Turn time-to-event data into a publication-ready survival curve.
Upload follow-up times, event status, and groups from CSV or Excel. SciFigure computes the Kaplan-Meier estimates, median survival, number-at-risk table, and log-rank test before generating a clean figure for your manuscript.
Upload survival data and describe the curve you need.
Costs 5 credits. Survival estimates, medians, and log-rank statistics are computed from your uploaded data, then drawn as a PNG.
预览
宽高比: 16:9
These are sample previews. Upload your dataset to generate your own Kaplan-Meier plot.
Use these examples as starting points for clinical trial arms, tumor grade comparisons, animal survival studies, and relapse-free survival figures with risk tables.
提示词
Based on the uploaded two_arm_trial.csv file, create a Kaplan-Meier survival plot of Months by Arm. Draw step curves with vertical censor ticks, print the log-rank p-value, list median survival for Control and Treatment, and add a number-at-risk table below the x-axis.
提示词
Based on the uploaded tumor_grade_survival.xlsx file, create a Kaplan-Meier overall survival plot for tumor Grade I, II, and III. Status is coded 2 = death and 1 = censored. Use an ordered color scale, print the three-group log-rank p-value, and add a risk table.
提示词
Based on the uploaded mouse_survival.csv file, create a Kaplan-Meier survival plot of Day by Genotype with shaded 95% confidence bands, survival shown in percent, censor marks, dashed median survival lines, and the log-rank p-value.
提示词
Based on the uploaded relapse_free_survival.tsv file, create a Kaplan-Meier relapse-free survival plot for low, intermediate, and high risk groups. Show landmark survival at 12 and 24 months, a number-at-risk table with group colors, and a compact legend with n per group.
提示词
Based on the uploaded dose_escalation_survival.csv file, create a Kaplan-Meier progression-free survival plot of Weeks by DoseArm for placebo, 5 mg, 10 mg, and 20 mg, with Event coded 1 = progression and 0 = censored. Use an ordered light-to-dark blue scale, add censor ticks, print the four-group log-rank p-value together with the log-rank test for trend across dose, list median PFS per arm, and put a number-at-risk table with n per arm under the x-axis.
提示词
Based on the uploaded transplant_competing_risks.xlsx file, create a cumulative incidence figure of Months after transplant. EventType is coded 0 = censored, 1 = relapse, 2 = non-relapse mortality. Draw one rising step curve per event type on a 0 to 1 axis, print the 24-month cumulative incidence with its 95% CI for each, add censor ticks and a shared number-at-risk table, and note in a footnote that these are 1 - S(t) cause-specific curves rather than Fine-Gray subdistribution estimates.
Add CSV, TSV, TXT, or Excel data with one row per subject: a numeric follow-up time, an event column (1 = event, 0 = censored), and an optional group or treatment arm.
Ask for confidence bands, percent scale, landmark times, median lines, or a cumulative incidence curve, and say which time unit to show on the x-axis.
SciFigure computes the survival estimates and log-rank test, draws the curves with a number-at-risk table, and lets you download a PNG for papers and slides.
It estimates the survival function for each group in your time-to-event table and draws the result as a stepped survival curve. Every drop in the curve happens at an observed event time, censored subjects appear as small tick marks, and the figure can include median survival, a log-rank p-value, and a number-at-risk table aligned with the time axis. All of these numbers are calculated from your uploaded rows before the image is generated, so the Kaplan-Meier plot generator works as a survival curve generator you can check against your statistics software rather than an AI sketch.
Use one row per subject with a numeric time column, a numeric event column (1 = event such as death or relapse, 0 = censored), and an optional group column such as treatment arm, genotype, or tumor grade. Up to 6 groups are supported.
The tool calculates them. The product-limit estimates, censor positions, median survival, confidence limits, number at risk, and log-rank test are computed from your uploaded values in code, then passed to the image model as exact values to draw. The image model handles layout and styling only, so review the final figure against the listed numbers before publishing.
For 2 or more groups, the generator sums observed and expected events at every event time, builds the variance matrix, and reports the chi-square statistic, degrees of freedom (groups minus 1), and p-value. It does not adjust for covariates; use Cox regression in your own analysis for that.
Yes. The number-at-risk table is included by default at about six evenly spaced time points. Ask for "95% confidence bands" to add Greenwood log-log intervals, or write "no risk table" to leave the table out.
Yes. Ask for a "cumulative incidence" or "1 - survival" curve and the values are converted to 1 - S(t). This is the complement of the Kaplan-Meier curve, not a competing-risks estimate.
Yes, after you check the event coding, group sizes, median survival, and p-value against your analysis plan. State the time origin, censoring rules, and test used in the figure legend.
Generate Kaplan-Meier plots, forest plots, ROC curves, and other research-ready visualizations with SciFigure AI.
