Method and apparatus for performing block retrieval on block to be processed of urine sediment image

Patent metadata

Field	Value
Publication number	US-11386340-B2
Application number	US-202016987870-A
Country	US
Kind code	B2
Filing date	Aug 7, 2020
Priority date	Apr 30, 2014
Publication date	Jul 12, 2022
Grant date	Jul 12, 2022

Abstract

The inventive concepts herein relate to performing block retrieval on a block to be processed of a urine sediment image. The method comprises: using a plurality of decision trees to perform block retrieval on the block to be processed, wherein each of the plurality of decision trees comprises a judgment node and a leaf node, and the judgment node judges the block to be processed to make it reach the leaf node by using a block retrieval feature in a block retrieval feature set to form a block retrieval result at the leaf node, and at least two decision trees in the plurality of decision trees are different in structures thereof and/or judgments performed by the judgment nodes thereof by using the block retrieval feature; and integrating the block retrieval results of the plurality of decision trees so as to form a final block retrieval result.

First claim

The invention claimed is:

1. A method for performing block retrieval on a first block to be processed of a urine sediment image, the first block being a portion of the urine sediment image, the method comprising:
using a plurality of decision trees to perform block retrieval on the first block to be processed of the urine sediment image, wherein each of the plurality of decision trees comprises a judgment node and a leaf node, and the judgment node judges the first block to be processed to make it reach the leaf node by using a block retrieval feature in a block retrieval feature set to form a block retrieval result at the leaf node, wherein the block retrieval result comprises an index corresponding to one or more second block stored in a database, the index to the second block belonging to the leaf node, and at least two decision trees in the plurality of decision trees are different in structures thereof and/or judgments performed by the judgment nodes thereof by using the block retrieval feature;
using the indexes of the block retrieval results to retrieve second blocks in the database whereby retrieved second blocks are included in the block retrieval results; and
integrating the block retrieval results of the plurality of decision trees so as to form a final block retrieval result.

2. The method according to claim 1 , characterized in that the step of integrating the block retrieval results of the plurality of decision trees comprises:
voting for the retrieved second blocks by the plurality of decision trees, wherein if there are m decision trees in the plurality of decision trees altogether which retrieve a specific second block, a ballot of the specific second block is m, with m being a positive integer; and
arranging the retrieved second blocks by the plurality of decision trees in a descending order of the ballot.

3. The method according to claim 2 , characterized in that only the retrieved second blocks with ballots greater than a preset threshold value are listed.

4. The method according to claim 1 , characterized in that the step of using a plurality of decision trees to perform block retrieval on the first block to be processed of a urine sediment image comprises: on each decision tree, in response to the first block to be processed being judged by the judgment node and reaching the leaf node, acquiring a second block belonging to the leaf node as a block retrieval result, wherein the second block belonging to the leaf node is set in a manner as follows:
training the plurality of decision trees by using a training sample block in a training sample block set so that on each decision tree, the training sample block is judged by the judgment node and reaches a corresponding leaf node, and becomes the second block belonging to the corresponding leaf node.

5. The method according to claim 4 , characterized in that a classification tag is preset for the training sample block in the training sample block set so that the retrieved second blocks comprised in the block retrieval result also carry classification tags.

6. A device for performing block retrieval on a first block to be processed of a urine sediment image, comprising:
a memory for storing executable instructions, the executable instructions, when executed, implementing the method of claim 1 ; and
a processor for executing the executable instructions.

7. A non-transitory computer readable medium on which an executable instruction is stored, wherein when the executable instruction is executed, a machine is caused to perform the method of claim 1 .

8. An apparatus for performing block retrieval on a first block to be processed of a urine sediment image, comprising:
a block retrieval unit configured to use a plurality of decision trees to perform block retrieval on the first block to be processed of the urine sediment image, wherein each of the plurality of decision trees comprises a judgment node and a leaf node, and the judgment node judges the first block to be processed to make it reach the leaf node by using a block retrieval feature in a block retrieval feature set to form a block retrieval result at the leaf node, wherein the block retrieval result comprises an index corresponding to one or more second block stored in a database, the index to the second block belonging to the leaf node, and at least two decision trees in the plurality of decision trees are different in structures thereof and/or judgments performed by the judgment nodes thereof by using the block retrieval feature, the block retrieval unit configured to use the indexes of the block retrieval results to retrieve second blocks from the database whereby retrieved second block are included in the block retrieval results; and,
an integration unit configured to integrate the block retrieval results of the plurality of decision trees so as to form a final block retrieval result;
wherein the block retrieval unit and the integration unit include a processor and a memory storing an executable instruction.

9. The apparatus according to claim 8 , characterized in that the integration unit is further configured to:
vote for the retrieved second blocks by the plurality of decision trees, wherein if there are m decision trees in the plurality of decision trees altogether which retrieve a specific second block, a ballot of the specific block is m, with m being a positive integer; and
arrange the retrieved second blocks by the plurality of decision trees in a descending order of the ballot.

10. The apparatus according to claim 9 , characterized in that the integration unit is further configured to only list the retrieved second blocks with ballots greater than a preset threshold value.

11. The apparatus according to claim 8 , characterized in that the block retrieval unit is configured to, on each decision tree, in response to the first block to be processed being judged by the judgment node and reaching the leaf node, acquire a second block belonging to the leaf node as a block retrieval result, wherein the second block belonging to the leaf node is set in a manner as follows:
training the plurality of decision trees by using a training sample block in a training sample block set so that on each decision tree, the training sample block is judged by the judgment node and reaches a corresponding leaf node, and becomes a second block belonging to the corresponding leaf node.

12. The apparatus according to claim 11 , characterized in that a classification tag is preset for the training sample block in the training sample block set so that the retrieved second blocks comprised in the block retrieval result also carry classification tags.

Assignees

• Siemens Healthcare Diagnostics Inc
• Siemens Healthcare Diagnostic Inc

Inventors

• Shen Tian
• Xu Juan
• Zhang Xiaofan

Classifications

• G06N5/045 Primary

  Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence · CPC title

• G06N20/00 Primary

  Machine learning · CPC title

• G06F18/24323

  Tree-organised classifiers · CPC title

• G06F16/9027

  Trees · CPC title

• G01N33/493

  urine · CPC title