Method of synthesizing a radiopharmaceutical

A controller for synthesizing a pharmaceutical using an automated radiosynthesizer during a run, the radiosynthesizer having a number of individual activity detectors operable associated therewith, wherein the controller is configured to:

• record activity data from each activity detector;
• detect precursors of yield drop in the recorded activity data based on historic data accessible from a data storage,
• wherein the detecting precursors of yield drop comprises detecting anomalies by processing historic data from multiple earlier runs to identify behavior that give an early warning signal for yield,
• the processing of historic data comprises normalizing data from multiple earlier runs on particular points,
• fitting a mathematical function to the historic data for a selected region corresponding to each activity detector and evaluating the mathematical function relative to the activity data, and
• the mathematical function is y=1−Ae−λt,
• wherein y is the yield, A and λ, are constants and t is time, and the evaluation is based on the magnitude of λ, where a λ, indicates precursors of yield drop;
• predict yield when synthesizing a tracer with the radiosynthesizer based on the detected precursors of yield drop;
• and
• recommend actions related to a level of predicted yield, and
• where the precursor events are indicative of a radiolabelling reaction yield or a solid phase extraction cartridge trapping below a first predetermined threshold, and
• where initiating actions add more material to maintain an output from the radiosynthesizer when the level of predicted yield is lower than a second predetermined threshold.

A method of synthesizing a radiopharmaceutical using an automated radiosynthesizer during a run, the radiosynthesizer having a number of individual activity detectors operably associated therewith, comprising the steps of:

• recording activity data from each activity detector;
• accessing historic data from a data storage;
• detecting precursors of yield drop in the recorded activity data based on the historic data, wherein the detecting precursors of yield drop comprises detecting anomalies by processing historic data from multiple earlier runs to identify behavior that give an early warning signal for yield,
• the processing of historic data comprises normalizing data from multiple earlier runs on particular points,
• fitting a mathematical function to the historic data for a selected region corresponding to the number of individual activity detectors and evaluating the mathematical function relative to the activity data, and
• the mathematical function is y=1−Ae−λt,
• wherein y is the yield, A and λ, are constants and t is time, and the evaluation is based on the magnitude of λ, where a λ, is indicative of precursors in yield drop;
• predicting yield when synthesizing a tracer with the radiosynthesizer based on the detected precursors of yield drop;
• and
• initiating actions related to a level of predicted yield,
• where the precursor events are indicative of a radiolabelling reaction yield or a solid phase extraction cartridge trapping below a first predetermined threshold, and
• the initiating actions maintain an output from the radiosynthesizer based on the level of predicted yield being lower than a second predetermined threshold,
• wherein material is introduced into the automated radiosynthesizer when synthesizing the tracer, and the initiating actions comprise adding more material.

A controller for synthesizing a pharmaceutical using an automated radiosynthesizer during a run, the radiosynthesizer having a number of individual activity detectors operable associated therewith, wherein the controller is configured to:

• record activity data from each activity detector;
• detect precursors of yield drop in the recorded activity data based on historic data accessible from a data storage,
• wherein the detecting precursors of yield drop comprises detecting anomalies by processing historic data from multiple earlier runs to identify behavior that give an early warning signal for yield,
• the processing of historic data comprises normalizing data from multiple earlier runs on particular points,
• fitting a mathematical function to the historic data for a selected region corresponding to each activity detector and evaluating the mathematical function relative to the activity data, and
• the mathematical function is y=1−Ae−λt,
• wherein y is the yield, A and λ, are constants and t is time, and the evaluation is based on the magnitude of λ, where a λ, indicates precursors of yield drop;
• predict yield when synthesizing a tracer with the radiosynthesizer based on the detected precursors of yield drop;
• and
• recommend actions related to a level of predicted yield, and
• where the precursor events are indicative of a radiolabelling reaction yield or a solid phase extraction cartridge trapping below a first predetermined threshold, and
• where initiating actions maintain an output from the radiosynthesizer when the level of predicted yield is lower than a second predetermined threshold, wherein material is introduced into the automated radiosynthesizer when synthesizing the tracer, and the initiating actions comprise adding more material.