energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT yuexi energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT yudenglin energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT heqian energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis AT huaqiangwu energyefficienthighfidelityimagereconstructionwithmemristorarraysformedicaldiagnosis |