The main goal of this lab report is to provide a comprehensive comparative analysis of confocal and two-photon microscopy for depth-resolved imaging of YG Invitrogen fluorescent beads, Convallaria rhizome, and Triple-Negative Breast Cancer (TNBC) cells embedded in a collagen matrix. In particular, we aim to determine lateral and axial resolution of the confocal microscope using 0.5 𝜇m fluorescent beads with varying pinhole sizes. Additionally, we aim to perform 3D reconstruction of the Convallaria rhizome to understand the optical sectioning capabilities of the confocal microscope. Another aim is to measure the optical redox ratio of cells in two-photon imaging modality in order to compare and contrast results of metabolic analysis performed with two-photon and confocal microscopy. Finally, we aim to compare continuous wave (CW) and pulsed laser depth penetration, using TNBC cells and calculating respective Signal-to-Background ratios (SBR). One of the key findings include 1 Airy Unit (AU) to be the most accurate lateral resolution measurement for the 0.5 𝜇m fluorescent beads. We also found that image sharpness and contrast decreased with increasing imaging depth in the Convallaria rhizome, likely due to scattering and absorption properties of the sample. Finally, the expected result of two-photon imaging providing better depth penetration was not observed in deep imaging.

Introduction

Biomedical imaging plays a critical role in medical diagnostics and research by enabling the visualization of biological structures and processes with high spatial resolution. However, conventional imaging techniques often suffer from limitations in depth penetration, contrast, and optical sectioning, making it difficult to obtain high-quality images from thick biological samples. Depth-resolved imaging addresses these challenges by providing improved axial resolution, enabling three-dimensional visualization of structures in tissues and cells. This capability is particularly beneficial for applications such as neuroscience, cancer research, and regenerative medicine, where understanding cellular microenvironments and tissue architecture is essential4,2.

The development of depth-resolved imaging has evolved over several decades, beginning with the advent of confocal microscopy in the 1950s. Marvin Minsky first conceptualized the confocal microscope in 1957 to improve optical sectioning by eliminating out-of-focus light, but it was not until the 1980s that advances in laser technology and computing allowed for practical applications3. Around the same time, multiphoton excitation microscopy emerged as a breakthrough technique, pioneered by Denk et al. in 1990. This method allowed for deeper imaging in biological tissues while reducing phototoxicity, making it particularly useful for live-cell imaging and in vivo studies1.

Among the key modalities of depth-resolved imaging, confocal microscopy and multiphoton excitation microscopy offer distinct advantages. Confocal microscopy enhances optical sectioning by using a pinhole to reject out-of-focus light, thereby improving contrast and resolution in thick specimens5. However, its depth penetration is limited due to light scattering and absorption. Multiphoton excitation microscopy, on the other hand, uses nonlinear optical interactions to achieve fluorescence excitation at greater depths while reducing phototoxicity and photobleaching 1. This technique allows for improved imaging of live tissues, benefiting from longer excitation wavelengths that scatter less in biological media. Key sources of contrast in these modalities include fluorescence from endogenous or exogenous labels, second-harmonic generation (SHG), and metabolic imaging based on fluorescence lifetime or redox ratio measurements6.

This report details the methodology and findings from depth-resolved imaging experiments using confocal and multiphoton microscopy. The samples examined include Invitrogen YG fluorescent beads (0.5 µm) to assess resolution, Convallaria rhizome sections stained with acridine orange to evaluate optical sectioning, and human-induced neural stem cells (hiNSCs) in a 3D bioengineered brain model to investigate metabolic imaging. The imaging methods employed include z-stack acquisition, stitched imaging, second-harmonic generation, and fluorescence-based metabolic analysis. By comparing the performance of confocal and multiphoton microscopy, this study highlights the strengths and limitations of each technique in different biological contexts. The results provide insights into the optimal imaging approaches for specific biomedical applications, particularly in cases requiring deep tissue penetration and reduced phototoxicity.

Background

In this lab report, we will focus on the principles of confocal microscopy, including a comparative analysis of two-photon and single-photon excitation modalities. A confocal microscope consists of key components such as a laser light source, an objective lens, a scanning system, a dichroic mirror or beam splitter, a pinhole, a detector (e.g., a photomultiplier tube), and specialized software for image processing.

When observing an optically turbid sample, such as a milk-water mixture, with a conventional microscope, the image appears hazy and opaque due to extensive light scattering. In a standard widefield microscope, illumination occurs across the entire sample, allowing both in-focus and out-of-focus light to reach the detector. This results in a blurred, low-contrast image where internal structures are difficult to distinguish.

Confocal microscopy overcomes this limitation by using a laser as a point light source instead of illuminating the entire sample. The laser selectively excites a small region, and the emitted fluorescence is collected by the objective lens. To enhance image clarity, the light passes through a pinhole aperture, which blocks out-of-focus light, ensuring that only in-focus signals reach the detector. For depth resolution and 3D reconstruction, the laser scans the sample in a point-by-point zigzag pattern. At each point, the objective lens collects light, which is filtered through the pinhole to isolate focal plane signals. The final image is reconstructed based on the intensity data measured at each pixel by the detector.

Figure 1: Energy diagrams for single photon (left), and two-photon (right) excitation processes.

Figure 1 illustrates the energy level dynamics of two-photon excitation compared to single-photon excitation microscopy. In single-photon excitation, a high-energy photon excites the fluorophore, followed by fluorescence emission at a lower wavelength corresponding to the energy difference between the ground and virtual states. In two-photon excitation, two lower-energy photons interact simultaneously to excite the fluorophore to the same excited state as in single-photon excitation.

The two-photon excited fluorescence (TPEF) action cross-section quantifies the probability of two photons being absorbed simultaneously, leading to fluorescence emission. Measured in Göeppert-Mayer (GM) units, this probability is significantly lower than in single-photon excitation. To compensate, a high instantaneous photon flux is required, necessitating specialized laser illumination such as femtosecond pulsed lasers. These lasers deliver short, high-intensity pulses with low average power—unachievable with continuous-wave lasers—minimizing photodamage to the sample.

Methods

Images were acquired using an inverted multiphoton microscope TCS MP8 (Leica Microsystems, Deerfield, Illinois), which is integrated into a confocal software interface. The microscope is configured for white, ultraviolet, fluorescence, laser, and femtosecond laser illumination. The microscope was equipped with 10X, 20X, 25X, and 40X objectives, infrared excitation lasers, HyD detectors, and an infrared-optimized optical transmission. Backwards detection, non-descanned detectors were used, with included bandwidths of detectors PMT NDD 1 (500-550 nm), PMT NDD 2 (435-485 nm), PMT NDD 3 (604-644 nm), and PMT NDD 4 (418-442 nm).

Two TCS MP8 microscopes were used during this lab, one for the confocal imaging and the other for the multiphoton imaging. This was due to the microscopes having different detectors set up, but both microscopes were the same model and used the same software. The confocal microscope had lasers set to detect UV light (405 nm) and visible light (488 nm, 552 nm, and 638 nm). Lasers on the multiphoton microscope were set to include ultraviolet (405 nm), visible (488 nm, 514 nm, 552 nm, and 638 nm), and multiphoton (variable; 755 nm and 860 nm) light sources.

We used the included LasX software, which facilitates objective selection, exposure control, detector selection, pinhole size, pixel dwell time, and line and frame averaging. Focus and image-acquisition were aided through the use of a joystick and dials to adjust the x-, y-, and z-planes. All images were stored in the Biophotonics2025 directory, categorized by magnification and modality, and later analyzed using the Fiji/ImageJ image processing package. Three samples were imaged for this lab: 0.5 µm fluorescent beads for confocal imaging; Convallaria rhizome stained with acridine orange for confocal imaging; and human-induced neural stem cells (hiNSCs) in a 3D bioengineered brain model of a porous silk scaffold infused with a collagen gel for two-photon and SHG images.

For the first confocal images, the 0.5 µm fluorescent bead slide was placed facing downward, as trans-illumination microscopy requires the sample and coverslip to be oriented toward the objective. A drop of water was added to the slide to allow for the use of a 40X water immersion objective. To view the beads, the 488 nm laser was selected with a detection range of 498-588 nm and a gain of 419 nm. The pinhole size was set to 1 airy unit (AU). Frame averaging was set to 1 and then 8, and an image of the beads was captured for each frame average for illustrative purposes. Next, three z-stack images were captured with the following parameters: pinhole size of 1 AU; pinhole size of 0.5 AU; and pinhole size of 7.0 AU. For all three images, the frame averaging was kept at 8, and the step size was set to 0.25 µm. Analysis of this step included using the z-stack images to compare the bead resolution with the different pinhole sizes.

To enhance visualization of Convallaria rhizome, a confocal set-up with two-color imaging was used. The lasers used in this step were 405 nm (UV light), 488 nm (blue), 552 nm (green), and 638 nm (red). Using a 10X objective, a pinhole size of 1 AU, and a step size of 4.28 µm, a z-stack image was captured of the Convallaria. During analysis, this z-stack was used to create a 3D reconstruction of the rhizome to look at optical sectioning. A stitched image was also captured with the 10X objective, but notably, we deviated from the lab instructions and did not capture a stitched image at 40X due to difficulties detecting the image at that magnification with the laser setting previously chosen.

Two-photon imaging was next used to image hiNSCs in a collagen/silk scaffold. During this step, the temperature control on the microscope was set to 35°C to keep the samples warm while on the microscope stage. While in multiphoton mode, the microscope was set to use a 12 bit depth and enable line averaging during live image acquisition. Bi-directional scanning was set at 600 Hz and the image size was 1024x1024 with no pixel binning. The 40X 1.1 W objective was used. Once the sample was located on the slide, NDD detectors 1 and 2 (NADH and FAD, respectively) were enabled and a cell was viewed using 860 nm illumination. An image of the cell was captured using the 860 nm illumination, and then without moving the slide, another image was captured using 755 nm illumination. This process was repeated twice more on two different areas of the slide to capture a total of three cells. For analysis, cell images were provided to us from a previous class. They included shallow MP 860 nm, shallow MP 755 nm, deep MP 860 nm, deep MP 755 nm, shallow confocal, and deep confocal. The optical redox ratio was calculated using these images and the equation [FAD/(FAD+NADH)].

Additionally, images were provided to us for analysis that were of triple negative breast cancer (TNBC) cells in a collagen matrix for use in CW versus pulse laser depth penetration analysis. Protocol for imaging these TNBCs would have included capturing z-stack images with the multiphoton mode. For this step, the bit depth would have been set to 12 bit, and live averaging during live acquisition would be enabled. The operator would use the 40X 1.1 W objective, bi-directional scanning at 600 Hz, and an image size of 1024 x 1024. The confocal laser would be on, NDD detectors 1 and 4 enabled, and the MP laser set to 976 nm. Three regions of cells would be located and captured to collect fluorescence data. Next, the MP laser and NDD detectors would be turned off, and each z-stack image would be repeated using the 488 nm CW laser and descanned detectors with the pinhole at 1 AU. PMT 1 would be set to collect SHD, and HyD 1 would be set to collect NADH fluorescence. The images that were given to us from this method were entered into ImageJ for analysis using the signal to background ratio of a cell and cell-free region.

Results and Analysis

Resolution

In order to measure the resolution of the beads using confocal microscopy, three beads were selected and had their full-width half-maximum (FWHM) measurements calculated through the ImageJ software. This was done by drawing a line across the center of the bead in the lateral plane, and plotting the intensity profiles. Each intensity profile was fitted with a Gaussian fit function, which gives an output of a d value, or the smallest resolvable feature size, which can be multiplied by 2.355 to calculate the FWHM (Table 1). A similar process was repeated for calculating the axial FWHM, except a z-stack image of each bead was used to calculate the resolution instead of a single-layer image (Table 2). The beads had been imaged at pinhole sizes of 0.5 AU, 1 AU, and 7 AU, and the same three beads were measured during analysis for consistency.

Table 1. Lateral FWHM calculations for three selected beads at each of the observed pinhole sizes are shown. The table shows the measured d value from the Image J software, as calculated by the Gaussian fit function, in the third column. d was multiplied by 2.355 to get the FWHM size, in micrometers, as shown in the fourth column. In the fifth column, micrometers were translated to nanometers to give the bead diameter in nanometers. The FWHM had average and standard deviations calculated for each pinhole size.

Table 2. Axial FWHM calculations for three selected beads at each of the observed pinhole sizes are shown. The table shows the measured d value from the Image J software, as calculated by the Gaussian fit function, in the third column. d was multiplied by 2.355 to get the FWHM size, in micrometers, as shown in the fourth column. In the fifth column, micrometers were translated to nanometers to give the bead diameter in nanometers. The FWHM had average and standard deviations calculated for each pinhole size.

The lateral FWHM was calculated to be 469 nm for the 0.5 AU condition, 492 nm for the 1 AU condition, and 547 nm for the 7 AU condition. The axial FWHM was calculated to be 5221 nm for the 0.5 AU condition, 4757 nm for the 1 AU condition, and 2267 nm for the 7 AU condition. We know the beads have a diameter, or lateral FWHM of 500 nm, so we can compare that value to the measurements taken in this lab. The results indicate that a 1 AU pinhole gives the best resolution and most accurate measurement. A pinhole size too large (7 AU) and the resolution changes so that there is room for more airy units, or rings of light that are seen around the bead and thus measured. A pinhole size too small (0.5 AU) may not let enough light in to hit the entirety of the object, so the entire bead may not be in view. While the averages for 0.5 AU and 7 AU conditions are similar to the known 500 nm bead size, there is enough of a difference that it might have some effects on analysis, and highlights the importance of choosing the ideal pinhole size. From Figure 2, we can see that lateral resolution slightly increases with increased pinhole size, while the axial resolution decreases with increased pinhole size.

Figure 2. Bar charts displaying the lateral (left) and axial (right) resolution of the confocal microscope at varying pinhole sizes. Three beads were measured at each pinhole size of 0.5 AU, 1 AU, and 7 AU for both lateral and axial resolution. The lateral resolution ranged from 463-560 nm, and the axial resolution ranged from 2178-5787 nm.

When using the brightfield microscope, the beads were calculated to have an average lateral FWHM of 171 and an axial FWHM of 2729 nm. Immediately, we can see that using a confocal microscope significantly improved the resolution of the images so that we are able to obtain a much more accurate lateral diameter of 492 nm at 1 AU pinhole size. The axial resolution is also improved, and we are able to measure a greater depth of the object. The axial resolution of the brightfield microscope is about equal to the resolution of the confocal microscope with a pinhole size of 7 AU.

Figure 3. Above, 0.5 µm YG Invitrogen fluorescent beads were observed through a confocal microscope through three different pinhole size conditions: (A) 0.5 AU, (B) 1 AU, and (C) 7 AU. The beads are the most visible in condition B, while the least visible in condition C. Each image was captured at a step size of 0.25 µm and a frame averaging of 8.

When looking at the images of the beads with the naked eye, it is also relatively clear that the 1 AU pinhole size provides the best resolution (Figure 3B). The beads are visible and have clearly defined borders. While the beads observed at pinhole size 0.5 AU (Figure 3A) are less bright and a bit harder to detect, they do still also have clearly defined shapes. In both of these conditions, bead clusters can be identified, but the individual beads can still be distinguished. In the 7 AU pinhole condition (Figure 3C), the beads are very difficult to locate without zooming in on the image. The borders of the beads are very fuzzy, and the bead clusters look more like blobs.

Optical Sectioning

Figure 4. Z-stack images of Convallaria rhizome at different depths (Z-Slices 0, 2, 4, 6, and 8). Each panel is labeled (A-E) in the top-left corner. The variation in image contrast and sharpness across depths highlights the effects of optical sectioning, with shallower slices appearing clearer and deeper slices exhibiting increased blurring and reduced signal quality.

Figure 5. (Left) 3D reconstruction of the Convallaria rhizome using optical sectioning, where high-intensity regions indicate well-resolved structures across different z-sections. The Z-axis represents optical depth in slices. (Right) Quantitative analysis of imaging depth on image quality, showing a decline in contrast (blue) and sharpness (red) as depth increases. This demonstrates how deeper imaging leads to reduced clarity due to light scattering and signal attenuation.

Optical sectioning plays a crucial role in capturing high-resolution images at varying depths within a specimen. Factors such as depth, resolution, and step size significantly influence the quality of the resulting images. In this experiment, a z-stack of Convallaria rhizome sections was acquired, allowing for the analysis of how imaging depth impacts contrast and sharpness.

From the individual z-sections, a noticeable variation in image clarity was observed as the imaging depth increased. The shallower slices exhibited higher contrast and sharper features, while deeper slices appeared progressively more blurred. This trend is quantified through the analysis of contrast (measured as the standard deviation of pixel intensities) and sharpness (determined using the Sobel edge detection filter). A plot of these values against imaging depth shows a clear decline in both contrast and sharpness as depth increases. The reduction in contrast is likely due to light scattering and absorption effects, which decrease the signal-to-noise ratio at greater depths. Similarly, the decline in sharpness suggests that out-of-focus contributions from adjacent planes reduce the ability to resolve fine structures.

A 3D reconstruction of the z-stack further illustrates the effects of optical sectioning, highlighting high-intensity regions that represent well-resolved structures. This visualization supports the conclusion that optical sectioning is most effective at shallower depths, where resolution is highest. As depth increases, the quality of imaging degrades due to reduced light penetration and increased interference from surrounding tissue layers.

Additionally, a quantitative analysis of image quality shows that both contrast and sharpness decline with increasing depth. This was demonstrated through a plot comparing contrast (measured as intensity standard deviation) and sharpness (measured using the Sobel filter) across different z-slices. The results confirm that image degradation at greater depths is a result of optical blur and signal attenuation, reinforcing the need for optimized imaging parameters.

Metabolic Analysis

Figure 6. The optical redox ratio of cells imaged with shallow MP. Cells were imaged using 860 nm and 755 nm lasers to detect levels of FAD and NADH, respectively. The images were analyzed using Fiji ImageJ to give an outputted redox ratio as shown in this figure. Three cells can be seen, as designated by blue-green blobs in the top section and bottom right corner of the image. A contrast key is displayed in the top right corner.

Figure 7. The optical redox ratio of cells imaged with shallow MP. Cells were imaged using 860 nm and 755 nm lasers to detect levels of FAD and NADH, respectively. The images were analyzed using Fiji ImageJ to give an outputted redox ratio as shown in this figure. This image was processed in the same manner as Figure 6, except without excluding the non-cell areas. Lots of noise can be seen throughout the image. Three cells are identified, as designated by blue-green blobs in the top section and bottom right corner of the image. A contrast key is displayed in the top right corner.

Excluding the non-cell areas removes a significant amount of noise and makes it easier to distinguish the cells from the background. In Figure 7, the cells are barely discernible from the background, and are only identifiable by a more consistent level of contrast rather than a varied mix of contrasts.

Figure 8. The optical redox ratio of cells imaged with shallow confocal microscopy. Cells were imaged using 860 nm and 755 nm lasers to detect levels of FAD and NADH, respectively. The images were processed and analyzed in Fiji (ImageJ) to generate the redox ratio map displayed in this figure. Three cells are visible, appearing as blue-green blobs in the top section and bottom right corner of the image. A contrast key indicating the redox ratio scale is shown in the top right corner, and a scale bar of 20 μm is provided at the bottom.

Figure 9. The optical redox ratio of cells imaged with deep MP. Cells were imaged using an 860 nm laser to detect levels of FAD and NADH. The images were analyzed using Fiji ImageJ to generate the redox ratio map displayed in this figure. Two cells can be seen, appearing as yellow-green blobs in the center and lower right portion of the image. A contrast key indicating the redox ratio scale is displayed in the top right corner, and a scale bar of 20 μm is provided at the bottom.

Figure 10. The optical redox ratio of cells imaged with deep MP. Cells were imaged using a 755 nm laser to detect levels of FAD and NADH. The images were analyzed using Fiji ImageJ to generate the redox ratio map displayed in this figure. Two cells can be seen, appearing as white blobs in the center and lower right portion of the image. A contrast key indicating the redox ratio scale is displayed in the top right corner, and a scale bar of 20 μm is provided at the bottom.

Figure 11. The optical redox ratio of cells imaged with deep confocal microscopy. Cells were imaged using 755 nm and 860 nm lasers to detect levels of FADand NADH, respectively. The redox ratio [FAD / (FAD + NADH)] was calculated and visualized using a heat map, with the color scale ranging from low (blue, NADH-dominant) to high (red, FAD-dominant) metabolic activity. A scale bar of 20 µm is included in the bottom right corner for reference.

CW vs Pulse Laser Depth Penetration

Figure 12 presents a comparative analysis of two-photon excitation (976 nm) and single-photon continuous wave (CW) excitation (488 nm) for imaging Triple Negative Breast Cancer (TNBC) cells embedded in a collagen matrix at both shallow and deep depths. The images were acquired at depth positions 1873 (shallow) and 1927 (deep), with a step size of approximately 4.28 µm. Imaging was conducted using a 1024 × 1024 pixel resolution, a 16-bit depth, a 40× 1.1W water-immersion objective, and a bi-directional scanning speed of 600 MHz. This comparison aims to assess the effectiveness of two-photon excitation in reducing background fluorescence and preserving contrast at increasing depths within the collagen matrix.

To find the SBR (signal-to-background ratio) we used a formula:

SBR=Mean Signal IntensityMean Background Intensity

The shallow imaging modality SBR with a two-photon excitation was measured as 35.60% and with a one-photon excitation as 51.70%. We were not able to measure the SBR for deep imaging modality since the cells were not visually present on the pictures.

Figure 12: (A, B) Shallow depth imaging of Triple Negative Breast Cancer (TNBC) cells in a collagen matrix using two-photon excitation at 976 nm (A, red) and continuous wave (CW) excitation at 488 nm (B, green), with measured signal-to-background ratios (SBR) of 35.60% and 51.70%, respectively. (C, D) Deep depth imaging under the same conditions, where SBR could not be determined due to the absence of clearly identifiable cell regions. While two-photon excitation is theoretically expected to maintain better contrast and structural clarity at greater depths, both imaging modalities showed significant signal degradation at depth.

Conclusion

In conclusion, this laboratory practice provided us with knowledge on key modalities of depth-resolved imaging: confocal microscopy and multiphoton excitation microscopy, highlighting their respective strengths and weaknesses. Depth-resolved imaging is fundamental in medical diagnostics and research through enabling the visualization of biological structures and processes with high spatial resolution. Depth-resolved imaging addresses the challenges of limitations in depth penetration, contrast, and optical sectioning in conventional imaging techniques by providing improved axial resolution, enabling three-dimensional visualization of structures in tissues and cells.

Key findings of this lab included 1 Airy Unit (AU) to be the most accurate lateral resolution measurement for the 0.5 𝜇m fluorescent beads. We also found that image sharpness and contrast decreased with increasing imaging depth in the Convallaria rhizome, likely due to scattering and absorption properties of the sample. Finally, the expected result of two-photon imaging providing better depth penetration was not observed in deep imaging.

Confocal microscopy enhances optical sectioning by using a pinhole to reject out-of-focus light, thereby improving contrast and resolution in thick specimens. However, its depth penetration is limited due to light scattering and absorption. Multiphoton excitation microscopy, on the other hand, uses nonlinear optical interactions to achieve fluorescence excitation at greater depths while reducing phototoxicity and photobleaching.. This technique allows for improved imaging of live tissues, benefiting from longer excitation wavelengths that scatter less in biological media. Confocal microscopy would be ideal for imaging thin specimens and relatively shallow depth. Multiphoton techniques would be more beneficial for thicker specimens that need more depth penetration.

Author Contributions: KS and MZ were responsible for data collection in the lab. JRS performed data analysis based on the gathered results. KS, JRS, and MZ contributed equally to the conceptualization, methodology, and initial drafting of the manuscript. JRS and KS led the revision and editing process, while MZ coordinated the final submission. MS provided overall guidance, supervision, and project oversight. All authors reviewed and approved the final version of the report.

Funding: This work was supported by the Tufts University School of Engineering and made possible through the use of facilities at the Tufts Advanced Microscopic Imaging Center (TAMIC). We thank both institutions for providing access to microscopy equipment and resources as part of the BME-0156: Biophotonics Laboratory course.

Acknowledgments: We sincerely thank Maria Savvidou, Ph.D., Postdoctoral Researcher in the Department of Biomedical Engineering at Tufts University, for her guidance, mentorship, and support throughout the project, particularly in advising and assisting with lab data generation.

Conflicts of Interest: The authors declare no conflict of interest.

References

Denk, W., Strickler, J. H., & Webb, W. W. (1990). Two-photon laser scanning fluorescence microscopy. Science, 248(4951), 73-76.
Helmchen, F., & Denk, W. (2005). Deep tissue two-photon microscopy. Nature Methods, 2(12), 932-940.
Minsky, M. (1988). Memoir on inventing the confocal scanning microscope. Scanning, 10(4), 128-138.
Pawley, J. B. (2006). Handbook of biological confocal microscopy. Springer.
Sheppard, C. J. R., & Shotton, D. M. (1997). Confocal laser scanning microscopy. BIOS Scientific Publishers.
Zipfel, W. R., Williams, R. M., & Webb, W. W. (2003). Nonlinear magic: multiphoton microscopy in the biosciences. Nature Biotechnology, 21(11), 1369-1377.

Kyona Schacht, AG’25

School of Arts and Sciences, Tufts University

Jose Rodriguez Sanchez, E’26

School of Engineering, Tufts University

Malika Zakarina, EG’1G

School of Engineering, Tufts University

Maria Savvidou, Ph.D.

Senior Postdoctoral Researcher, Tufts University

Page updated

Google Sites

Report abuse