# US 2023/0067655 A1 · Machine learning enabled detection of infusion pump misloads

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0067655 A1
- Document ID: US20230067655A1
- Status: Published application
- Kind: A1
- Application number: 17/934,686
- Filed: 2022-09-23
- Published: 2023-03-02
- Earliest claimed: 2021-09-24
- Assignees: CareFusion 303, Inc.
- Applicants: CareFusion 303, Inc.
- Inventors: Daniel ABAL; Brendan John Burgess; Ramkumar Subramanian
- Web page: https://patentgrep.com/patent/US20230067655A1

## CPC

- A61M 5/142: Pressure infusion, e.g. using pumps
- A61M 5/16831: {Monitoring, detecting, signalling or eliminating infusion flow anomalies (low-level float-valves causing cut-off A61M5/40; indicating or recording presence, absence or direction of flow in general G01P13/0066)}
- A61M 5/172: electrical or electronic
- A61M 5/31566: {Means improving security or handling thereof}
- G06T 7/001: {using an image reference approach}
- G06T 7/13: Edge detection
- G06T 7/74: {involving reference images or patches}
- G06V 10/147: Details of sensors, e.g. sensor lenses (fingerprint or palmprint sensors G06V40/13; vascular sensors G06V40/145; eye sensors G06V40/19)
- G06V 10/77: Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G16H 20/17: delivered via infusion or injection
- G16H 40/40: for the management of medical equipment or devices, e.g. scheduling maintenance or upgrades
- G16H 40/63: for local operation
- G16H 40/67: for remote operation
- A61M 2205/3306: Optical measuring means
- A61M 2205/52: with memories providing a history of measured variating parameters of apparatus or patient
- G06V 2201/07: Target detection
- A61M: DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY (introducing media into or onto the bodies of animals A61D7/00; means for inserting tampons A61F13/26; devices for administering food or medicines orally A61J; containers for collecting, storing or administering blood or medical fluids A61J1/05); DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY (surgery A61B; chemical aspects of surgical articles A61L); DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR {(Electrotherapy, e.g. producing anaesthesia by the use of alternating or intermittent currents A61N1/36021)}
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

A method may include capturing, by a camera at an infusion pump, one or more images of the pump loaded with an infusion set. A machine learning model may be applied to the images to detect nonconformities that may be present in the images of the pump loaded with the infusion set. Examples of nonconformities include a misload of the intravenous set in which an upper fitment, a lower fitment, and/or a tubing of the intravenous set is misplaced within the pump. In response to an output of the machine learning model indicating a presence of a nonconformity in the one or more images of the pump loaded with the infusion set, a corrective action may be performed. For example, the pump may be prevented from performing an infusion and a message identifying the nonconformities may be generated. Related methods and articles of manufacture are also disclosed.

## Claims

### Claim 1

A system, comprising:
at least one data processor; and
at least one memory storing instructions which, when executed by the at least one data processor, result in operations comprising:
receiving, from a camera at an infusion pump, one or more images of the infusion pump loaded with an infusion set;
applying a machine learning model trained to detect one or more nonconformities present in the one or more images of the infusion pump loaded with the infusion set; and
in response to an output of the machine learning model indicating a presence of a nonconformity in the one or more images of the infusion pump loaded with the infusion set, performing a corrective action.

### Claim 2 (depends on 1)

The system of claim 1, wherein the operations further comprise:
identifying, within the one or more images, a first component of the infusion pump and a second component of the infusion set, the machine learning model detecting the one or more conformities based at least on a relative position of the first component and the second component.

### Claim 3 (depends on 2)

The system of claim 2, wherein the first component and the second component are identified by (i) applying an edge detection technique and/or (ii) based on a graphical feature disposed on each one of the first component and the second component.

### Claim 4 (depends on 2)

The system of claim 2, wherein the first component comprises a bezel, a membrane seal, a door, a platen, a locator feature, or an air-in-line detector, and wherein the second component comprises an upper fitment, a lower fitment, or a pumping segment of a tubing.

### Claim 5 (depends on 1)

The system of claim 1, wherein the one or more nonconformities include at least one of (i) a misload of the infusion set including a misplacement of an upper fitment of the infusion set, a lower fitment of the infusion set, and/or a pumping segment of a tubing of the infusion set, (ii) the intravenous set being reused, past an expiration date, an incorrect type, or a counterfeit product, (iii) one or more components of the infusion pump and/or the intravenous set being missing and/or damaged, and (iv) a presence of contaminants in the infusion pump and/or the intravenous set.

### Claim 6 (depends on 1)

The system of claim 1, wherein the machine learning model detects the one or more nonconformities by at least comparing, to one or more images of a correctly loaded infusion set, the one or more images of the infusion pump loaded with the infusion set.

### Claim 7 (depends on 1)

The system of claim 1, wherein the camera is mounted to a door of the infusion pump, and wherein the camera is mounted in a location where the camera has a field of view that includes at least a portion of the infusion pump loaded with the intravenous set and excludes one or more areas surveillance is unsuitable, prohibited, or unnecessary.

### Claim 8 (depends on 7)

The system of claim 7, wherein the one or more images include a first image captured while the door is in an open position and a second image captured while the door is in a partially open position, and wherein the one or more nonconformities are detected based on the first image and the second image.

### Claim 9 (depends on 1)

The system of claim 1, wherein the camera comprises a visible light camera, an infrared camera, and/or an ultraviolet camera, and wherein the one or more nonconformities are detected based on a graphical feature that is detectable under visible light, infrared light, or ultraviolet light.

### Claim 10 (depends on 1)

The system of claim 1, wherein the corrective action includes at least one of (i) preventing the infusion pump from performing an infusion and (ii) generating a message identifying the one or more nonconformities.

### Claim 11

An infusion pump, comprising:
a bezel having one or more locator features for receiving an intravenous set;
a camera mounted to a door of the infusion pump, the camera being mounted in a location where the camera has a field of view that includes at least a portion of the infusion pump loaded with the intravenous set, and the camera being configured to capture one or more images of the infusion pump loaded with the intravenous set; and
a controller comprising at least data processor and at least one memory storing instructions which, when executed by the at least one data processor, cause the controller to perform operations comprising:
applying a machine learning model trained to detect one or more nonconformities present in the one or more images of the infusion pump loaded with the infusion set; and
in response to an output of the machine learning model indicating a presence of a nonconformity in the one or more images of the infusion pump loaded with the infusion set, performing a corrective action.

### Claim 12 (depends on 11)

The infusion pump of claim 11, wherein the controller is further caused to perform operations comprising:
identifying, within the one or more images, a first component of the infusion pump and a second component of the infusion set, the machine learning model detecting the one or more conformities based at least on a relative position of the first component and the second component.

### Claim 13 (depends on 12)

The infusion pump of claim 12, wherein the first component and the second component are identified by (i) applying an edge detection technique and/or (ii) based on a graphical feature disposed on each one of the first component and the second component.

### Claim 14 (depends on 12)

The infusion pump of claim 12, wherein the first component comprises the bezel, the door, the one or more locator features, a membrane seal, a platen, or an air-in-line detector, and wherein the second component comprises an upper fitment, a lower fitment, or a pumping segment of a tubing.

### Claim 15 (depends on 11)

The infusion pump of claim 11, wherein the one or more nonconformities include at least one of (i) a misload of the infusion set including a misplacement of an upper fitment of the infusion set, a lower fitment of the infusion set, and/or a pumping segment of a tubing of the infusion set, (ii) the intravenous set being reused, past an expiration date, an incorrect type, or a counterfeit product, (iii) one or more components of the infusion pump and/or the intravenous set being missing and/or damaged, and (iv) a presence of contaminants in the infusion pump and/or the intravenous set.

### Claim 16 (depends on 11)

The infusion pump of claim 11, wherein the machine learning model detects the one or more nonconformities by at least comparing, to one or more images of a correctly loaded infusion set, the one or more images of the infusion pump loaded with the infusion set.

### Claim 17 (depends on 11)

The infusion pump of claim 11, wherein the field of view of the camera further excludes one or more areas surveillance is unsuitable, prohibited, or unnecessary.

### Claim 18 (depends on 11)

The infusion pump of claim 11, wherein the one or more images include a first image captured while the door is in an open position and a second image captured while the door is in a partially open position, and wherein the one or more nonconformities are detected based on the first image and the second image.

### Claim 19 (depends on 11)

The infusion pump of claim 11, wherein the camera comprises a visible light camera, an infrared camera, and/or an ultraviolet camera, and wherein the one or more nonconformities are detected based on a graphical feature that is detectable under visible light, infrared light, or ultraviolet light.

### Claim 20 (depends on 11)

The infusion pump of claim 11, wherein the corrective action includes at least one of (i) preventing the infusion pump from performing an infusion and (ii) generating a message identifying the one or more nonconformities.
