NMR and MRI physics · Introductory

T1 recovery (longitudinal)

Written by SignalsLab

Early edition: these lessons are being reviewed.

Content

Summary

T1 recovery describes how longitudinal magnetization regrows toward its equilibrium value M0 after a radiofrequency pulse disturbs it. The regrowth is exponential with time constant T1, which depends on tissue type and field strength. Differences in T1 are the basis of T1-weighted MRI contrast.

Intuition

Think of the spins as a population of tiny magnets aligned with a strong static field. An RF pulse knocks some of that alignment away, so the longitudinal component shrinks or even flips. The spins then exchange energy with their surroundings, called the lattice, and slowly return to alignment. The rate of return is proportional to how far the system is from equilibrium: far away means fast change, close means slow change. That is the signature of an exponential. T1 is the time for about 63 percent of the missing magnetization to return. Fat has a short T1 and recovers quickly, so it looks bright on T1-weighted images. Water has a long T1 and recovers slowly, so it looks dark.

Key equations

dMzdt=−Mz−M0T1
(1)

Longitudinal relaxation part of the Bloch equation: the rate of change is proportional to the deviation from equilibrium.

Mz(t)=M0−(M0−Mz(0))e−t/T1
(2)

General solution for recovery from any starting value M_z(0).

Mz(t)=M0(1−e−t/T1)
(3)

Recovery after a 90 degree pulse, starting from M_z(0)=0.

Mz(t)=M0(1−2e−t/T1)
(4)

Recovery after a 180 degree inversion pulse, starting from M_z(0)=-M_0.

Derivation

  1. Start with the longitudinal Bloch equation: dMzdt=−(Mz−M0)T1.
  2. Define the deviation D(t)=Mz(t)−M0. Then dDdt=−DT1.
  3. Separate variables: dD/D = -dt/T1, and integrate to get ln D = -t/T1 + constant.
  4. Exponentiate: D(t) = D(0) e^{-t/T1}.
  5. Substitute back: Mz(t)=M0+(Mz(0)−M0)e−t/T1.
  6. For a 90 degree pulse Mz(0)=0, giving Mz=M0(1−e−t/T1).
  7. For a 180 degree pulse Mz(0)=−M0, giving Mz=M0(1−2e−t/T1), which crosses zero at t=T1ln⁡2, about 0.69 T1.

Worked example

Problem

A tissue has T1=800 ms. After a 90 degree pulse, what fraction of M0 has recovered at t=800 ms and at t=1600 ms? Also, after a 180 degree inversion pulse, when is the null point?

  1. Use Mz/M0=1−e−t/T1.
  2. At t=800 ms, t/T1=1, so Mz/M0=1−e−1=1−0.368=0.632.
  3. At t=1600 ms, t/T1=2, so Mz/M0=1−e−2=1−0.135=0.865.
  4. For inversion, set 1 - 2e^{-t/T1} = 0, so e^{-t/T1} = 0.5.
  5. Then t = T1 ln 2 = 800 x 0.693 = 554 ms.

Answer 63% recovered at 800 ms, about 86.5% at 1600 ms, and the inversion null point is about 554 ms.

Common mistakes

  • Confusing T1 with T2: T1 is longitudinal recovery from energy exchange with the lattice, while T2 is transverse decay from dephasing; T2 is never longer than T1.
  • Thinking full recovery happens at t = T1; only 63% has recovered, and about 5 T1 is needed for over 99%.
  • Using the 90 degree formula after an inversion pulse and forgetting the factor of 2.
  • Assuming T1 is the same in all tissues or at all field strengths; T1 lengthens as field strength increases.
  • Believing long T1 tissues look bright on T1-weighted images; they look dark because they recover less between pulses.

Animation

Animation coming soon

The animated version of this lesson is being prepared for the web. In the Signals AI app, lessons are drawn step by step.

Code

scene.py(66 lines)
import sys
import numpy as np
from manim import *
from blocks import *

M0 = 1.0
T1 = 1.0


def mz(t, t1=T1, m0=M0):
    """Longitudinal recovery after a 90-degree pulse: Mz = M0 (1 - exp(-t/T1))."""
    return m0 * (1.0 - np.exp(-t / t1))


def milestones():
    """Computed fractions of M0 recovered at t = T1, 2*T1, 3*T1."""
    return [(k, float(mz(k * T1))) for k in (1, 2, 3)]


class Main(Scene):
    def construct(self):
        self.camera.background_color = "#101418"
        # Title
        TitleCard(self, "T1 recovery", "Longitudinal magnetisation")

        # Curve plot, shifted left to leave room for text
        plot = CurvePlot(mz, x_range=(0, 5, 1), y_range=(0, 1.2, 0.5),
                         x_label="t / T1", y_label="Mz / M0",
                         x_length=7.0, y_length=3.8, color=C_MAIN)
        # x axis in units of T1 (T1 = 1)
        plot.shift(LEFT * 2.2 + DOWN * 0.2)
        self.play(Create(plot.axes), FadeIn(plot.labels))
        self.play(Create(plot.curve), run_time=2)

        # Equation
        eq = MathTex(r"M_z=M_0\left(1-e^{-t/T_1}\right)", font_size=32)
        eq.to_edge(RIGHT, buff=0.5).shift(UP * 1.8)
        self.play(Write(eq))

        # Milestone dots from computed values
        rows = VGroup()
        for k, v in milestones():
            d = plot.dot_at(k)
            d.move_to(plot.axes.c2p(k, v))
            line = DashedLine(plot.axes.c2p(k, 0), plot.axes.c2p(k, v),
                              color=C_ACCENT, stroke_width=2)
            lab = Text(f"t={k}T1: {100 * v:.0f}%", font_size=26, color=C_ACCENT)
            rows.add(lab)
            rows.arrange(DOWN, aligned_edge=LEFT, buff=0.3)
            rows.next_to(eq, DOWN, buff=0.5).align_to(eq, LEFT)
            self.play(Create(line), FadeIn(d), FadeIn(lab), run_time=0.8)

        StepCaption(self, "After about 3 T1, Mz has recovered ~95% of M0.")
        self.wait(1.5)


if __name__ == "__main__" and "--selftest" in sys.argv:
    assert abs(mz(0.0)) < 1e-12
    assert abs(mz(T1) - (1 - np.exp(-1))) < 1e-12
    ms = dict(milestones())
    assert abs(ms[1] - 0.632) < 1e-3
    assert abs(ms[2] - 0.865) < 1e-3
    assert abs(ms[3] - 0.950) < 1e-3
    ts = np.linspace(0, 5, 50)
    assert np.all(np.diff(mz(ts)) > 0) and mz(50.0) > 0.999
    print("ok")

The Manim scene code that draws this lesson's animation in the Signals AI app (Python, Manim Community).

ccBYNC© 2026 SignalsLab. Lesson text shared as is under CC BY-NC 4.0.