Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: (a) Schematic showing the features of monochrome, red-green-blue (RGB), spectroscopy, multispectral, and HSI [17]. As shown in the figure, both spectroscopy and HSI can store wavelength information over the entire spectrum. However, spectroscopy cannot provide precise spatial (location within the sample) information. RGB imaging does not allow for spectral information at all (information across multiple wavelengths) and is insensitive to components that are at different wavelengths than red (630 nm), green (545 nm) and blue (435 nm) but does enable spatial information. Spectroscopy allows for spectral information to be gleaned but doesn't allow for spatial information. HSI (300–2600 nm) can collect spatial, spectral, multicomponent while being sensitive to a variety of different wavelengths or components. (b) Detailed comparison showing the differences between HSI and RGB imaging [15]. The figure depicts light reflectance curve of a single pixel from an arbitrary sample imaged using hyperspectral spectroscopy and RGB imaging. The hyperspectral image contains information in a continuous visible near-infrared spectrum compared to the intensity curve from RGB imaging that provides data at only three prominent wavelengths. The additional spectral information contained with the continuous hyperspectral image can be utilized to more accurately analyze and understand micro- and nanoscale features that are not feasible using the discrete RGB imaging dataset. The caption text refers to online color version of the figure.

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Spectroscopy, Imaging

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Schematic showing different approaches used for HSI [17]: (a) Whiskbroom, (b) push broom, (c) staring, (d) snapshot. Briefly, the dispersive element for whiskbroom, push broom, and snapshot is either a prism, a grating, or a prism gating prism while for spectral scan, it is a tunable filter or an interferometer. The wavelength range is wide for whiskbroom, push broom, and snapshot while it is medium for staring. The wavelength selection is partial for both whiskbroom and push broom and complete for staring and unavailable in snapshot. The spectral resolution is high for both whiskbroom and push broom while it is low for snapshot and medium for staring. Whiskbroom and staring are hyperspectral while snapshot is multispectral. The throughput is high for whiskbroom, push broom, and snapshot and low for staring. The data cube collection is relatively long for both whiskbroom and push broom while it is short for staring and fast for snapshot. However, the complexity is high for whiskbroom and push broom while it is simple for staring and medium for snapshot. The associated costs are low for both whiskbroom and push broom, medium for snapshot, and high for staring.

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Selection

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: A summary of the current commercially available HSI systems

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Imaging, Software

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: A summary of the single cell analysis/applications using HSI modalities

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Software, Microscopy, Transmission Assay, Expressing, Imaging, Fluorescence, Irradiation

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Hyperspectral dark-field imaging using plasmonic nanoprobes to quantify 5-carboxylcytosine (5caC) modification on DNA in single cells [12]. The study by Wang et al. [12] revealed the distribution of 5caC at different cell-cycle stages and demonstrated that 5caC is an inherited epigenetic marker. As stated by the authors, the hyperspectral dark-field imaging efficiently removes scattering noises from nonspecifically aggregated nanoprobes. The image shows the filter function applied to: (a) Raw image (b) converted to spectrally mapped images from 520–620 nm, (c) image of cell at a wavelength above 635 nm, and (d) number of gold nanoparticle (shown as green dots) inside the cell. The caption text refers to online color version of the figure.

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Imaging, Modification, Marker

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Study of 3D rotational dynamics of gold nanorods inside live HEK293 cells using CytoViva HSI system by Chaudhari and Pradeep [90]. (a) Scattering spectra of a single gold nanorod attached on the cell membrane. The inset shows the corresponding hyperspectral image. (b) Scattering spectra of the gold nanorod in (a) after being absorbed by the cell. The inset shows the corresponding hyperspectral image. (c) Actual image of cell being monitored to study rotational dynamics. The gold nanorod is marked with a square. Inset shows an enlarged view of the gold nanorod. Light scattered in the Z direction was collected through analyzer, whose orientation is shown by yellow double arrow. (d) Time variation of scattering intensity of the gold nanorod. Time scale corresponds to the axis of the graph shown below. Pink vertical bars show the region where microscope focus was adjusted on the particle after it went out of the focal plane. (e) Time variation of width of gold nanorod spot in two-dimensional Gaussian width of the gold nanorod. See Chaudhari and Pradeep [90] for further details. (f) Representation of gold nano particle path inside the HEK293 cell. Green arrow shows the time point from where temporal data of the GNR is shown. Color of the trace corresponds to the time scale of graphs (D, E). Please note that the background image is just to give a rough idea of the position of GNR inside the cell.

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Microscopy

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Hyperspectral fluorescence imaging of SHSY5Y cells containing iron from Oh et al. [18]. (a) Dark-field images of the SHSY5Ycells incubated with iron (specifically ferric ammonium nitrate) for 1 h. The areas mapped with HSI are marked by colored boxes. (b) Spectral profiles collected from each region shown in (a). Peaks are mainly observed between 450 to 650 nm. All bulk iron areas have a peak absorbance near 600 nm, whereas the peak signal in the cells is near 500 nm and the signal on the glass plate is almost zero. (c) Magnified or zoomed-in images (17×) of pixels containing HSI data from glass, cytoplasm, nucleus, and bulk iron.

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques: Fluorescence, Imaging, Incubation

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Label-free dark-field HSI of human RBCs from Conti et al. [13]: (a) Hyperspectral image of erythrocyte sample. (b) Spectral library composed of different endmembers with random color code. (c) Zoomed-in image of one erythrocyte. (d) Color mapping matching spectra of the spectral library. (e) Mapping five main components of RBCs namely phospholipid, cholesterol, hemoglobin, spectrin, and protoporphyrin. Briefly, 5 μl of whole blood was loaded in the center of glass slide and sandwiched with coverslip. After 120 min, to allow for image stability, the optical acquisition was started. Each image consisted of approximately 30 regularly shaped RBC as shown in (a) with no other cells. One RBC was chosen as shown in (c), (d), and (e) for further image analysis. For the RBCs, eight spectra (b) were individuated with optimal coverage of the optical image (d). Applying the SAM function, the spectral distribution of the 8 endmember spectra in the samples was then determined (data not shown) as described in Conti et al. [13]. This study demonstrated a fast, easy, and repeatable protocol to study large number of cells and to the possibility of mapping single molecules, proteins as well as structure of cell membranes with applications in personalized medicine and membrane-targeted therapies [13]. The caption text refers to online color version of the figure.

Article Snippet: Photo-thermal tumor ablation , Irradiation of nano-aggregates to produce heat killing cancer cells , MDA-MB-231 cells , Cytoviva Hyperspectral System , ENVI 4.4 , 400–1000 nm , 550 nm, 785 nm , 2.5 nm , Nanoparticles , [ 84 ].

Techniques:

Journal: Scientific Reports

Article Title: Hyperspectral imaging and deep learning for parasite detection in white fish under industrial conditions

doi: 10.1038/s41598-024-76808-w

Figure Lengend Snippet: Example of different results for the automatic detection of nematodes with hyperspectral imaging and the proposed processing approach. ( a , b ) Successful detection of nematodes. ( b ) Example of a false negative. ( c ) Example of a false positive.

Article Snippet: The hyperspectral camera is a HySpex Baldur V-1024N (Norsk Elektro Optikk AS, Oslo, Norway), which is a pushbroom camera covering the VNIR (visual and near-infrared) spectral range (485-960 nm) with a spectral resolution of 5.5 nm (88 spectral bands) and 1024 spatial pixels.

Techniques: Imaging

Journal: Scientific Reports

Article Title: Hyperspectral imaging and deep learning for parasite detection in white fish under industrial conditions

doi: 10.1038/s41598-024-76808-w

Figure Lengend Snippet: Workflow for the identification and labeling of nematodes. ( a ) Manual examination of the fillets for nematodes using a candling table, and digital annotation using an iPad. ( b ) Example of manual annotation of nematodes in the digital images (blue color circling). ( c ) Example of digital annotation of nematodes in a synthetic RGB image extracted from the hyperspectral cube (yellow color circling).

Article Snippet: The hyperspectral camera is a HySpex Baldur V-1024N (Norsk Elektro Optikk AS, Oslo, Norway), which is a pushbroom camera covering the VNIR (visual and near-infrared) spectral range (485-960 nm) with a spectral resolution of 5.5 nm (88 spectral bands) and 1024 spatial pixels.

Techniques: Labeling

Journal: Scientific Reports

Article Title: Hyperspectral imaging and deep learning for parasite detection in white fish under industrial conditions

doi: 10.1038/s41598-024-76808-w

Figure Lengend Snippet: Processing workflow for the detection of nematodes from hyperspectral image data.

Article Snippet: The hyperspectral camera is a HySpex Baldur V-1024N (Norsk Elektro Optikk AS, Oslo, Norway), which is a pushbroom camera covering the VNIR (visual and near-infrared) spectral range (485-960 nm) with a spectral resolution of 5.5 nm (88 spectral bands) and 1024 spatial pixels.

Techniques:

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: (a) Schematic showing the features of monochrome, red-green-blue (RGB), spectroscopy, multispectral, and HSI [17]. As shown in the figure, both spectroscopy and HSI can store wavelength information over the entire spectrum. However, spectroscopy cannot provide precise spatial (location within the sample) information. RGB imaging does not allow for spectral information at all (information across multiple wavelengths) and is insensitive to components that are at different wavelengths than red (630 nm), green (545 nm) and blue (435 nm) but does enable spatial information. Spectroscopy allows for spectral information to be gleaned but doesn't allow for spatial information. HSI (300–2600 nm) can collect spatial, spectral, multicomponent while being sensitive to a variety of different wavelengths or components. (b) Detailed comparison showing the differences between HSI and RGB imaging [15]. The figure depicts light reflectance curve of a single pixel from an arbitrary sample imaged using hyperspectral spectroscopy and RGB imaging. The hyperspectral image contains information in a continuous visible near-infrared spectrum compared to the intensity curve from RGB imaging that provides data at only three prominent wavelengths. The additional spectral information contained with the continuous hyperspectral image can be utilized to more accurately analyze and understand micro- and nanoscale features that are not feasible using the discrete RGB imaging dataset. The caption text refers to online color version of the figure.

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Spectroscopy, Imaging

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Schematic showing different approaches used for HSI [17]: (a) Whiskbroom, (b) push broom, (c) staring, (d) snapshot. Briefly, the dispersive element for whiskbroom, push broom, and snapshot is either a prism, a grating, or a prism gating prism while for spectral scan, it is a tunable filter or an interferometer. The wavelength range is wide for whiskbroom, push broom, and snapshot while it is medium for staring. The wavelength selection is partial for both whiskbroom and push broom and complete for staring and unavailable in snapshot. The spectral resolution is high for both whiskbroom and push broom while it is low for snapshot and medium for staring. Whiskbroom and staring are hyperspectral while snapshot is multispectral. The throughput is high for whiskbroom, push broom, and snapshot and low for staring. The data cube collection is relatively long for both whiskbroom and push broom while it is short for staring and fast for snapshot. However, the complexity is high for whiskbroom and push broom while it is simple for staring and medium for snapshot. The associated costs are low for both whiskbroom and push broom, medium for snapshot, and high for staring.

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Selection

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: A summary of the current commercially available HSI systems

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Imaging, Software

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: A summary of the single cell analysis/applications using HSI modalities

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Software, Microscopy, Transmission Assay, Expressing, Imaging, Fluorescence, Irradiation

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Hyperspectral dark-field imaging using plasmonic nanoprobes to quantify 5-carboxylcytosine (5caC) modification on DNA in single cells [12]. The study by Wang et al. [12] revealed the distribution of 5caC at different cell-cycle stages and demonstrated that 5caC is an inherited epigenetic marker. As stated by the authors, the hyperspectral dark-field imaging efficiently removes scattering noises from nonspecifically aggregated nanoprobes. The image shows the filter function applied to: (a) Raw image (b) converted to spectrally mapped images from 520–620 nm, (c) image of cell at a wavelength above 635 nm, and (d) number of gold nanoparticle (shown as green dots) inside the cell. The caption text refers to online color version of the figure.

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Imaging, Modification, Marker

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Study of 3D rotational dynamics of gold nanorods inside live HEK293 cells using CytoViva HSI system by Chaudhari and Pradeep [90]. (a) Scattering spectra of a single gold nanorod attached on the cell membrane. The inset shows the corresponding hyperspectral image. (b) Scattering spectra of the gold nanorod in (a) after being absorbed by the cell. The inset shows the corresponding hyperspectral image. (c) Actual image of cell being monitored to study rotational dynamics. The gold nanorod is marked with a square. Inset shows an enlarged view of the gold nanorod. Light scattered in the Z direction was collected through analyzer, whose orientation is shown by yellow double arrow. (d) Time variation of scattering intensity of the gold nanorod. Time scale corresponds to the axis of the graph shown below. Pink vertical bars show the region where microscope focus was adjusted on the particle after it went out of the focal plane. (e) Time variation of width of gold nanorod spot in two-dimensional Gaussian width of the gold nanorod. See Chaudhari and Pradeep [90] for further details. (f) Representation of gold nano particle path inside the HEK293 cell. Green arrow shows the time point from where temporal data of the GNR is shown. Color of the trace corresponds to the time scale of graphs (D, E). Please note that the background image is just to give a rough idea of the position of GNR inside the cell.

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Microscopy

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Hyperspectral fluorescence imaging of SHSY5Y cells containing iron from Oh et al. [18]. (a) Dark-field images of the SHSY5Ycells incubated with iron (specifically ferric ammonium nitrate) for 1 h. The areas mapped with HSI are marked by colored boxes. (b) Spectral profiles collected from each region shown in (a). Peaks are mainly observed between 450 to 650 nm. All bulk iron areas have a peak absorbance near 600 nm, whereas the peak signal in the cells is near 500 nm and the signal on the glass plate is almost zero. (c) Magnified or zoomed-in images (17×) of pixels containing HSI data from glass, cytoplasm, nucleus, and bulk iron.

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: Fluorescence, Imaging, Incubation

Journal: Journal of Biomechanical Engineering

Article Title: Single-Cell Analysis Using Hyperspectral Imaging Modalities

doi: 10.1115/1.4038638

Figure Lengend Snippet: Label-free dark-field HSI of human RBCs from Conti et al. [13]: (a) Hyperspectral image of erythrocyte sample. (b) Spectral library composed of different endmembers with random color code. (c) Zoomed-in image of one erythrocyte. (d) Color mapping matching spectra of the spectral library. (e) Mapping five main components of RBCs namely phospholipid, cholesterol, hemoglobin, spectrin, and protoporphyrin. Briefly, 5 μl of whole blood was loaded in the center of glass slide and sandwiched with coverslip. After 120 min, to allow for image stability, the optical acquisition was started. Each image consisted of approximately 30 regularly shaped RBC as shown in (a) with no other cells. One RBC was chosen as shown in (c), (d), and (e) for further image analysis. For the RBCs, eight spectra (b) were individuated with optimal coverage of the optical image (d). Applying the SAM function, the spectral distribution of the 8 endmember spectra in the samples was then determined (data not shown) as described in Conti et al. [13]. This study demonstrated a fast, easy, and repeatable protocol to study large number of cells and to the possibility of mapping single molecules, proteins as well as structure of cell membranes with applications in personalized medicine and membrane-targeted therapies [13]. The caption text refers to online color version of the figure.

Article Snippet: Parkinson's disease , Cellular iron , Neuroblastoma dopaminergic cells (SHSY5Y) , CytoViva Hyperspectral Imager , CytoViva Imaging, t-test SPSS19 , 600 nm , Near 500 nm , 40–90 nm , No external labelInternal label- Fe 2+ to Fe 3+ , [ 18 ].

Techniques: