Suppressing motion artefacts in MRI using an Inception-ResNet network with motion simulation augmentation

The suppression of motion artefacts from MR images is a challenging task. The purpose of this paper was to develop a standalone novel technique to suppress motion artefacts in MR images using a data-driven deep learning approach. A simulation framework was developed to generate motion-corrupted images from motion-free images using randomly generated motion profiles. An Inception-ResNet deep learning network architecture was used as the encoder and was augmented with a stack of convolution and upsampling layers to form an encoder-decoder network. The network was trained on simulated motion-corrupted images to identify and suppress those artefacts attributable to motion. The network was validated on unseen simulated datasets and real-world experimental motion-corrupted in vivo brain datasets. The trained network was able to suppress the motion artefacts in the reconstructed images, and the mean structural similarity (SSIM) increased from 0.9058 to 0.9338. The network was also able to suppress the motion artefacts from the real-world experimental dataset, and the mean SSIM increased from 0.8671 to 0.9145. The motion correction of the experimental datasets demonstrated the effectiveness of the motion simulation generation process. The proposed method successfully removed motion artefacts and outperformed an iterative entropy minimization method in terms of the SSIM index and normalized root mean squared error, which were 5–10% better for the proposed method. In conclusion, a novel, data-driven motion correction technique has been developed that can suppress motion artefacts from motion-corrupted MR images. The proposed technique is a standalone, post-processing method that does not interfere with data acquisition or reconstruction parameters, thus making it suitable for routine clinical practice.     

Rapid motion correction in MR‐guided high‐intensity focused ultrasound heating using real‐time ultrasound echo information

The objective of this study was to evaluate the feasibility of integrating real‐time ultrasound echo guidance in MR‐guided high‐intensity focused ultrasound (HIFU) heating of mobile targets in order to reduce latency between displacement analysis and HIFU treatment. Experiments on a moving phantom were carried out with MRI‐guided HIFU during continuous one‐dimensional ultrasound echo detection using separate HIFU and ultrasound imaging transducers. Excellent correspondence was found between MR‐ and ultrasound‐detected displacements. Real‐time ultrasound echo‐based target tracking during MR‐guided HIFU heating is shown with the dimensions of the heated area similar to those obtained for a static target. This work demonstrates that the combination of the two modalities opens up perspectives for motion correction in MRI‐guided HIFU with negligible latency. Copyright © 2010 John Wiley & Sons, Ltd.     

Single‐shot single‐voxel lactate measurements using FOCI‐LASER and a multiple‐quantum filter

Measurement of tissue lactate using ¹H MRS is often confounded by overlap with intense lipid signals at 1.3 ppm. Single‐voxel localization using PRESS is also compromised by the large chemical shift displacement between voxels for the 4.1 ppm (–CH) resonance and the 1.3 ppm –CH₃ resonance, leading to subvoxels with signals of opposite phase and hence partial signal cancellation. To reduce the chemical shift displacement to negligible proportions, a modified semi‐LASER sequence was written (“FOCI‐LASER”, abbreviated as fLASER) using FOCI pulses to permit high RF bandwidth even with the limited RF amplitude characteristic of clinical MRI scanners. A further modification, MQF‐fLASER, includes a selective multiple‐quantum filter to detect lactate and reject lipid signals. The sequences were implemented on a Philips 3 T Achieva TX system. In a solution of brain metabolites fLASER lactate signals were 2.7 times those of PRESS. MQF‐fLASER lactate was 47% of fLASER (the theoretical maximum is 50%) but still larger than PRESS lactate. In oil, the main 1.3 ppm lipid peak was suppressed to less than 1%. Enhanced suppression was possible using increased gradient durations. The minimum detectable lactate concentration was approximately 0.5 mM. Coherence selection gradients needed to be at the magic angle to avoid large water signals derived from intermolecular multiple‐quantum coherences. In pilot patient measurements, lactate peaks were often observed in brain tumours, but not in cervix tumours; lipids were effectively suppressed. In summary, compared with PRESS, the fLASER sequence yields greatly superior sensitivity for direct detection of lactate (and equivalent sensitivity for other metabolites), while the single‐voxel single‐shot MQF‐fLASER sequence surpasses PRESS for lactate detection while eliminating substantial signals from lipids. This sequence will increase the potential for in vivo lactate measurement as a biomarker in targeted anti‐cancer treatments as well as in measurements of tissue hypoxia. Copyright © 2015 John Wiley & Sons, Ltd.     

Real-time motion and B0 correction for localized adiabatic selective refocusing (LASER) MRSI using echo planar imaging volumetric navigators

A method is presented to correct the effects of motion and motion‐related B₀ perturbations on spectroscopic imaging in real time through the use of a volumetric navigator. It is demonstrated that, for an axial slice, lifting the chin significantly disrupts the B₀ homogeneity in the zero‐order (frequency), first‐order Y (coronal) axis and second‐order ZY term. This volumetric navigator is able to measure and correct in real time both head pose and zero‐ to first‐order B₀ inhomogeneities. The volumetric navigator was validated in six volunteers who deliberately lifted and then dropped their chin during the scan. These scans show that motion correction alone is not sufficient to recover the spectral quality. By applying real‐time shim adjustments, spectral quality was fully recovered to linewidths below 0.08 ppm and the signal‐to‐noise ratio to within acceptable limits in five of six subjects. In the sixth subject, 83% of the spectra within the volume of interest were recovered, compared with the worst case nonshim‐corrected scan, where none of the voxels fell within these quality bounds. It is shown that the use of a volumetric navigator comes at no additional cost to the scan time or spectral signal‐to‐noise ratio. Copyright © 2011 John Wiley & Sons, Ltd.     

Motion correction and frequency stabilization for MRS of the human spinal cord

Subject motion is challenging for MRS, because it can falsify results. For spinal cord MRS in particular, subject movement is critical, since even a small movement > 1 mm) can lead to a voxel shift out of the desired measurement region. Therefore, the identification of motion corrupted MRS scans is essential. In this investigation, MR navigators acquired simultaneously with the MRS data are used to identify a displacement of the spinal cord due to subject motion. It is shown that navigators are able to recognize substantial subject motion (>1 mm) without impairing the MRS measurement. In addition, navigators are easy to apply to the measurement, because no additional hardware and just a minor additional user effort are needed. Moreover, no additional scan time is required, because navigators can be applied in the deadtime of the MRS sequence. Furthermore, in this work, retrospective motion correction combined with frequency stabilization is presented by combining navigators with non‐water‐suppressed ¹H‐MRS, resulting in an improved spectral quality of the spinal cord measurements. Copyright © 2016 John Wiley & Sons, Ltd.     

Quantitative three‐dimensional ultrashort echo time cones imaging of the knee joint with motion correction

Knee degeneration involves all the major tissues in the joint. However, conventional MRI sequences can only detect signals from long T2 tissues such as the superficial cartilage, with little signal from the deep cartilage, menisci, ligaments, tendons and bone. It is highly desirable to develop new sequences that can detect signal from all major tissues in the knee. We aimed to develop a comprehensive quantitative three‐dimensional ultrashort echo time (3D UTE) cones imaging protocol for a truly “whole joint” evaluation of knee degeneration. The protocol included 3D UTE cones actual flip angle imaging (3D UTE‐Cones‐AFI) for T₁ mapping, multiecho UTE‐Cones with fat suppression for T₂* mapping, UTE‐Cones with adiabatic T_1ρ (AdiabT_1ρ) preparation for AdiabT_1ρ mapping, and UTE‐Cones magnetization transfer (UTE‐Cones‐MT) for MT ratio (MTR) and modeling of macromolecular proton fraction (f). An elastix registration technique was used to compensate for motion during scans. Quantitative data analyses were performed on the registered data. Three knee specimens and 15 volunteers were evaluated at 3 T. The elastix motion correction algorithm worked well in correcting motion artifacts associated with relatively long scan times. Much improved curve fitting was achieved for all UTE‐Cones biomarkers with greatly reduced root mean square errors. The averaged T₁, T₂*, AdiabT_1ρ, MTR and f for knee joint tissues of 15 healthy volunteers were reported. The 3D UTE‐Cones quantitative imaging techniques (ie, T₁, T₂*, AdiabT_1ρ, MTR and MT modeling) together with elastix motion correction provide robust volumetric measurement of relaxation times, MTR and f of both short and long T₂ tissues in the knee joint.     

Volumetric navigated MEGA‐SPECIAL for real‐time motion and shim corrected GABA editing

Mescher–Garwood (MEGA) editing with spin echo full intensity acquired localization (MEGA‐SPECIAL, MSpc) is a technique to acquire γ‐aminobutyric acid (GABA) without macromolecule (MM) contamination at a T_E of 68 ms. However, due to the requirement of multiple shot‐localization, it is often susceptible to subject motion and B₀ inhomogeneity. A method is presented for real‐time shim and motion correction (ShMoCo) using volumetric navigators to correct for motion and motion‐related B₀ inhomogeneity during MSpc acquisition. A phantom experiment demonstrates that ShMoCo restores the GABA peak and improves spectral quality in the presence of motion and zero‐ and first‐order shim changes. The ShMoCo scans were validated in three subjects who performed up–down and left–right head rotations. Qualitative assessment of these scans indicates effective reduction of subtraction artefacts and well edited GABA peaks, while quantitative analysis indicates superior fitting and spectral quality relative to scans with no correction. Copyright © 2015 John Wiley & Sons, Ltd.     

Two‐voxel spectroscopy with dynamic B0 shimming and flip angle adjustment at 7 T in the human motor cortex

The aim of this study was to acquire high‐quality in vivo ¹H spectra concurrently from two voxels at ultra‐high field (7 T) without specialized hardware. To this end, an acquisition scheme was developed in which first‐order shims and flip angles are dynamically updated to acquire spectra from both of the brain's motor cortices in an alternating fashion. To validate this acquisition scheme, separate, static, single‐voxel acquisitions were also performed for comparison. Six subjects were examined using semi‐LASER spectroscopy at 7 T. Barium titanate pads were used to increase the extent of the effective transmit field (B₁⁺). Spectra were obtained from the hand area of both motor cortices for both acquisition schemes. LCModel was used to determine neurochemical profiles in order to examine variations between acquisition schemes and volumes of interest. The dynamic two‐voxel acquisition protocol produced water linewidths (full width at half‐maximum between 11.6 and 12.8 Hz) and signal‐to‐noise ratios similar to those from static single‐voxel measurements. The concentrations of 13 individual and 3 combined metabolites with Cramér–Rao lower bounds below 30% were reliably detected for both acquisition schemes, and agreed well with previous postmortem assay and spectroscopy studies. The results show that high spectral quality from two voxels can be acquired concurrently without specialized hardware. This practical technique can be applied to many neuroscience applications. Copyright © 2015 John Wiley & Sons, Ltd.     

Compressed sensing with signal averaging for improved sensitivity and motion artifact reduction in fluorine‐19 MRI

Fluorine‐19 (¹⁹F) MRI of injected perfluorocarbon emulsions (PFCs) allows for the non‐invasive quantification of inflammation and cell tracking, but suffers from a low signal‐to‐noise ratio and extended scan time. To address this limitation, we tested the hypotheses that a ¹⁹F MRI pulse sequence that combines a specific undersampling regime with signal averaging has both increased sensitivity and robustness against motion artifacts compared with a non‐averaged fully sampled pulse sequence, when both datasets are reconstructed with compressed sensing. As a proof of principle, numerical simulations and phantom experiments were performed on selected variable ranges to characterize the point spread function of undersampling patterns, as well as the vulnerability to noise of undersampling and reconstruction parameters with paired numbers of x signal averages and acceleration factor x (NAx ‐AFx ). The numerical simulations demonstrated that a probability density function that uses 25% of the samples to fully sample the k‐space central area allowed for an optimal balance between limited blurring and artifact incoherence. At all investigated noise levels, the Dice similarity coefficient (DSC) strongly depended on the regularization parameters and acceleration factor. In phantoms, the motion robustness of an NA8‐AF8 undersampling pattern versus NA1‐AF1 was evaluated with simulated and real motion patterns. Differences were assessed with the DSC, which was consistently higher for the NA8‐AF8 compared with the NA1‐AF1 strategy, for both simulated and real cyclic motion patterns (P < 0.001). Both strategies were validated in vivo in mice (n = 2) injected with perfluoropolyether. Here, the images displayed a sharper delineation of the liver with the NA8‐AF8 strategy than with the NA1‐AF1 strategy. In conclusion, we validated the hypotheses that in ¹⁹F MRI the combination of undersampling and averaging improves both the sensitivity and the robustness against motion artifacts.     

Combined DCE‐MRI and single‐voxel 2D MRS for differentiation between benign and malignant breast lesions

Dynamic contrast‐enhanced (DCE) magnetic resonance imaging (MRI) and proton (1H) magnetic resonance spectroscopy (MRS) provide structural and biochemical information, including vascular volume, vascular permeability and tissue metabolism. In this study, we performed analysis of the enhancement characteristic from DCE‐MRI and the biochemical information provided by two‐dimensional (2D) Localized Correlated Spectroscopy (L‐COSY) MRS to determine the sensitivity and specificity of using DCE‐MRI alone compared to the combination with 2D MRS. The metabolite ratios from the 2D MRS spectra were analyzed using multivariate statistical analyses to determine a method capable of automatic separation of the patient cohort into malignant and benign lesions. A total of 24 lesions were studied with 21 diagnosed accurately using the enhancement characteristics alone resulting in sensitivity and specificity of 100% and 73%, respectively. Analysis of the 2D MRS data demonstrated a significant difference (p < 0.05) in 12 of 18 metabolite ratios analyzed for malignant compared to benign lesions. Previous research focused on utilizing the choline signal to noise ratio (SNR) as a marker for malignancy has been verified using 2D MRS in this study. Using Fisher's linear discriminant test using water (WAT)/olefinic fat diagonal (UFD), choline (CHO)/fat (FAT), CHO/UFD, and FAT/methyl fat (FMETD) as predictors the sensitivity and specificity increased to 92% and 100%, respectively. Using the Classification and Regression Tree (CART) statistical analysis the resulting sensitivity and specificity were 100% and 91%, respectively, with the most accurate predictor for differentiating malignant and benign determined to be FAT/FMETD. The cases within the study that presented a indeterminate diagnosis using DCE‐MRI alone were able to be accurately diagnosed when the metabolic information from 2D MRS was incorporated. The results suggest improved breast cancer detection through the combination of morphological and enhancement information from DCE‐MRI and metabolic information from 2D MRS. Copyright © 2010 John Wiley & Sons, Ltd.     

An evaluation of motion compensation strategies and repeatability for abdominal (1)H MR spectroscopy measurements in volunteer studies and clinical trials

Increased expression of choline kinase has frequently been shown in tumours and is thought to be associated with disease progression. Studies using magnetic resonance spectroscopy have shown an increase in total choline-containing metabolites (tCho) in tumour compared with healthy tissue. Subsequent reductions in tCho following successful treatment support the use of tCho as a biomarker of disease and response. However, accurate measurement of tCho using MRS in abdominal tumours is complicated by respiratory motion, blurring the acquisition volume and degrading the lineshape and signal-to-noise ratio (SNR) of metabolites. Motion compensation using prospectively gated acquisitions or offline correction of phase and frequency distortions can help restore the SNR and linewidth of metabolites. Prospectively gated acquisitions have the advantage of confining the volume of acquisition to the prescribed volume but are constrained by the repetition time (TR) of the respiratory motion. In contrast, data acquired for offline correction may use a shorter repetition time and therefore yield an increased SNR per unit time. In this study abdominal spectra acquired from single-voxel 'free-breathing' measurements in liver of healthy volunteers and in abdominal tumours of cancer patients were compared with those of prospective gating and with an implementation of offline correction. The two motion compensation methodologies were assessed in terms of SNR, linewidth and repeatability. Our experiments show that prospective gating and offline correction result in a 12-22% reduction in median tCho linewidth, while offline correction also provides a significant increase in SNR. The repeatability coefficient (the expected interval for 95% of repeat measurements) for tCho/water ratio was reduced by 37% (prospective gating) and 41% (offline correction). Both methods of motion compensation substantially improved the reproducibility of the tCho/water measurement and the tCho linewidth. While offline correction also leads to a significant improvement in SNR, it may suffer more from out-of-voxel contamination.     

A semi‐LASER,single‐voxel spectroscopic sequence with a minimal echo time of 20.1 ms in the human brain at 3 T

An optimized semi‐LASER sequence that is capable of acquiring artefact‐free data with an echo time (TE) of 20.1 ms on a standard clinical 3 T MR system was developed. Simulations were performed to determine the optimal TEs that minimize the expected Cramér‐Rao lower bound (CRLB) as proxy for quantification accuracy of metabolites. Optimized RF pulses, crusher gradients and phase cycling were used to achieve the shortest TE in a semi‐LASER sequence to date on a clinical system. Synthetic spectra were simulated using the density matrix formalism for TEs spanning from 20.1 to 220.1 ms. These simulations were used to calculate the expected CRLB for each of the 18 metabolites typically considered in ¹H MRS. High quality spectra were obtained in six healthy volunteers in the prefrontal cortex, which is known for spurious echoes due to its proximity to the paranasal sinuses, and in the parietal‐occipital cortex. Spectral transients were sufficient in quality to enable phase and frequency alignment prior to summation over all repetitions. Automated high‐quality water suppression was obtained for all voxels without manual adjustment. The shortest TE minimized the CRLB for all brain metabolites except glycine due to its overlap with myo‐inositol at this TE. It is also demonstrated that the CRLBs increase rapidly with TE for certain coupled metabolites.     

Noise correlations and SNR in phased‐array MRS

The acquisition of magnetic resonance spectroscopy (MRS) signals by multiple receiver coils can improve the signal‐to‐noise ratio (SNR) or alternatively can reduce the scan time maintaining a reliable SNR. However, using phased array coils in MRS studies requires efficient data processing and data combination techniques in order to exploit the sensitivity improvement of the phased array coil acquisition method. This paper describes a novel method for the combination of MRS signals acquired by phased array coils, even in presence of correlated noise between the acquisition channels. In fact, although it has been shown that electric and magnetic coupling mechanisms produce correlated noise in the coils, previous algorithms developed for MRS data combination have ignored this effect. The proposed approach takes advantage of a noise decorrelation stage to maximize the SNR of the combined spectra. In particular Principal Component Analysis (PCA) was exploited to project the acquired spectra in a subspace where the noise vectors are orthogonal. In this subspace the SNR weighting method will provide the optimal overall SNR. Performance evaluation of the proposed method is carried out on simulated ¹H‐MRS signals and experimental results are obtained on phantom ¹H‐MR spectra using a commercially available 8‐element phased array coil. Noise correlations between elements were generally low due to the optimal coil design, leading to a fair SNR gain (about 0.5%) in the center of the field of view (FOV). A greater SNR improvement was found in the peripheral FOV regions. Copyright © 2009 John Wiley & Sons, Ltd.     

High‐precision calibration of MRS thermometry using validated temperature standards: effects of ionic strength and protein content on the calibration

Currently, there is very limited ability to measure the temperature of the brain, but a direct technique for its estimation in vivo could improve the detection of patients at risk of temperature‐related brain damage, help in the diagnosis of stroke and tumour, and provide useful information on the mechanisms of thermoregulation of the brain. In this article, new calibrations in vitro of MRS thermometry using temperature‐stabilised reference phantoms are reported. The phantoms comprise two concentric glass spheres: the inner sphere contains the phantom material to be measured by MRS, and the outer sphere contains a substance with a known temperature stable to within 0.2 °C. The substances were freezing organic fixed‐point compounds (diphenyl ether and ethylene carbonate, freezing at 26.3 and 35.8 °C, respectively) or temperature‐controlled circulating water. The phantom temperature was continuously monitored with a fluoroptic probe calibrated at the National Physical Laboratory with traceability to the International Temperature Scale 1990 (ITS‐90). The MRS temperature calibration was obtained by measuring the chemical shift of water relative to N‐acetylaspartate (NAA) in a single voxel as a function of temperature using a 1.5‐T Philips Intera scanner. Measurements were made for several phantom materials to assess the effect of tissue composition on the water–NAA chemical shift against temperature calibration. The phantom mixtures contained 25 mm of NAA buffered to pH 6.5 or 7.5 and several ionic salts or bovine serum albumin (BSA). Spectra were acquired from 25 to 45 °C. The correlation between frequency differences and phantom temperature was very linear with small residuals. However, the linear fitting parameters varied with ionic composition and BSA concentration. The ‘apparent’ temperature (calibrated using the water–NAA frequency differences) decreased by approximately 1 °C for every 100 mm increase in ionic concentration and increased proportionally to the concentration of BSA. Copyright © 2012 John Wiley & Sons, Ltd.     

Stroke assessment with intravoxel incoherent motion diffusion‐weighted MRI

Intravoxel incoherent motion (IVIM) diffusion‐weighted MRI can simultaneously measure diffusion and perfusion characteristics in a non‐invasive way. This study aimed to determine the potential utility of IVIM in characterizing brain diffusion and perfusion properties for clinical stroke. The multi‐b‐value diffusion‐weighted images of 101 patients diagnosed with acute/subacute ischemic stroke were retrospectively evaluated. The diffusion coefficient D, representing the water apparent diffusivity, was obtained by fitting the diffusion data with increasing high b‐values to a simple mono‐exponential model. The IVIM‐derived perfusion parameters, pseudodiffusion coefficient D*, vascular volume fraction f and blood flow‐related parameter fD*, were calculated with the bi‐exponential model. Additionally, the apparent diffusion coefficient (ADC) was fitted according to the mono‐exponential model using all b‐values. The diffusion parameters for the ischemic lesion and normal contralateral region were measured in each patient. Statistical analysis was performed using the paired Student t‐test and Pearson correlation test. Diffusion data in both the ischemic lesion and normal contralateral region followed the IVIM bi‐exponential behavior, and the IVIM model showed better goodness of fit than the mono‐exponential model with lower Akaike information criterion values. The paired Student t‐test revealed significant differences for all diffusion parameters (all P < 0.001) except D* (P = 0.218) between ischemic and normal areas. For all patients in both ischemic and normal regions, ADC was significantly positively correlated with D (both r = 1, both P < 0.001) and f (r = 0.541, P < 0.001; r = 0.262, P = 0.008); significant correlation was also found between ADC and fD* in the ischemic region (r = 0.254, P = 0.010). For all pixels within the region of interest from a representative subject in both ischemic and normal regions, ADC was significantly positively correlated with D (both r = 1, both P < 0.001), f (r = 0.823, P < 0.001; r = 0.652, P < 0.001) and fD* (r = 0.294, P < 0.001; r = 0.340, P < 0.001). These findings may have clinical implications for the use of IVIM imaging in the assessment and management of acute/subacute stroke patients. Copyright © 2016 John Wiley & Sons, Ltd.     

Difference spectroscopy using PRESS asymmetry: application to glutamate,glutamine, and myo‐inositol

A simple, clinically viable technique utilizing PRESS and strong coupling properties is presented for discrimination of coupled brain metabolites. The method relies on signal variation due to alteration of inter‐echo timings (PRESS asymmetry) while maintaining a constant total echo time. Spin response of singlets and weakly coupled spins is unchanged due to PRESS asymmetry, allowing difference spectroscopy to detect unobstructed strongly coupled resonances. No changes to the standard PRESS sequence are required except variation of inter‐echo timings. The procedure is illustrated for the separate detection of glutamate from glutamine and the detection of myo‐inositol in simulation, phantom, and in vivo experiments at 4.7 T. The subtraction yields calculated from the simulation were 53% for glutamate and 75% for myo‐inositol, and a resultant contribution of 96% glutamate to the total glutamate/glutamine multiplet in the 2.04–2.14 ppm range. To extend the treatment to other field strengths and metabolites, an analytical approximation based on a strongly coupled AB system was used to model individual spin groups. Subtraction spectroscopy yields for different combinations of coupling parameters were calculated for the detection of various strongly coupled metabolites at common clinical field strengths. The approximation also predicts adequate glutamate/glutamine discrimination at 3.0 T using the difference spectroscopy method. Copyright © 2009 John Wiley & Sons, Ltd.     

Combined spiroergometry and 31P–MRS of human calf muscle during high‐intensity exercise

Simultaneous measurements of pulmonary oxygen consumption (VO₂), carbon dioxide exhalation (VCO₂) and phosphorus magnetic resonance spectroscopy (³¹P–MRS) are valuable in physiological studies to evaluate muscle metabolism during specific loads. Therefore, the aim of this study was to adapt a commercially available spirometric device to enable measurements of VO₂ and VCO₂ whilst simultaneously performing ³¹P–MRS at 3 T. Volunteers performed intense plantar flexion of their right calf muscle inside the MR scanner against a pneumatic MR‐compatible pedal ergometer. The use of a non‐magnetic pneumotachograph and extension of the sampling line from 3 m to 5 m to place the spirometric device outside the MR scanner room did not affect adversely the measurements of VO₂ and VCO₂. Response and delay times increased, on average, by at most 0.05 s and 0.79 s, respectively. Overall, we were able to demonstrate a feasible ventilation response (VO₂ = 1.05 ± 0.31 L/min; VCO₂ = 1.11 ± 0.33 L/min) during the exercise of a single calf muscle, as well as a good correlation between local energy metabolism and muscular acidification (τ_{PCr fast} and pH; R² = 0.73, p < 0.005) and global respiration (τ_{PCr fast} and VO₂; R² = 0.55, p = 0.01). This provides improved insights into aerobic and anaerobic energy supply during strong muscular performances.