1
|
Kumru HT, Gordin V, Cortes D. Predicting spatio-temporal radiofrequency ablation temperature using deep neural networks. Med Eng Phys 2024; 124:104089. [PMID: 38418015 DOI: 10.1016/j.medengphy.2023.104089] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/03/2022] [Revised: 12/07/2023] [Accepted: 12/10/2023] [Indexed: 03/01/2024]
Abstract
Radiofrequency ablation (RFA) of the medial branch nerve is a widely used therapeutic intervention for facet joint pain. However, denervation of the multifidus muscle is an inevitable consequence of RFA. New ablation techniques with the potential to prevent muscle denervation can be designed using computational simulations. However, depending on the complexity of the model, they could be computationally expensive. As an alternative approach, deep neural networks (DNNs) can be used to predict tissue temperature during RFA procedure. The objective of this paper is to predict the tissue spatial and temporal temperature distributions during RFA using DNNs. First, finite element (FE) models with a range of distances between the probes were run to obtain the temperature readings. The measured temperatures were then used to train the DNNs that predict the spatio-temporal temperature distribution within the tissue. Finally, a separate data obtained from FE simulations were used to test the efficacy of the network. The results presented in this paper demonstrate that the network can achieve an error rate as low as 0.05%, accompanied by a 92% reduction in time compared to FE simulations. The approach proposed in this study will play a major role in the design of new RFA treatments for facet joint pain.
Collapse
Affiliation(s)
- Hanife Tugba Kumru
- Department of Mechanical Engineering, The Pennsylvania State University, State College, PA, United States
| | - Vitaly Gordin
- Department of Anesthesia and Perioperative Medicine, Hershey Medical Center, Harrisburg, PA, United States
| | - Daniel Cortes
- Department of Mechanical Engineering, The Pennsylvania State University, State College, PA, United States.
| |
Collapse
|
2
|
Ross A, Leroux N, De Riz A, Marković D, Sanz-Hernández D, Trastoy J, Bortolotti P, Querlioz D, Martins L, Benetti L, Claro MS, Anacleto P, Schulman A, Taris T, Begueret JB, Saïghi S, Jenkins AS, Ferreira R, Vincent AF, Mizrahi FA, Grollier J. Multilayer spintronic neural networks with radiofrequency connections. NATURE NANOTECHNOLOGY 2023; 18:1273-1280. [PMID: 37500772 DOI: 10.1038/s41565-023-01452-w] [Citation(s) in RCA: 7] [Impact Index Per Article: 7.0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 11/10/2022] [Accepted: 06/12/2023] [Indexed: 07/29/2023]
Abstract
Spintronic nano-synapses and nano-neurons perform neural network operations with high accuracy thanks to their rich, reproducible and controllable magnetization dynamics. These dynamical nanodevices could transform artificial intelligence hardware, provided they implement state-of-the-art deep neural networks. However, there is today no scalable way to connect them in multilayers. Here we show that the flagship nano-components of spintronics, magnetic tunnel junctions, can be connected into multilayer neural networks where they implement both synapses and neurons thanks to their magnetization dynamics, and communicate by processing, transmitting and receiving radiofrequency signals. We build a hardware spintronic neural network composed of nine magnetic tunnel junctions connected in two layers, and show that it natively classifies nonlinearly separable radiofrequency inputs with an accuracy of 97.7%. Using physical simulations, we demonstrate that a large network of nanoscale junctions can achieve state-of-the-art identification of drones from their radiofrequency transmissions, without digitization and consuming only a few milliwatts, which constitutes a gain of several orders of magnitude in power consumption compared to currently used techniques. This study lays the foundation for deep, dynamical, spintronic neural networks.
Collapse
Affiliation(s)
- Andrew Ross
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Nathan Leroux
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Arnaud De Riz
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Danijela Marković
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Dédalo Sanz-Hernández
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Juan Trastoy
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Paolo Bortolotti
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France
| | - Damien Querlioz
- Centre de Nanosciences et de Nanotechnologies, Université Paris-Saclay, CNRS, Palaiseau, France
| | - Leandro Martins
- International Iberian Nanotechnology Laboratory (INL), Braga, Portugal
| | - Luana Benetti
- International Iberian Nanotechnology Laboratory (INL), Braga, Portugal
| | - Marcel S Claro
- International Iberian Nanotechnology Laboratory (INL), Braga, Portugal
| | - Pedro Anacleto
- International Iberian Nanotechnology Laboratory (INL), Braga, Portugal
| | | | - Thierry Taris
- Laboratoire de l'Intégration du Matériau au Système (IMS; UMR 5218), Univ. Bordeaux, CNRS, Bordeaux INP, Talence, France
| | - Jean-Baptiste Begueret
- Laboratoire de l'Intégration du Matériau au Système (IMS; UMR 5218), Univ. Bordeaux, CNRS, Bordeaux INP, Talence, France
| | - Sylvain Saïghi
- Laboratoire de l'Intégration du Matériau au Système (IMS; UMR 5218), Univ. Bordeaux, CNRS, Bordeaux INP, Talence, France
| | - Alex S Jenkins
- International Iberian Nanotechnology Laboratory (INL), Braga, Portugal
| | - Ricardo Ferreira
- International Iberian Nanotechnology Laboratory (INL), Braga, Portugal
| | - Adrien F Vincent
- Laboratoire de l'Intégration du Matériau au Système (IMS; UMR 5218), Univ. Bordeaux, CNRS, Bordeaux INP, Talence, France
| | - Frank Alice Mizrahi
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France.
| | - Julie Grollier
- Unité Mixte de Physique CNRS/Thales, CNRS, Thales, Université Paris-Saclay, Palaiseau, France.
| |
Collapse
|
3
|
Wang X, Li W, Zhang K, Sun J, Yang J, Zhang A, Xu L. A Novel Local Tumor Progression Prediction Method for Multimode Ablation Treatment. IEEE Trans Biomed Eng 2021; 69:1386-1397. [PMID: 34591754 DOI: 10.1109/tbme.2021.3116607] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/10/2022]
Abstract
OBJECTIVE The multimode ablation of liver cancer, which uses radio-frequency heating after a pre-freezing process to treat the tumor, has shown significantly improved therapeutic effects and enhanced anti-tumor immune response. Unlike open surgery, the ablated lesions remain in the body after treatment, so it is critical to assess the immediate outcome and to monitor disease status over time. Here we propose a novel tumor progression prediction method for simultaneous postoperative evaluation and prognosis analysis. METHODS We propose to leverage the intraoperative therapeutic information extracted from thermal dose distribution. For tumors with specific sensitivity reflected in medical images, different thermal doses implicitly indicate the degree of instant damage and long-term inhibition excited under specific ablation energy. We further propose a survival analysis framework for the multimode ablation treatment. It extracts carefully designed features from clinical, preoperative, intraoperative, and postoperative data, then uses random survival forest for feature selection and deep neural networks for survival prediction. RESULTS We evaluated the proposed methods using clinical data. The results show that our method outperforms the state-of-the-art survival analysis methods with a C-index of 0.8550.090. The thermal dose information contributes significantly to the prediction accuracy by taking up 21.7% of the overall feature importance. CONCLUSION The proposed methods have been demonstrated to be a powerful tool in tumor progression prediction of multimode ablation therapy. SIGNIFICANCE This kind of data-driven prognosis analysis may benefit personalized medicine and simplify the follow-up process.
Collapse
|