Imaging Assessment of the Efficacy of Chemotherapy in Primary Malignant Bone Tumors: Recent Advances in Qualitative and Quantitative Magnetic Resonance Imaging and Radiomics.

Recent studies have shown that MRI demonstrates promising results for evaluating the chemotherapy efficacy in bone sarcomas. This article reviews current methods for evaluating the efficacy of malignant bone tumors and the application of MRI in this area, and emphasizes the advantages and limitations of each modality. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 2.

Reducing optical loss of dual-ion beam sputtered HfO2 films via optimization of coating and annealing parameters.

HfO2 films are widely used for optical coatings due to the high refractive index and low absorption, especially in the ultraviolet (UV) band. In this work, HfO2 film samples were prepared with the optimized assistant source power and deposition temperature by dual-ion beam sputtering (DIBS), followed by annealing treatments in vacuum and atmosphere, respectively. For samples with different annealing temperatures from 200 to 450 °C, the microstructure, morphology, film stress and optical properties from 200 to 1000 nm were systematically investigated. A monoclinic phase, a refractive index inhomogeneity along the film thickness and an absorption of shoulder-shape in the 250-300 nm band were found in the as-deposited samples. For samples annealed in vacuum, 400 °C annealing leaded to more oxygen defects, which in turn caused aggravated UV absorption. For samples annealed in atmosphere, the shoulder-shaped absorption weakened obviously above 300 °C annealing, which was suspected due to the reduction of oxygen defects during the crystallization process with sufficient oxygen. Scattering loss was investigated and found negligible for as-deposited and annealed samples. Additionally, film stress varied from compressive state to tensile state with increasing annealing temperature, and the zero-stress temperature is between 300-350 °C, which is due to the obvious crystallization behavior. Production methods and physical mechanisms for low absorption and scattering loss DIBS deposited HfO2 films were proposed and discussed in detail.

High-flux wavelength tunable XUV source in the 12-40.8†eV photon energy range with adjustable energy and time resolution for Tr-ARPES applications.

High-order harmonic generation (HHG) has a broad spectrum covering vacuum ultraviolet to extreme ultraviolet (XUV) bands, which is useful for applications involving material analyses at different information depths. Such an HHG light source is perfect for time- and angle-resolved photoemission spectroscopy. Here, we demonstrate a high-photon flux HHG source driven by a two-color field. Applying a fused silica compression stage to reduce the driving pulse width, we obtained a high XUV photon flux of 2 × 1012 phs/s @21.6 eV on target. We designed a classical diffraction mounted (CDM) grating monochromator that can achieve a wide range of photon energy from 12 to 40.8 eV, while the time resolution is improved by reducing the pulse front tilt after the harmonic selection. We designed a spatial filtering method to adjust the time resolution using the CDM monochromator and significantly reduced the pulse front tilt of the XUV pulses. We also demonstrate a detailed prediction of the energy resolution broadening which is caused by the space charge effect.

Advances in The Application and Research of Magnetic Resonance Diffusion Kurtosis Imaging in The Musculoskeletal System.

Magnetic resonance diffusion kurtosis imaging (DKI) is an emerging magnetic resonance imaging (MRI) technique that can reflect microstructural changes in tissue through non-Gaussian diffusion of water molecules. Compared to traditional diffusion weighted imaging (DWI), the DKI model has shown greater sensitivity for diagnosis of musculoskeletal diseases and can help formulate more reasonable treatment plans. Moreover, DKI is an important auxiliary examination for evaluation of the motor function of the musculoskeletal system. This article briefly introduces the basic principles of DKI and reviews the application and research of DKI in the evaluation of disorders of the musculoskeletal system (including bone tumors, soft tissue tumors, spinal lesions, chronic musculoskeletal diseases, musculoskeletal trauma, and developmental disorders) as well as the normal musculoskeletal tissues. EVIDENCE LEVEL: 5 TECHNICAL EFFICACY: 1.

Correlation of Intravoxel Incoherent Motion and Diffusion Kurtosis MR Imaging Models With Reactive Stromal Grade in Prostate Cancer.

BACKGROUND: Reactive stroma is recognized as one of the independent prognostic factors in prostate cancer (PCa). Intravoxel incoherent motion (IVIM) and diffusion kurtosis imaging (DKI) may be useful for assessing the reactive stromal grade (RSG). PURPOSE: To investigate whether IVIM and DKI models can evaluate RSG in PCa patients. STUDY TYPE: Retrospective. SUBJECTS: A total of 56 PCa patients aged 73 years on average confirmed by MRI and transrectal ultrasound (MRI/TRUS) fusion biopsy divided into two subgroups (18 high RSG and 38 low RSG). FIELD STRENGTH/SEQUENCE: A 3 T/T1 WI-fs, T1 WI, T2 WI-fs, T2 WI, DWI, IVIM, and DKI. ASSESSMENT: Apparent diffusion coefficient (ADC), pure diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), mean diffusion (MD), mean kurtosis (MK) were obtained. Patients were divided into high RSG PCa (>50% reactive stroma) and low RSG PCa (≤50% reactive stroma) groups on hematoxylin and eosin (H&E) stained sections. STATISTICAL TESTS: Spearman correlation and independent sample t-test or Wilcoxon's rank sum test was used to investigate the relationship between each imaging parameter and RSG. The combined parameters were calculated using a binary logistic regression model. Receiver operating characteristic (ROC) analysis was used to explore the value of individual and combined parameters to differentiate between high and low RSG group. Area under the ROC curves (AUC) > 0.7 were used as reference standards. RESULTS: ADC, D, f, and MD values showed positive correlation with RSG (r = 0.489, 0.619, 0.318, and 0.544, respectively); MK showed negative correlation with RSG (r = -0.444). ADC, D, f, and MD values were significantly lower in the low RSG group than in the high RSG group. The combined model showed the best diagnostic ability to differentiate low and high RSG groups (AUC = 0.887). DATA CONCLUSION: Parameters of IVIM and DKI may be promising methods for assessment of RSG in PCa patients. EVIDENCE LEVEL: 3. TECHNICAL EFFICACY: Stage 3.

Correlation of Apparent Diffusion Coefficient With Proliferation and Apoptotic Indexes in a Murine Model of Fibrosarcoma: Comparison of Four Methods for MRI Region of Interest Positioning.

BACKGROUND: Diffusion-weighted imaging (DWI) has demonstrated great potential in predicting the expression of tumor cell proliferation and apoptosis indexes. PURPOSE: To evaluate the impact of four region of interest (ROI) methods on interobserver variability and apparent diffusion coefficient (ADC) values and to examine the correlation of ADC values with Ki-67, Bcl-2, and P53 labeling indexes (LIs) in a murine model of fibrosarcoma. STUDY TYPE: Prospective, animal model. ANIMAL MODEL: A total of 22 female BALB/c mice bearing intramuscular fibrosarcoma xenografts. FIELD STRENGTH/SEQUENCE: A 3.0 T/T1-weighted fast spin-echo (FSE), T2-weighted fast relaxation fast spin-echo, and DWI PROPELLER FSE sequences. ASSESSMENT: Four radiologists measured ADC values using four ROI methods (oval, freehand, small-sample, and whole-volume). Immunohistochemical assessment of Ki-67, Bcl-2, and P53 LIs was performed. STATISTICAL TESTS: Interclass correlation coefficient (ICC), one-way analysis of variance followed by LSD-t post hoc analysis, and Pearson correlation test were performed. The statistical threshold was defined as a P-value of <0.05. RESULTS: All ROI methods for ADC measurements showed excellent interobserver agreement (ICC range, 0.832-0.986). The ADC values demonstrated significant differences among the four ROI methods. The ADC values for oval, freehand, small-sample, and whole-volume ROI methods showed a moderately negative correlation with Ki-67 (r = -0.623; r = -0.629; r = -0.642, and r = -0.431) and Bcl-2 (r = -0.590; r = -0.597; r = -0.659, and r = -0.425) LIs, but no correlation with P53 LI (r = 0.364, P = 0.104; r = 0.350, P = 0.120; r = 0.379, P = 0.091; r = 0.390, P = 0.080). DATA CONCLUSION: The ADC value can be used to evaluate cell proliferation and apoptosis indexes in a murine model of fibrosarcoma, employing the small-sample ROI as a reliable method. EVIDENCE LEVEL: 1 TECHNICAL EFFICACY: Stage 3.

Correlation between IVIM parameters and microvessel architecture: direct comparison of MRI images and pathological slices in an orthotopic murine model of rhabdomyosarcoma.

OBJECTIVE: This study aimed to explore the correlation between intravoxel incoherent motion (IVIM) parameters and microvessel architecture (microvessel density (MVD), vasculogenic mimicry (VM), and pericyte coverage index (PCI)) in an orthotopic murine model of rhabdomyosarcoma. METHODS: The murine model was established by injecting rhabdomyosarcoma-derived (RD) cells into the muscle. Nude mice underwent routine magnetic resonance imaging (MRI) and IVIM examinations with ten b values (0, 50, 100, 150, 200, 400, 600, 800, 1000, and 2000 s/mm2). D, D*, and f values were calculated with the ADW4.7 workstation. MRI images and pathological slices were directly compared to ensure that radiology parameters accurately reflect pathology. MVD, VM, PCI, and cellularity were obtained by histological analysis. The correlations were assessed between IVIM parameters (D, D*, f, and fD* values) and pathological markers (MVD, VM, PCI, and cellularity). RESULTS: The average of D, D*, f, and fD* values were 0.55 ± 0.07 × 10-3 mm2/s, 5.25 ± 0.73 × 10-3 mm2/s, 13.39 ± 7.68%, and 0.73 ± 0.49 × 10-3 mm2/s, respectively. The average of MVD, VM, PCI, and cellularity were 41.91 ± 10.98, 1.16 ± 0.83, 0.49 ± 0.18, and 39.15 ± 9.00%. D*, f, and fD* values showed a positive correlation with MVD separately, while the D value did not correlate with MVD. D value negatively correlated to VM moderately, and other parameters did not associate with VM. D* and fD* values were positively correlated with PCI, but no correlation was observed between other parameters and PCI. CONCLUSIONS: IVIM may evaluate the tumor microvessel architecture. D*, f, and fD* may reflect the endothelial lining blood vessel; D could indirectly reflect the VM; D* and fD* could reflect PCI(the normal degree of the tumor blood vessel). CLINICAL RELEVANCE STATEMENT: An intravoxel incoherent motion may be useful in assessing rhabdomyosarcoma microvessel structure to predict the target and effectiveness of anti-angiogenic therapy. KEY POINTS: • IVIM may be used to evaluate the tumor microvessel architecture in the mouse rhabdomyosarcoma model. • The MRI-pathology control method achieves correspondence between MRI slices and pathology slices, which ensures the consistency of the ROI of MRI and the pathology observation region.

Correlations between DWI, IVIM, and HIF-1α expression based on MRI and pathology in a murine model of rhabdomyosarcoma.

PURPOSE: To investigate the correlation between DWI, intravoxel incoherent motion (IVIM), and hypoxia-inducible factor 1-alpha (HIF-1α) expression in a nude mouse model of rhabdomyosarcoma based on imaging and pathological comparisons. METHODS: Human rhabdomyosarcoma-derived (RD) cells were inoculated into the right thigh muscle of 20 BALB/c female nude mice. Mice were imaged using 3.0 Tesla MRI system. T1 -weighted imaging, T2 -weighted imaging, DWI, and IVIM images were obtained. ADW4.7 (GE Healthcare, ChicagoAQ34, IL, USA) was used for image processing of ADC, Dslow , Dfast , and f values. All parameter values were independently analyzed by 2 observers. Immunohistochemistry of HIF-1α was performed. We used a specific image-pathology comparison method to ensure correct overlap between the image plane and the pathological section. Mann-Whitney U test or independent sample t test, Pearson or Spearman correlation test, the intragroup correlation coefficient, Kolmogorov-Smirnov test, and receiver operating characteristic curve were used. The correlation between DWI and intravoxel incoherent motion parameter values and HIF-1α expression was determined. RESULTS: There were 10 mice in the low-expression group and 7 in the high-expression group. The ADC and Dslow values were negatively correlated with HIF-1α with correlation coefficients of -0.491 and - 0.702 (P = 0.045 and 0.002). The f value positively correlated with HIF-1α expression (r = 0.485, P = 0.048). ADC, Dslow , and f were significantly different between the high-HIF-1α expression tumors and the low-HIF-1α expression tumors. ADC showed the best predictive performance among all parameters (area under the curve = 0.652, sensitivity = 83.3%, specificity = 63.6%). CONCLUSION: The parameter values of DWI and intravoxel incoherent motion can be used to evaluate the expression of HIF-1α in rhabdomyosarcoma. ADC, Dslow , and f value showed correlation with the expression of HIF-1α.

Power-scalable thin-disk Ti:sapphire laser amplifier.

We experimentally demonstrate a 38-fs chirped-pulse amplified (CPA) Ti:sapphire laser system based on the power-scalable thin-disk scheme with an average output power of 1.45 W at a repetition rate of 1 kHz, corresponding to peak power of 38 GW. The beam profile close to the diffraction limit with a measured M2 value of approximately 1.1 is obtained. It demonstrates the potential for an ultra-intense laser with high beam quality compared with the conventional bulk gain amplifier. To the best of our knowledge, this is the first reported Ti:sapphire regenerative amplifier based on the thin-disk approach reaching 1 kHz.

DWI and IVIM Imaging in a Murine Model of Rhabdomyosarcoma: Correlations with Quantitative Histopathologic Features.

BACKGROUND: High cellularity and abnormal interstitial structures are some of the unfavorable factors that affect the treatment outcomes and survival of rhabdomyosarcoma (RMS) patients. PURPOSE: To explore the correlation between diffusion-weighted imaging (DWI) and intravoxel incoherent motion (IVIM) with quantitative histopathologic features in a murine model of RMS. STUDY TYPE: Prospective. ANIMAL MODEL: Murine model of RMS (31 female BALB/c nude mice). FIELD STRENGTH/SEQUENCE: 3.0 T; fast spin-echo (FSE) T1-weighted imaging, fast relaxation fast spin-echo (FRFSE) T2-weighted imaging, DWI PROPELLER FSE imaging sequence, and IVIM echo planar imaging sequence; 10 different b-values (0, 50, 100, 150, 200, 400, 600, 800, 1000, and 1200 s/mm2 ). ASSESSMENT: Magnetic resonance imaging (MRI) was performed after 30-45 days of implantation. The following MRI parameters were calculated: apparent diffusion coefficient (ADC), pure diffusion coefficient (D), pseudo-diffusion coefficient (D*), and perfusion fraction (f). Histopathologic features, which contained nuclear, cytoplasmic, and stromal fractions, and the nuclear-to-cytoplasmic ratio within the tumor were measured using image-based segmentation. STATISTICAL TESTS: Pearson's correlation, multiple linear regression analysis, and receiver operating characteristic curve analysis were performed. A P < 0.05 was considered statistically significant. RESULTS: The ADC value showed moderate negative correlation with nuclear fraction (r = -0.540), and moderate positive correlation with stroma fraction (r = 0.474). The D value showed moderate negative correlation with nuclear fraction (r = -0.491), and moderate positive correlation with stroma fraction (r = 0.421). The f value showed a moderate negative correlation with stroma fraction (r = -0.423). The D value showed the best diagnostic ability. The optimal cut-off D value of 0.460 was associated with 77.8% sensitivity and 68.2% specificity (area under the curve, 0.747). DATA CONCLUSION: The ADC, D, and f values obtained from DWI and IVIM images showed moderate correlation with the quantitative histopathologic features in a murine model of RMS. LEVEL OF EVIDENCE: 1 TECHNICAL EFFICACY STAGE: 3.

An Update in Imaging Evaluation of Histopathological Grade of Soft Tissue Sarcomas Using Structural and Quantitative Imaging and Radiomics.

Over the past two decades, considerable efforts have been made to develop non-invasive methods for determining tumor grade or surrogates for predicting the biological behavior, aiding early treatment decisions, and providing prognostic information. The development of new imaging tools, such as diffusion-weighted imaging, diffusion kurtosis imaging, perfusion imaging, and magnetic resonance spectroscopy have provided leverage in the diagnosis of soft tissue sarcomas. Artificial intelligence is a new technology used to study and simulate human thinking and abilities, which can extract and analyze advanced and quantitative image features from medical images with high throughput for an in-depth characterization of the spatial heterogeneity of tumor tissues. This article reviews the current imaging modalities used to predict the histopathological grade of soft tissue sarcomas and highlights the advantages and limitations of each modality. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 2.

Generation of annular femtosecond few-cycle pulses by self-compression and spatial filtering in solid thin plates.

Annular-shaped femtosecond few-cycle pulses are generated by 40fs laser pulses propagating through 6 solid thin plates in numerical simulations as well as in experiments. The generation of such pulses takes advantage of the conical emission caused by plasma effect, which introduces continuously varying off-axis plasma density along the radial direction of the propagating beam. The negative dispersion induced by the plasma causes the pulse at particular radial location to be self-compressed and to form an annular beam of short pulse, which can be extracted simply by spatial filtering. Meanwhile, by adjusting the input pulse energy and position of each thin plate relative to the laser focus, we control the plasma density in thin plates which changes the ratio between ionization and effects providing positive dispersion, and obtain a higher compression ratio indicating that the scheme of solid thin plates has the flexibility to regulate the laser intensity so as to plasma density, thus the negative dispersion the pulse experiences during propagation. Few-cycle pulses as short as 8.8 fs are generated in experiments, meanwhile the shortest pulse duration found in the simulations is 5.0 fs, which corresponds to two optical cycles at its central wavelength 761 nm. This method has great potential in high-power few-cycle pulse generation.

Double-pass pre-chirp managed amplification with high gain and high average power.

We demonstrate, to the best of our knowledge, the first double-pass pre-chirp managed fiber amplifier. The double-pass fiber amplifier exhibits high gain allowing us to amplify chirped picosecond pulses from 20 mW to 113 W in a rod-type Yb-fiber corresponding to 38 dB gain. We study the dependence of static mode degradation (SMD) on the nonlinear phase shift (NPS) accumulated by the amplified pulse. Our results indicate that a larger nonlinear phase shift results in stronger nonlinear polarization evolution of the fundamental mode and leads to a lower threshold for SMD. After optimization, our pre-chirp managed amplifier seeded by 80 mW pulses delivers 102 W amplified power from the main output. The amplified pulses are compressed to 37 (55) fs with 90 (100) W average power by a grating pair (chirped mirrors). The double-pass configuration significantly simplifies the implementation of pre-chirp managed fiber amplifiers leading to an extremely compact system.

Phase-error-compensation-based surface recovery algorithm using spectrum selection for white light interferometry.

White light interferometry is a well-established surface recovery technique. In this paper, a white light signal processing algorithm based on phase error compensation using spectrum selection is proposed. The derived nonlinear phase distribution from the correlogram is modeled as the combination of random errors and systemic deviations. By developing a new, to the best of our knowledge, recovery algorithm, the phase noise can be separated from the linear map and significantly attenuated. Based on the proposed algorithm, the spectrum features of white light LEDs and halogen lamps are investigated in detail. The inner products defined by three selected points are employed to generate a coefficient to evaluate the linearity of an unwrapped phase map within a certain spectrum region. The optimal spectrum range corresponding to the best measurement performance can then be located where the coefficient approximates 1 and the spectrum energy stays relatively high. The simulations are carried out under different levels of SNR and scan step noises, which show that the new method can effectively reduce additional disturbance from the recovered topography. In experiments, the system with the proposed method is first calibrated by a step height standard (VLSI, 182.7±2.0nm) with the repeatability of 0.44 nm. A silicon wafer and three roughness standards are also tested to further verify the robustness of the new method.

Ultra-low-noise carrier-envelope phase stabilization of a Kerr-lens mode-locked Yb:CYA laser frequency comb with a feed-forward method.

We demonstrate ultra-low-noise carrier-envelope phase stabilization of a Kerr-lens mode-locked Yb:CYA laser frequency comb, which is the first time, to the best of our knowledge, that the feed-forward method has been applied to 1 µm all-solid-state lasers. We obtain a signal-to-noise ratio of more than 38 dB at a 100 kHz resolution bandwidth for carrier-envelope phase offset beat signals in two standard in-loop and out-of-loop f-2f interferometers. The residual integrated phase noise amounts to 79.3 mrad from 1 Hz to 1 MHz, corresponding to a timing jitter of 44 as. We also investigate long-term performances of the CEP stabilization in the feed-forward scheme, phase-locked feedback systems and the combining locking techniques in terms of sub-hertz frequency-resolved phase noise and Allan deviation in 1000 s. The results indicate that, although the feed-forward CEP stabilization method dominates in the range of 1 Hz to 1 MHz Fourier frequency, feedback methods with a time integration effect are superior in sub-hertz Fourier frequency phase noise suppression.

Highly-stable mode-locked PM Yb-fiber laser with 10 nJ in 93-fs at 6 MHz using NALM.

We demonstrate highly stable mode-locked Yb-doped fiber oscillators using a nonlinear amplifying loop mirror, delivering linearly polarized laser pulses with high energy at a low repetition rate of several MHz. These lasers are composed of polarization-maintaining fibers and fiber-based components without intra-cavity dispersion compensation. The spectral and temporal characteristics are systematically investigated at different repetition rates. Spectral bandwidth of 31 nm is realized in the case of 6 MHz repetition rate, and the pulse energy reaches 10 nJ. A pair of gratings compresses the output pulse to 93 fs. RMS power stability is as low as 0.04% in 10 hours, which shows excellent stability. We believe that this type of fiber oscillator is an ideal seed for further high power amplification.

Watt-level widely tunable femtosecond mid-infrared KTiOAsO4 optical parametric oscillator pumped by a 1.03 µm Yb:KGW laser.

A high-power, high-repetition-rate, broadband tunable femtosecond optical parametric oscillator (OPO) is constructed based on KTiOAsO4 crystal, pumped by a 75.5 MHz mode-locked Yb:KGW laser. With 7 W pump power, the OPO generates as much as 2.32 W of signal power at 1.55 µm and 1.31 W of idler power at 3.05 µm, corresponding to a total conversion efficiency of 51.8%. Operating at 151 MHz repetition rate, the wavelength of the signal covers 1.41-1.71 µm with a tunable idler range of 2.61-3.84 µm. The idler bandwidth is more than 180 nm over the entire mid-infrared range. By compensating intracavity dispersion, the signal pulse has a nearly Fourier transform-limited duration of 129 fs at 1.52 µm.

65-fs Yb-doped all-fiber laser using tapered fiber for nonlinearity and dispersion management.

We implement an ultrafast Yb-doped all-fiber laser which incorporates tapered single-mode fibers for managing nonlinearity and dispersion. The tapered fiber placed in the oscillator cavity aims to broaden the optical spectrum of the intracavity pulse. At the oscillator output, we use another tapered fiber to perform pulse compression. The resulting 66.1-MHz Yb-doped all-fiber oscillator self-starts and generates 0.4-nJ, 65-fs pulses, which can serve as a compact and robust seed source for subsequent high-power, high-energy amplifiers.

Tungsten disulfide saturable absorbers for 67 fs mode-locked erbium-doped fiber lasers.

In this paper, we demonstrate 67 fs pulse emitting with tungsten disulfide (WS2) in mode-locked erbium-doped fiber (EDF) lasers. Using the pulsed laser deposition method, WS2 is deposited on the surface of the tapered fiber to form the evanescent field. The fiber-taper WS2 saturable absorber (SA) with the large modulation depth is fabricated to support the ultrashort pulse generation. The influences of the WS2 SA are analyzed through contrastive experiments on fiber lasers with or without the WS2 SA. The pulse duration is measured to be 67 fs, which is the shortest pulse duration obtained in the mode-locked fiber lasers with two dimensional (2D) material SAs. Compared to graphene, topological insulator, and other transition metal dichalcogenides (TMDs) SAs, results in this paper indicate that the fiber-taper WS2 SA with large modulation depth is a more promising photonic device in mode-locked fiber lasers with the wide spectrum and ultrashort pulse duration.

High-efficiency supercontinuum generation in solid thin plates at 0.1 TW level.

Supercontinuum generation in a solid-state medium was investigated experimentally. A continuum covering 460 to 950 nm was obtained when 0.8 mJ/30 fs Ti:sapphire laser pulses were applied to seven thin fused silica plates at a 1 kHz repetition rate. The primary processes responsible for spectral broadening were self-phase modulation (SPM) and self-steepening, while SPM and self-focusing were balanced to optimize the spectral broadening and suppress the multiphoton process. The output was compressed to a 5.4 fs and a 0.68 mJ pulse, corresponding to two optical cycles and 0.13 TW of peak power.