A precise fit in the wrong coordinates
An analyst fits a compact dipole to a clean MEG transient and obtains excellent agreement with the sensor fields. On the structural display, however, the location is implausible and one landmark is labeled on the wrong side. A colleague argues that the high fit percentage makes further checking unnecessary. The analyst recognizes that model fit and anatomical correspondence are separate quality questions. Sensor data, head position, digitized landmarks, MRI identity and transformation need review. Other source methods may also need consideration if the field is not well represented by one focal model. The result should remain a documented estimate with limitations rather than be reported as a definitive epileptogenic focus solely because a numerical fit is high.
1. Why does the fit not resolve the concern?
Read the worked answer
Sensor-space fit does not validate subject identity, landmark labeling or coregistration. Incorrect geometry can make the displayed anatomical location misleading despite a good fit.
2. What should the record preserve?
Read the worked answer
Retain raw data, coordinate information, model and preprocessing settings, identified error and correction history. The final clinical interpretation should integrate the corrected estimate and broader evidence.