Review



horizontal puller p-1000 next generation micropipette puller  (Sutter Instrument Company)


Bioz Verified Symbol Sutter Instrument Company is a verified supplier
Bioz Manufacturer Symbol Sutter Instrument Company manufactures this product  
  • Logo
  • About
  • News
  • Press Release
  • Team
  • Advisors
  • Partners
  • Contact
  • Bioz Stars
  • Bioz vStars
    • 90
      Buy from Supplier

    Structured Review

    Sutter Instrument Company horizontal puller p-1000 next generation micropipette puller
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Horizontal Puller P 1000 Next Generation Micropipette Puller, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/horizontal puller p-1000 next generation micropipette puller/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    horizontal puller p-1000 next generation micropipette puller - by Bioz Stars, 2026-03
    90/100 stars

    Images

    1) Product Images from "Mapping dendritic spines using 2D two-photon laser scanning"

    Article Title: Mapping dendritic spines using 2D two-photon laser scanning

    Journal: bioRxiv

    doi: 10.1101/2025.07.10.664064

    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green horizontal lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Figure Legend Snippet: A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green horizontal lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.

    Techniques Used: Comparison

    A. Digital-to-analog square functions showing timing relationship between blue-light delivery (blue trace, top) and PMT gate (black trace, bottom). The PMT gate is triggered shortly before light onset and remains active for 10 ms to prevent photodamage. This results in a horizontal dark artifact in the image (shown in panel D). Scalebar: 5 ms. B. Schematic (left) and representative data (right) showing ChR2-evoked activity in vCA1 PNs. Top traces depict calcium signals in a responsive and a non-responsive spine following light stimulation. Individual trials are plotted in gray and the average trace in green. Bottom trace shows the corresponding somatic EPSC recorded via whole-cell patch clamp. Individual trials are plotted in gray and the average trace in red. Calcium trace scalebars: 1 s, 0.5 ΔF/F 0 . EPSC traces scalebars: 100 ms, 50 pA. C. Top: schematic of the scanner path and PMT gating during single z-layer imaging. Red rectangles indicate scanned sparse ROIs; curved arrows represent connecting travel (fly-to next ROI and fly-back to start position). Bottom: zoom-in of the PMT gate (black) activation just before scanning the ROI (as indicated by the partially black colored arrow). During this inactive period, blue light illuminates the sample (light blue line). Scalebar: 20 µm. D. PMT gating artifact correction. Left column: raw frames from before (n – 1), during (n), and after (n + 1) stimulation showing a segment of A594-filled dendrite. Middle column: row-wise average intensity for each frame. Gray lines represent the mean row intensity across all 50 frames, with shaded areas indicating the tolerance threshold of 3 σ . Pink traces correspond to the intensity profile of the current frame; rows exceeding the threshold are marked (*). Right column: corrected frame (n), where affected rows are interpolated using the corresponding rows from adjacent frames. Scalebar: 50 nm.
    Figure Legend Snippet: A. Digital-to-analog square functions showing timing relationship between blue-light delivery (blue trace, top) and PMT gate (black trace, bottom). The PMT gate is triggered shortly before light onset and remains active for 10 ms to prevent photodamage. This results in a horizontal dark artifact in the image (shown in panel D). Scalebar: 5 ms. B. Schematic (left) and representative data (right) showing ChR2-evoked activity in vCA1 PNs. Top traces depict calcium signals in a responsive and a non-responsive spine following light stimulation. Individual trials are plotted in gray and the average trace in green. Bottom trace shows the corresponding somatic EPSC recorded via whole-cell patch clamp. Individual trials are plotted in gray and the average trace in red. Calcium trace scalebars: 1 s, 0.5 ΔF/F 0 . EPSC traces scalebars: 100 ms, 50 pA. C. Top: schematic of the scanner path and PMT gating during single z-layer imaging. Red rectangles indicate scanned sparse ROIs; curved arrows represent connecting travel (fly-to next ROI and fly-back to start position). Bottom: zoom-in of the PMT gate (black) activation just before scanning the ROI (as indicated by the partially black colored arrow). During this inactive period, blue light illuminates the sample (light blue line). Scalebar: 20 µm. D. PMT gating artifact correction. Left column: raw frames from before (n – 1), during (n), and after (n + 1) stimulation showing a segment of A594-filled dendrite. Middle column: row-wise average intensity for each frame. Gray lines represent the mean row intensity across all 50 frames, with shaded areas indicating the tolerance threshold of 3 σ . Pink traces correspond to the intensity profile of the current frame; rows exceeding the threshold are marked (*). Right column: corrected frame (n), where affected rows are interpolated using the corresponding rows from adjacent frames. Scalebar: 50 nm.

    Techniques Used: Activity Assay, Patch Clamp, Imaging, Activation Assay



    Similar Products

    horizontal puller p-1000 next generation micropipette puller  (Sutter Instrument Company)
    90
    Sutter Instrument Company horizontal puller p-1000 next generation micropipette puller
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Horizontal Puller P 1000 Next Generation Micropipette Puller, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/horizontal puller p-1000 next generation micropipette puller/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    horizontal puller p-1000 next generation micropipette puller - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier
    p-2000 horizontal micropipette puller  (Sutter Instrument Company)
    90
    Sutter Instrument Company p-2000 horizontal micropipette puller
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    P 2000 Horizontal Micropipette Puller, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/p-2000 horizontal micropipette puller/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    p-2000 horizontal micropipette puller - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier
    horizontal micropipette puller p-1000  (Sutter Instrument Company)
    90
    Sutter Instrument Company horizontal micropipette puller p-1000
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Horizontal Micropipette Puller P 1000, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/horizontal micropipette puller p-1000/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    horizontal micropipette puller p-1000 - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier
    horizontal micropipette puller p-97  (Sutter Instrument Company)
    90
    Sutter Instrument Company horizontal micropipette puller p-97
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Horizontal Micropipette Puller P 97, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/horizontal micropipette puller p-97/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    horizontal micropipette puller p-97 - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier
    p-1000 horizontal micropipette puller  (Sutter Instrument Company)
    90
    Sutter Instrument Company p-1000 horizontal micropipette puller
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    P 1000 Horizontal Micropipette Puller, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/p-1000 horizontal micropipette puller/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    p-1000 horizontal micropipette puller - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier
    horizontal micropipette puller flaming- brown p- 97  (Sutter Instrument Company)
    90
    Sutter Instrument Company horizontal micropipette puller flaming- brown p- 97
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Horizontal Micropipette Puller Flaming Brown P 97, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/horizontal micropipette puller flaming- brown p- 97/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    horizontal micropipette puller flaming- brown p- 97 - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier
    horizontal micropipette puller  (Sutter Instrument Company)
    90
    Sutter Instrument Company horizontal micropipette puller
    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green <t>horizontal</t> lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.
    Horizontal Micropipette Puller, supplied by Sutter Instrument Company, used in various techniques. Bioz Stars score: 90/100, based on 1 PubMed citations. ZERO BIAS - scores, article reviews, protocol conditions and more
    https://www.bioz.com/result/horizontal micropipette puller/product/Sutter Instrument Company
    Average 90 stars, based on 1 article reviews
    horizontal micropipette puller - by Bioz Stars, 2026-03
    90/100 stars
    		  Buy from Supplier

    Image Search Results


    A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green horizontal lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.

    Journal: bioRxiv

    Article Title: Mapping dendritic spines using 2D two-photon laser scanning

    doi: 10.1101/2025.07.10.664064

    Figure Lengend Snippet: A. Model performance on test dataset. Left: representative ground truth z-scores for class 1 (red) and class 0 (gray) aligned to stimulation onset. Stacked bar plot indicates the class distribution of the test dataset. Center: F1 score as a function of the decision threshold on the test set. The maximum F1 value is indicated. Top-right: confusion matrix at the selected threshold, showing predicted vs. true labels. Bottom-right: detected z-scores are shown in gray and average in red. Scalebars: 1 s, 2 z-score units. B. Performance of threshold-based methods. Left: example traces with CNN model classification annotated in red. Stimulation is shown as a red vertical line. Gray shaded areas indicates the σ window. Blue and green horizontal lines indicate the threshold, with their length corresponding to the detection window. Blue and green annotations shown comparison with CNN classifier performance. TP: true-positive, TN: true-negative, FP: false-positive. Center and right: confusion matrices for detection using thresholds of 1.96 σ (left) and 2.5 σ (right). Thresholding at 2.5 σ reduces false positives but fails to detect most positive samples. Detected z-scores and average trace are shown below. Scalebars: 1 s, 2 z-score units. C. Logistic regression benchmark. Left: classifier output as a function of the logit (dot product of weights) is indicated with a black line. Overlayed dots are individual z-scores as classified by the regression model, colored by their ground truth annotation. Center: ROC curve with AUC. Top-right: confusion matrix using the decision boundary that maximizes the F1 score. Bottom-right: detected z-scores are shown in gray and average in violet. Scalebars: 1 s, 2 z-score units.

    Article Snippet: Patch pipettes were prepared using a horizontal puller (P-1000 Next Generation Micropipette Puller, Sutter Instrument) from borosilicate glass capillaries (Warner Instruments, LLC, Hamden, USA).

    Techniques: Comparison

    A. Digital-to-analog square functions showing timing relationship between blue-light delivery (blue trace, top) and PMT gate (black trace, bottom). The PMT gate is triggered shortly before light onset and remains active for 10 ms to prevent photodamage. This results in a horizontal dark artifact in the image (shown in panel D). Scalebar: 5 ms. B. Schematic (left) and representative data (right) showing ChR2-evoked activity in vCA1 PNs. Top traces depict calcium signals in a responsive and a non-responsive spine following light stimulation. Individual trials are plotted in gray and the average trace in green. Bottom trace shows the corresponding somatic EPSC recorded via whole-cell patch clamp. Individual trials are plotted in gray and the average trace in red. Calcium trace scalebars: 1 s, 0.5 ΔF/F 0 . EPSC traces scalebars: 100 ms, 50 pA. C. Top: schematic of the scanner path and PMT gating during single z-layer imaging. Red rectangles indicate scanned sparse ROIs; curved arrows represent connecting travel (fly-to next ROI and fly-back to start position). Bottom: zoom-in of the PMT gate (black) activation just before scanning the ROI (as indicated by the partially black colored arrow). During this inactive period, blue light illuminates the sample (light blue line). Scalebar: 20 µm. D. PMT gating artifact correction. Left column: raw frames from before (n – 1), during (n), and after (n + 1) stimulation showing a segment of A594-filled dendrite. Middle column: row-wise average intensity for each frame. Gray lines represent the mean row intensity across all 50 frames, with shaded areas indicating the tolerance threshold of 3 σ . Pink traces correspond to the intensity profile of the current frame; rows exceeding the threshold are marked (*). Right column: corrected frame (n), where affected rows are interpolated using the corresponding rows from adjacent frames. Scalebar: 50 nm.

    Journal: bioRxiv

    Article Title: Mapping dendritic spines using 2D two-photon laser scanning

    doi: 10.1101/2025.07.10.664064

    Figure Lengend Snippet: A. Digital-to-analog square functions showing timing relationship between blue-light delivery (blue trace, top) and PMT gate (black trace, bottom). The PMT gate is triggered shortly before light onset and remains active for 10 ms to prevent photodamage. This results in a horizontal dark artifact in the image (shown in panel D). Scalebar: 5 ms. B. Schematic (left) and representative data (right) showing ChR2-evoked activity in vCA1 PNs. Top traces depict calcium signals in a responsive and a non-responsive spine following light stimulation. Individual trials are plotted in gray and the average trace in green. Bottom trace shows the corresponding somatic EPSC recorded via whole-cell patch clamp. Individual trials are plotted in gray and the average trace in red. Calcium trace scalebars: 1 s, 0.5 ΔF/F 0 . EPSC traces scalebars: 100 ms, 50 pA. C. Top: schematic of the scanner path and PMT gating during single z-layer imaging. Red rectangles indicate scanned sparse ROIs; curved arrows represent connecting travel (fly-to next ROI and fly-back to start position). Bottom: zoom-in of the PMT gate (black) activation just before scanning the ROI (as indicated by the partially black colored arrow). During this inactive period, blue light illuminates the sample (light blue line). Scalebar: 20 µm. D. PMT gating artifact correction. Left column: raw frames from before (n – 1), during (n), and after (n + 1) stimulation showing a segment of A594-filled dendrite. Middle column: row-wise average intensity for each frame. Gray lines represent the mean row intensity across all 50 frames, with shaded areas indicating the tolerance threshold of 3 σ . Pink traces correspond to the intensity profile of the current frame; rows exceeding the threshold are marked (*). Right column: corrected frame (n), where affected rows are interpolated using the corresponding rows from adjacent frames. Scalebar: 50 nm.

    Article Snippet: Patch pipettes were prepared using a horizontal puller (P-1000 Next Generation Micropipette Puller, Sutter Instrument) from borosilicate glass capillaries (Warner Instruments, LLC, Hamden, USA).

    Techniques: Activity Assay, Patch Clamp, Imaging, Activation Assay