Resource scheduling in adaptive radiation therapy planning

Patent metadata

Field	Value
Publication number	US-11633624-B2
Application number	US-201916558509-A
Country	US
Kind code	B2
Filing date	Sep 3, 2019
Priority date	Sep 4, 2018
Publication date	Apr 25, 2023
Grant date	Apr 25, 2023

Abstract

A resource management system for better operation of a plurality of devices. The system comprises an input interface (IN) for receiving input data including one or more characteristics of at least one work object (P1-P3) and/or including context data. The at least one work object (P1-P3) can be processed by one or more processing devices (Mij). The said processing is specified in a respective work specification (S1-S3). A predictor component (PC) of the system is configured to predict, based on the input data, a change to the work specification. The system comprises an output interface (OUT) for providing output data that represents said predicted change.

First claim

The invention claimed is:

1. A resource management system, comprising:
an input port for receiving input data including one or more characteristics of at least one work object and comprising context data, the at least one work object to be processed by one or more processing devices (M ij ), the processing being specified in a respective work specification;
a convolutional neural network (CNN) model created by training input data to predict, based on the input data, a change to the respective work specification;
determine when a plan adaption is required based on a work specification; and
an output port adapted to provide output data that represents the change to the respective work specification, wherein the output data indicate when, during a treatment, the plan adaption will occur.

2. The resource management system of claim 1 , wherein the training of the CNN model further comprises training an initial model based on a set of training data to produce the CNN model.

3. The resource management system of claim 1 , wherein in a deployment mode, the CNN model is provided new data to operate during use of the resource management system.

4. The resource management system of claim 1 , wherein the created CNN model further is repeatedly trained.

5. The resource management system of claim 4 , wherein the repeated training comprises repeatedly training the CNN model with new input data.

6. The resource management system of claim 4 , wherein the respective work specification specifies a period of time for the processing.

7. The resource management system of claim 6 , wherein the output data comprise a probability density or distribution over the period of time, the period of time comprising at least one time interval, the probability density, or distribution indicative of a probability of the predicted change to occur within a given one of the at least one time intervals.

8. The resource management system of claim 4 , wherein the output data comprise a probability density or distribution over a period of time, which comprises at least one time interval, the probability density or distribution indicative of a probability of the predicted change to occur within a given one of the at least one time intervals.

9. The resource management system of claim 4 , wherein the output data comprise an indication of an expected workload, due to the predicted change, for at least one of the one or more processing devices (M ij ) or for at least one category of devices.

10. The resource management system of claim 4 , creating of the CNN model further comprise providing a pre-trained machine learning component.

11. The resource management system of claim 4 , wherein the one or more processing devices comprises any one or a combination of: i) a radiation therapy device, ii) an imaging device, iii) a computer system configured to run a therapy planning algorithm, iv) a computer system configured to run an image segmentation algorithm.

12. The resource management system of claim 1 , wherein there is plurality of work objects, and the resource management system further comprises an accumulator configured to accumulate the output data for the plurality of work objects.

13. The resource management system of claim 1 , wherein the output data includes an indication of an expected workload, due to the predicted change, for at least one of the one or more processing devices or for at least one category of devices.

14. The resource management system of claim 13 , comprising a display configured to effect a visualization of the output data or of the accumulated output data on at least one display device.

15. A method of managing a resource adapted to be implement by a processor, wherein a tangible, non-transitory computer-readable medium stores instructions, which when executed by the processor, causes the processor to:
receive input data including one or more characteristics of at least one work object and comprising context data, the at least one work object to be processed by one or more processing devices (M ij ), the said processing being specified in a respective work specification;
create a convolutional neural network (CNN) model by training input data to predict, based on the input data, a change to the respective work specification;
determine when a plan adaptation is required based on a work specification; and
provide output data that represents the change to the respective work specification, wherein the output data indicate when, during a treatment, the plan adaption will occur.

16. The method of claim 15 , wherein the training of the CNN model further comprises training an initial model based on a set of training data to produce the CNN model.

17. The method of claim 15 , wherein the instructions, when executed by the processor further cause the processor to deploy the CNN model by providing new data to operate during the method.

18. The method of claim 15 , wherein the CNN network is created by repeating the training.

19. The method of claim 18 , where the repeating the training comprises repeating the training with new input data.

20. The method of claim 15 , wherein the output data comprise a probability density or distribution over a period of time, the period of time comprising at least one time interval, the probability density, or distribution indicative of a probability of the predicted change to occur within a given one of the at least one time intervals.

21. The resource management system of claim 1 , wherein the output data further indicate at least which category of the one or more processing devices (M ij ) is involved in the change.

22. The method of claim 15 , wherein the output data further indicate at least which category of the one or more processing devices (M ij ) is involved in the change.

Assignees

• Koninklijke Philips Nv

Inventors

• Heese Harald Sepp
• Vik Torbjoern
• Renisch Steffen
• Bystrov Daniel
• Schulz Heinrich

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• A61N5/1038 Primary

  taking into account previously administered plans applied to the same patient, i.e. adaptive radiotherapy · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• A61N2005/1041

  using a library of previously administered radiation treatment applied to other patients · CPC title