1
|
Hauptman A, Balasubramaniam GM, Arnon S. Machine Learning Diffuse Optical Tomography Using Extreme Gradient Boosting and Genetic Programming. Bioengineering (Basel) 2023; 10:bioengineering10030382. [PMID: 36978773 PMCID: PMC10045273 DOI: 10.3390/bioengineering10030382] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/20/2023] [Revised: 03/18/2023] [Accepted: 03/20/2023] [Indexed: 03/30/2023] Open
Abstract
Diffuse optical tomography (DOT) is a non-invasive method for detecting breast cancer; however, it struggles to produce high-quality images due to the complexity of scattered light and the limitations of traditional image reconstruction algorithms. These algorithms can be affected by boundary conditions and have a low imaging accuracy, a shallow imaging depth, a long computation time, and a high signal-to-noise ratio. However, machine learning can potentially improve the performance of DOT by being better equipped to solve inverse problems, perform regression, classify medical images, and reconstruct biomedical images. In this study, we utilized a machine learning model called "XGBoost" to detect tumors in inhomogeneous breasts and applied a post-processing technique based on genetic programming to improve accuracy. The proposed algorithm was tested using simulated DOT measurements from complex inhomogeneous breasts and evaluated using the cosine similarity metrics and root mean square error loss. The results showed that the use of XGBoost and genetic programming in DOT could lead to more accurate and non-invasive detection of tumors in inhomogeneous breasts compared to traditional methods, with the reconstructed breasts having an average cosine similarity of more than 0.97 ± 0.07 and average root mean square error of around 0.1270 ± 0.0031 compared to the ground truth.
Collapse
Affiliation(s)
- Ami Hauptman
- Department of Computer Science, Sapir Academic College, Sderot 7915600, Israel
| | - Ganesh M Balasubramaniam
- Department of Electrical and Computer Engineering, Ben-Gurion University of the Negev, Be'er Sheva 8441405, Israel
| | - Shlomi Arnon
- Department of Electrical and Computer Engineering, Ben-Gurion University of the Negev, Be'er Sheva 8441405, Israel
| |
Collapse
|
2
|
Salvadego D, Keramidas ME, Kölegård R, Brocca L, Lazzer S, Mavelli I, Rittweger J, Eiken O, Mekjavic IB, Grassi B. PlanHab * : hypoxia does not worsen the impairment of skeletal muscle oxidative function induced by bed rest alone. J Physiol 2018; 596:3341-3355. [PMID: 29665013 DOI: 10.1113/jp275605] [Citation(s) in RCA: 26] [Impact Index Per Article: 4.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/27/2017] [Accepted: 02/16/2018] [Indexed: 12/31/2022] Open
Abstract
KEY POINTS Superposition of hypoxia on 21 day bed rest did not worsen the impairment of skeletal muscle oxidative function induced by bed rest alone. A significant impairment of maximal oxidative performance was identified downstream of cardiovascular O2 delivery, involving both the intramuscular matching between O2 supply and utilization and mitochondrial respiration. These chronic adaptations appear to be relevant in terms of exposure to spaceflights and reduced gravity habitats (Moon or Mars), as characterized by low gravity and hypoxia, in patients with chronic diseases characterized by hypomobility/immobility and hypoxia, as well as in ageing. ABSTRACT Skeletal muscle oxidative function was evaluated in 11 healthy males (mean ± SD age 27 ± 5 years) prior to (baseline data collection, BDC) and following a 21 day horizontal bed rest (BR), carried out in normoxia ( PIO2 = 133 mmHg; N-BR) and hypoxia ( PIO2 = 90 mmHg; H-BR). H-BR was aimed at simulating reduced gravity habitats. The effects of a 21 day hypoxic ambulatory confinement ( PIO2 = 90 mmHg; H-AMB) were also assessed. Pulmonary O2 uptake ( V̇O2 ), vastus lateralis fractional O2 extraction (changes in deoxygenated haemoglobin + myoglobin concentration, Δ[deoxy(Hb + Mb)]; near-infrared spectroscopy) and femoral artery blood flow (ultrasound Doppler) were evaluated during incremental one-leg knee-extension exercise (reduced constraints to cardiovascular O2 delivery) carried out to voluntary exhaustion in a normoxic environment. Mitochondrial respiration was evaluated ex vivo by high-resolution respirometry in permeabilized vastus lateralis fibres. V̇O2peak decreased (P < 0.05) after N-BR (0.98 ± 0.13 L min-1 ) and H-BR (0.96 ± 0.17 L min-1 ) vs. BDC (1.05 ± 0.14 L min-1 ). In the presence of a decreased (by ∼6-8%) thigh muscle volume, V̇O2peak normalized per unit of muscle mass was not affected by both interventions. Δ[deoxy(Hb + Mb)]peak decreased (P < 0.05) after N-BR (65 ± 13% of limb ischaemia) and H-BR (62 ± 12%) vs. BDC (73 ± 13%). H-AMB did not alter V̇O2peak or Δ[deoxy(Hb + Mb)]peak . An overshoot of Δ[deoxy(Hb + Mb)] was evident during the first minute of unloaded exercise after N-BR and H-BR. Arterial blood flow to the lower limb during both unloaded and peak knee extension was not affected by any intervention. Maximal ADP-stimulated mitochondrial respiration decreased (P < 0.05) after all interventions vs. control. In 21 day N-BR, a significant impairment of oxidative metabolism occurred downstream of cardiovascular O2 delivery, affecting both mitochondrial respiration and presumably the intramuscular matching between O2 supply and utilization. Superposition of H on BR did not worsen the impairment induced by BR alone.
Collapse
Affiliation(s)
- Desy Salvadego
- Department of Medicine, University of Udine, Udine, Italy
| | - Michail E Keramidas
- Department of Environmental Physiology, Swedish Aerospace Physiology Centre, Royal Institute of Technology, Stockholm, Sweden
| | - Roger Kölegård
- Department of Environmental Physiology, Swedish Aerospace Physiology Centre, Royal Institute of Technology, Stockholm, Sweden
| | - Lorenza Brocca
- Department of Molecular Medicine, University of Pavia, Pavia, Italy
| | - Stefano Lazzer
- Department of Medicine, University of Udine, Udine, Italy
| | - Irene Mavelli
- Department of Medicine, University of Udine, Udine, Italy
| | - Jörn Rittweger
- Institute of Aerospace Medicine, German Aerospace Center, Cologne, Germany.,Department of Pediatrics and Adolescent Medicine, Medical Faculty, University of Cologne, Cologne, Germany
| | - Ola Eiken
- Department of Environmental Physiology, Swedish Aerospace Physiology Centre, Royal Institute of Technology, Stockholm, Sweden
| | - Igor B Mekjavic
- Department of Automation, Biocybernetics and Robotics, Jožef Stefan Institute, Ljubljana, Slovenia.,Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada
| | - Bruno Grassi
- Department of Medicine, University of Udine, Udine, Italy.,Institute of Bioimaging and Molecular Physiology, National Research Council, Milano, Italy
| |
Collapse
|